HEALTH SERVICES AND DELIVERY RESEARCH

Transcription

1 HEALTH SERVICES AND DELIVERY RESEARCH VOLUME 5 ISSUE 29 OCTOBER 2017 ISSN Optima NHS service deivery to care homes: a reaist evauation of the features and mechanisms that support effective working for the continuing care of oder peope in residentia settings Caire Goodman, Sue L Davies, Adam L Gordon, Tom Dening, Heather Gage, Juienne Meyer, Justine Schneider, Brian Be, Jake Jordan, Finbarr Martin, Steve Iiffe, Cive Bowman, John RF Gadman, Christina Victor, Andrea Mayrhofer, Meanie Handey and Maria Zubair DOI /hsdr05290

2

3 Optima NHS service deivery to care homes: a reaist evauation of the features and mechanisms that support effective working for the continuing care of oder peope in residentia settings Caire Goodman, 1 * Sue L Davies, 1 Adam L Gordon, 2 Tom Dening, 2 Heather Gage, 3 Juienne Meyer, 4 Justine Schneider, 5 Brian Be, 2 Jake Jordan, 3 Finbarr Martin, 6 Steve Iiffe, 7 Cive Bowman, 4 John RF Gadman, 2 Christina Victor, 8 Andrea Mayrhofer, 1 Meanie Handey 1 and Maria Zubair 2 1 Centre for Research in Primary and Community Care (CRIPACC), University of Hertfordshire, Hatfied, UK 2 Facuty of Medicine and Heath Sciences, University of Nottingham, Nottingham, UK 3 Schoo of Economics, University of Surrey, Guidford, UK 4 Schoo of Heath Sciences, City, University of London, London, UK 5 Schoo of Socioogy and Socia Poicy, University of Nottingham, Nottingham, UK 6 Guy s and St Thomas NHS Foundation Trust, London, UK 7 Research Department of Primary Care and Popuation Heath (PCPH), University Coege London, London, UK 8 Institute of Environment, Heath and Societies, Brune University London, London, UK *Corresponding author Phase 2 of the study Decared competing interests of authors: Caire Goodman is a Senior Investigator at the Nationa Institute for Heath Research. Pubished October 2017 DOI: /hsdr05290

4

5 This report shoud be referenced as foows: Goodman C, Davies SL, Gordon AL, Dening T, Gage H, Meyer J, et a. Optima NHS service deivery to care homes: a reaist evauation of the features and mechanisms that support effective working for the continuing care of oder peope in residentia settings. Heath Serv Deiv Res 2017;5(29).

6

7 Heath Services and Deivery Research ISSN (Print) ISSN (Onine) This journa is a member of and subscribes to the principes of the Committee on Pubication Ethics (COPE) ( Editoria contact: journas.ibrary@nihr.ac.uk The fu HS&DR archive is freey avaiabe to view onine at Print-on-demand copies can be purchased from the report pages of the NIHR Journas Library website: Criteria for incusion in the Heath Services and Deivery Research journa Reports are pubished in Heath Services and Deivery Research (HS&DR) if (1) they have resuted from work for the HS&DR programme or programmes which preceded the HS&DR programme, and (2) they are of a sufficienty high scientific quaity as assessed by the reviewers and editors. HS&DR programme The Heath Services and Deivery Research (HS&DR) programme, part of the Nationa Institute for Heath Research (NIHR), was estabished to fund a broad range of research. It combines the strengths and contributions of two previous NIHR research programmes: the Heath Services Research (HSR) programme and the Service Deivery and Organisation (SDO) programme, which were merged in January The HS&DR programme aims to produce rigorous and reevant evidence on the quaity, access and organisation of heath services incuding costs and outcomes, as we as research on impementation. The programme wi enhance the strategic focus on research that matters to the NHS and is keen to support ambitious evauative research to improve heath services. For more information about the HS&DR programme pease visit the website: This report The research reported in this issue of the journa was funded by the HS&DR programme or one of its preceding programmes as project number 11/1021/02. The contractua start date was in January The fina report began editoria review in Juy 2016 and was accepted for pubication in March The authors have been whoy responsibe for a data coection, anaysis and interpretation, and for writing up their work. The HS&DR editors and production house have tried to ensure the accuracy of the authors report and woud ike to thank the reviewers for their constructive comments on the fina report document. However, they do not accept iabiity for damages or osses arising from materia pubished in this report. This report presents independent research funded by the Nationa Institute for Heath Research (NIHR). The views and opinions expressed by authors in this pubication are those of the authors and do not necessariy refect those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Heath. If there are verbatim quotations incuded in this pubication the views and opinions expressed by the interviewees are those of the interviewees and do not necessariy refect those of the authors, those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Heath. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. Pubished by the NIHR Journas Library ( produced by Prepress Projects Ltd, Perth, Scotand (

8 Heath Services and Deivery Research Editor-in-Chief Professor Jo Rycroft-Maone Professor of Heath Services and Impementation Research, Bangor University, UK NIHR Journas Library Editor-in-Chief Professor Tom Waey Director, NIHR Evauation, Trias and Studies and Director of the EME Programme, UK NIHR Journas Library Editors Professor Ken Stein Chair of HTA and EME Editoria Board and Professor of Pubic Heath, University of Exeter Medica Schoo, UK Professor Andrée Le May Chair of NIHR Journas Library Editoria Group (HS&DR, PGfAR, PHR journas) Dr Martin Ashton-Key Consutant in Pubic Heath Medicine/Consutant Advisor, NETSCC, UK Professor Matthias Beck Chair in Pubic Sector Management and Subject Leader (Management Group), Queen s University Management Schoo, Queen s University Befast, UK Dr Tessa Criy Director, Crysta Bue Consuting Ltd, UK Dr Eugenia Cronin Senior Scientific Advisor, Wessex Institute, UK Dr Peter Davidson Director of the NIHR Dissemination Centre, University of Southampton, UK Ms Tara Lamont Scientific Advisor, NETSCC, UK Dr Catriona McDaid Senior Research Feow, York Trias Unit, Department of Heath Sciences, University of York, UK Professor Wiiam McGuire Professor of Chid Heath, Hu York Medica Schoo, University of York, UK Professor Geoffrey Meads Professor of Webeing Research, University of Winchester, UK Professor John Norrie Chair in Medica Statistics, University of Edinburgh, UK Professor John Powe Consutant Cinica Adviser, Nationa Institute for Heath and Care Exceence (NICE), UK Professor James Raftery Professor of Heath Technoogy Assessment, Wessex Institute, Facuty of Medicine, University of Southampton, UK Dr Rob Riemsma Reviews Manager, Keijnen Systematic Reviews Ltd, UK Professor Heen Roberts Professor of Chid Heath Research, UCL Institute of Chid Heath, UK Professor Jonathan Ross Professor of Sexua Heath and HIV, University Hospita Birmingham, UK Professor Heen Snooks Professor of Heath Services Research, Institute of Life Science, Coege of Medicine, Swansea University, UK Professor Jim Thornton Professor of Obstetrics and Gynaecoogy, Facuty of Medicine and Heath Sciences, University of Nottingham, UK Professor Martin Underwood Director, Warwick Cinica Trias Unit, Warwick Medica Schoo, University of Warwick, UK Pease visit the website for a ist of members of the NIHR Journas Library Board: Editoria contact: journas.ibrary@nihr.ac.uk NIHR Journas Library

9 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Abstract Optima NHS service deivery to care homes: a reaist evauation of the features and mechanisms that support effective working for the continuing care of oder peope in residentia settings Caire Goodman, 1 * Sue L Davies, 1 Adam L Gordon, 2 Tom Dening, 2 Heather Gage, 3 Juienne Meyer, 4 Justine Schneider, 5 Brian Be, 2 Jake Jordan, 3 Finbarr Martin, 6 Steve Iiffe, 7 Cive Bowman, 4 John RF Gadman, 2 Christina Victor, 8 Andrea Mayrhofer, 1 Meanie Handey 1 and Maria Zubair 2 1 Centre for Research in Primary and Community Care (CRIPACC), University of Hertfordshire, Hatfied, UK 2 Facuty of Medicine and Heath Sciences, University of Nottingham, Nottingham, UK 3 Schoo of Economics, University of Surrey, Guidford, UK 4 Schoo of Heath Sciences, City, University of London, London, UK 5 Schoo of Socioogy and Socia Poicy, University of Nottingham, Nottingham, UK 6 Guy s and St Thomas NHS Foundation Trust, London, UK 7 Research Department of Primary Care and Popuation Heath (PCPH), University Coege London, London, UK 8 Institute of Environment, Heath and Societies, Brune University London, London, UK *Corresponding author c.goodman@herts.ac.uk Phase 2 of the study Background: Care homes are the institutiona providers of ong-term care for oder peope. The OPTIMAL study argued that it is probabe that there are key activities within different modes of heath-care provision that are important for residents heath care. Objectives: To understand what works, for whom, why and in what circumstances?. Study questions focused on how different mechanisms within the various modes of service deivery act as the active ingredients associated with positive heath-reated outcomes for care home residents. Methods: Using reaist methods we focused on five outcomes: (1) medication use and review; (2) use of out-of-hours services; (3) hospita admissions, incuding emergency department attendances and ength of hospita stay; (4) resource use; and (5) user satisfaction. Phase 1: interviewed stakehoders and reviewed the evidence to deveop an expanatory theory of what supported good heath-care provision for further testing in phase 2. Phase 2 deveoped a minimum data set of resident characteristics and tracked their care for 12 months. We aso interviewed residents, famiy and staff receiving and providing heath care to residents. The 12 study care homes were ocated on the south coast, the Midands and the east of Engand. Heath-care provision to care homes was distinctive in each site. Findings: Phase 1 found that heath-care provision to care homes is reactive and inequitabe. The reaist review argued that incentives or sanctions, agreed protocos, cinica expertise and structured approaches to assessment and care panning coud support improved heath-reated outcomes; however, to achieve change Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. vii

10 ABSTRACT NHS professionas and care home staff needed to work together from the outset to identify, co-design and impement agreed approaches to heath care. Phase 2 tested this further and found that, athough there were few differences between the sites in residents use of resources, the differences in service integration between the NHS and care homes did refect how these institutions approached activities that supported reationa working. Key to this was how much time NHS staff and care home staff had had to earn how to work together and if the work was seen as egitimate, requiring ongoing investment by commissioners and engagement from practitioners. Residents appreciated the genera practitioner (GP) input and, when supported by other care home-specific NHS services, GPs reported that it was sustainabe and vaued work. Access to dementia expertise, ongoing training and support was essentia to ensure that both NHS and care home staff were equipped to provide appropriate care. Limitations: Findings were constrained by the numbers of residents recruited and retained in phase 2 for the 12 months of data coection. Concusions: NHS services work we with care homes when payments and roe specification endorse the importance of this work at an institutiona eve as we as with individua residents. GP invovement is important but needs additiona support from other services to be sustainabe. A focus on strategies that promote co-design-based approaches between the NHS and care homes has the potentia to improve residents access to and experience of heath care. Funding: The Nationa Institute for Heath Research Heath Services and Deivery Research programme. viii NIHR Journas Library

11 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Contents List of tabes List of figures List of boxes Gossary List of abbreviations Pain Engish summary Scientific summary xiii xvii xix xxi xxiii xxv xxvii Chapter 1 Background 1 Introduction 1 Care homes 1 Heath-care provision to care homes 2 Rationae for the research 2 Aims and objectives 2 Structure of the report 3 Chapter 2 Research approach and methods 5 Study organisation and management 5 Pubic invovement in research 5 User invovement in the study design and research process 5 Users as participants in recruitment with care home residents 6 Anaysis workshop 6 Reaist methods 6 Method 6 Phase 1 6 Admission to hospita, incuding emergency department attendances and ength of hospita stay 7 Use of out-of-hours services 8 Medication use and review 8 User satisfaction 8 Concept mining, scoping of the evidence and deveopment of programme theories 8 Stakehoder interviews 8 Recruitment 9 Interviews 9 Anaysis 10 Scoping of the pubished evidence 10 Theory refinement and testing 12 Anaysis and synthesis 12 Phase 2 13 Ethics approva 13 Samping and recruitment 13 Seecting and recruiting homes for invovement in the study 15 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. ix

12 CONTENTS Recruiting residents from participant homes 15 Sampe size 15 Conducting the case studies 16 Baseine data coection 16 Descriptive case studies of continuing care as deivered to care home residents 16 Care home staff satisfaction surveys (staff outcomes) 17 Resource use outcome measures 18 Anaysis and synthesis 18 Stage 1 18 Stage 2 20 Quantitative data anaysis 20 Considering associations between baseine variabes, costs and outcome variabes 20 Medication anaysis 21 Staff satisfaction 21 Anaysis workshop 22 Chapter 3 Resuts from the review of surveys and the review of reviews 23 Introduction 23 Review of surveys 23 Review of reviews 33 Summary 38 Chapter 4 Reaist synthesis 41 Introduction 41 Stage 1 stakehoder invovement 41 Care home organisation owner and care home representatives 42 Residents and residents representatives accounts 42 Commissioners accounts 43 The accounts of the reguator 43 Inferences from interviews 44 Scoping of the iterature 44 Discussion 52 Concusion 53 Chapter 5 Phase 2 case studies: comparative description of the study sites 55 Introduction 55 Case study sites, recruitment and participant detais 55 Changes in service provision and care home-specific changes across the study 56 Resident recruitment and retention 59 Recruitment of interview subjects 63 Resident characteristics at baseine 66 Services provided at each study site in detai 66 Site 1: age-appropriate care 67 Site 2: incentives, sanctions and targets 70 Site 3: Genera Medica Services pus investment in care home eadership and reationa working within the care home 74 Genera practitioner invovement with care homes across the sites 78 Continuum of association 80 Concusion 82 x NIHR Journas Library

13 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Chapter 6 Phase 2 case studies: case study outcomes of interest 83 Introduction 83 Service use data by site 83 Costs by site 83 Regression anaysis of outcomes taking account of baseine variabes by site 86 Interpreting service use data from the case studies 87 Substudy of unpanned hospitaisations 88 Substudy of medication data 88 Staff satisfaction substudy 90 Quaitative data on the outcomes of interest 92 Use of medications 92 Out-of-hours services 95 Hospitaisation and ength of stay 95 Resident, carer and staff satisfaction 96 Resident behaviour secondary to dementia and its impact on service provision and working reationships 99 Cross-case comparison 101 Concusion 103 Chapter 7 Using the phase 2 case study findings to refine context mechanism outcome configurations and the expanatory framework 105 Introduction 105 Achieving common ground 106 Learning and working 107 Living and dying with dementia 112 Summary 113 Chapter 8 Discussion and concusions 115 Introduction 115 Patterns of service provision 115 Reationa working 116 Importance of genera practitioners 117 Investment in care home-specific work 118 Access to age-appropriate expertise: the case of dementia care 119 Programme theory 120 Strengths and imitations 122 Concusion 123 Impications for practice 124 Recommendations for future research 125 Acknowedgements 127 References 131 Appendix 1 OPTIMAL Study Steering Committee members: 30 August Appendix 2 Screening form for OPTIMAL 147 Appendix 3 Data extraction form used for theory area Appendix 4 Manager summary 155 Appendix 5 Care home ink staff roe 159 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xi

14 CONTENTS Appendix 6 Staff satisfaction questionnaire 161 Appendix 7 Resident service og 163 Appendix 8 Unit cost tabe and references 167 Appendix 9 Incuded studies (for phases 1 and 2) 173 Appendix 10 Resident characteristics at baseine by site and care home 185 Appendix 11 Univariate anayses undertaken during step 1 of Poisson regression 189 Appendix 12 Mutipe hospitaisations 191 Appendix 13 Singe hospitaisations 195 Appendix 14 OPTIMAL fim script: key messages 199 xii NIHR Journas Library

15 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 List of tabes TABLE 1 Summary of generaist provision to care homes from five surveys 24 TABLE 2 Summary of data extracted from topic-specific surveys 28 TABLE 3 Review of reviews ist of incuded studies 34 TABLE 4 Stakehoder interviews 41 TABLE 5 Study site characteristics 55 TABLE 6 Care home recruitment process 56 TABLE 7 Factors that faciitated, and inhibited, the care home recruitment process 56 TABLE 8 Care home characteristics at baseine 57 TABLE 9 Heath-care services accessed by care homes at baseine 58 TABLE 10 Summary of NHS and care home changes identified over the study period 60 TABLE 11 Resident recruitment figures, excusions and those with/without capacity to consent 61 TABLE 12 Resident retention and oss to foow-up 63 TABLE 13 Resident data coected over 12 months from baseine incuding interrai, medication and service use 64 TABLE 14 Participants incuded in interviews and focus groups broken down by site and group 65 TABLE 15 Interviews conducted with HCPs, incuding community nurses and aied heath professionas 65 TABLE 16 Summary of GP research contact across the 3 sites 66 TABLE 17 The GP invovement in care homes 79 TABLE 18 Service contacts by site 84 TABLE 19 Costs by site 85 TABLE 20 Significant predictors of GP contacts: site 1 used as reference category 86 TABLE 21 Significant predictors of primary care contacts: site 1 used as reference category 87 TABLE 22 Significant predictors of out-of-hours contacts: site 1 used as reference category 87 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xiii

16 LIST OF TABLES TABLE 23 Unpanned hospitaisations by site 88 TABLE 24 Reasons for hospitaisations recorded from residents care home notes 89 TABLE 25 Overview of responses by site 90 TABLE 26 How working hours were organised 91 TABLE 27 Breakdown of the average working week 91 TABLE 28 Staff satisfaction with care provided within the care home 92 TABLE 29 Medication review by GP approaches taken across the three sites 93 TABLE 30 Main narratives across the sites arising from HCP interviews on working with care homes (n = 43) 102 TABLE 31 Consutations and services 103 TABLE 32 Revised CMO of the impact of investment in NHS services on resident and service outcomes 109 TABLE 33 Context mechanism outcome care home working within a system of care 111 TABLE 34 Context mechanism outcome of iving and dying with dementia 112 TABLE 35 Unit cost tabe and references 168 TABLE 36 Significant predictors of GP contacts 189 TABLE 37 Significant predictors of primary care contacts 189 TABLE 38 Significant predictors of out-of-hours contacts 189 TABLE 39 Significant predictors of secondary care non-admissions 189 TABLE 40 Significant predictors of ambuance use 190 TABLE 41 Significant predictors of admissions to secondary care 190 TABLE 42 Significant predictors of community care contacts 190 TABLE 43 Significant predictors of tota costs 190 TABLE 44 Site 1: detais of residents hospitaisations with more than one admission (n = 5) 192 TABLE 45 Site 2: detais of residents hospitaisations with more than one admission (n = 2) 192 TABLE 46 Site 3: detais of residents hospitaisations with more than one admission (n = 5) 193 xiv NIHR Journas Library

17 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 TABLE 47 Site 1: detais of residents hospitaisations with one admission ony (n = 9) 196 TABLE 48 Site 2: detais of residents hospitaisations with one admission ony (n = 9) 197 TABLE 49 Site 3: detais of residents hospitaisations with one admission ony (n = 9) 198 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xv

18

19 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 List of figures FIGURE 1 Data synthesis 19 FIGURE 2 Search strategy for stages 1 and 2 of the reaist review 46 FIGURE 3 Overview of heath services provision in site 1 68 FIGURE 4 Overview of heath service provision in site 2 71 FIGURE 5 Overview of heath service provision in site 3 75 FIGURE 6 Dementia nurse speciaist roe and integration with other services 76 FIGURE 7 Continuum of horizonta integration of heath services with care homes across the sites 81 FIGURE 8 Mean ACB score across the three study sites over 12 months 90 FIGURE 9 Expanatory theory for NHS work with care homes 121 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xvii

20

21 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 List of boxes BOX 1 Definition of context, mechanism, outcomes and programme theory 7 BOX 2 Search terms and incusion/excusion criteria for scoping of the iterature 11 BOX 3 Possibe CMO configuration to expain how incentives and sanctions paid to primary care can improve heath care in care homes 47 BOX 4 Possibe CMO configuration to expain how provision of expert practitioners in od age care can improve heath care in care homes 49 BOX 5 Possibe CMO configuration to expain how an intervention designed to improve reationa working achieves improved outcomes for care home residents and staff invoved 50 BOX 6 Exampe of escaation of service use in reation to dementia care 100 BOX 7 Phase 1 emergent programme theory 105 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xix

22

23 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Gossary Setting-specific definitions The foowing are the definitions of a number of terms that are often used interchangeaby. In this report care home is used as the overarching term for a residentia care for oder peope with on-site care services. The key difference between settings is whether or not on-site nursing provision is avaiabe. Care home A residentia setting where a number of oder peope ive, and have access to on-site care services, for exampe meas and persona care (such as hep with washing and eating). It may aso provide nursing care. UK specific: a home registered as a care home wi provide either persona care ony hep with washing, dressing and giving medication and/or care with on-site nursing. Cooquiay, this is often expressed as a residentia care home (i.e. no on-site nursing) or a nursing home. Some care homes are officiay recognised as speciaist providers, for exampe in dementia or paiative care. Long-term care faciity Synonymous with a care home or residentia aged care faciities, a residentia setting that provides on-site care of services designed to meet a person s heath or persona care needs during a short or ong period of time. Nursing home A home with registered nurses who can provide care for more compex heath needs. UK specific: homes registered for nursing care may accept peope who just have persona care needs but who may need nursing care in the future. Residentia aged care faciity A term used in Austraia to describe a faciity for oder peope that offers persona and/or nursing care, as we as accommodation. Synonymous with a care home in a UK context. Other definitions Cinica Commissioning Groups Repaced primary care trusts in Apri 2013 as the commissioners of most services funded by the NHS in Engand. Horizonta integration Integrated care is about bringing together input, deivery, management and organisation of services reated to diagnosis, treatment, care, rehabiitation and heath promotion in order to improve services in reation to access, quaity, user satisfaction and efficiency. Horizonta integration is about inking simiar services that work around and for the patient. Vertica integration is about inking different eves of care ike primary, secondary and tertiary care. Personaised care pans Impemented by genera practitioner practices for the 2% most vunerabe patients on their ist as part of the unpanned admissions Directed Enhanced Service contract worth 2.87 per patient. The personaised care pan is part of a proactive case-management approach for a patients on the register. It incudes detais of their medica history, current medication, preferred pace of care and an agreed pan for escaating care, incuding crisis management, and can be shared with the mutidiscipinary team and other reevant providers with the patient s consent. Reationa working Those activities and processes that emphasise shared decision-making, panning and earning, and continuity of contact between staff from different sectors. Thus reationa working is achieved through a process of mutua recognition of different perspectives. This becomes a means to address the competing priorities, inequity of power and imited resources present when the NHS as a service and individua practitioners work with care homes. It recognises that care homes are distinctive cutura spaces that inform how heath care is understood and prioritised. Froggatt and coeagues (Froggatt K, Hockey J, Parker D, Brazi K. A system ifeword perspective on dying in ong term care settings for oder peope: contested states in contested paces. Heath Pace 2011;17:263 8) characterise reationa working as an ambiguous position between medica and domestic domains of care. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxi

24

25 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 List of abbreviations A&E accident and emergency MAR Medication Administration Record ACB antichoinergic burden MDT mutidiscipinary team ADL activities of daiy iving PIR pubic invovement in research ANOVA anaysis of variance QOF Quaity and Outcomes Framework CCG CI CMO CPN CQC ENRICH Cinica Commissioning Group confidence interva context mechanism outcome community psychiatric nurse Care Quaity Commission Enabing Research in Care Homes sadlh sadlsf scomm scps interrai activities of daiy iving hierarchy scae interrai short activities of daiy iving hierarchy scae interrai communication scae interrai cognitive performance scae GMS Genera Medica Services SD standard deviation GP HCP interrai genera practitioner heath-care professiona internationa Resident Assessment Instrument spain_1 spurs SSC interrai cinica syndrome for pain interrai pressure ucer risk scae Study Steering Committee LTCF ong-term care faciity Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxiii

26

27 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Pain Engish summary Residents in care homes rey on primary heath-care services for access to medica care and speciaist services. The OPTIMAL study asked what features of heath-care provision to care homes are associated with positive outcomes for residents, the NHS and care home staff. First, we reviewed the evidence around working with care homes. Then, to test our deveoping theory of what works we and in what circumstances, we identified three sites that had organised heath care to care homes differenty. One had designated care home teams, one had invested in extra genera practitioner provision to care homes and one had imited care home-specific provision. The care home managers in this third site had aso received eadership training. We recruited four care homes per site and tracked the care that 242 residents received over 12 months. We interviewed residents, famiy members, heath-care professionas and commissioners. The findings indicate that NHS services to care homes shoud ensure that NHS and care home staff have time to discuss, pan and review care together both for individua patients and a residents. Commissioning of heath-care services shoud recognise the importance of this work to the NHS and invest in personne to work with care homes. It takes time for practitioners to earn how to work with care homes. When seeking to reduce demand from care homes on hospitas and other NHS services it is important that ongoing support, particuary for peope with dementia, is sti avaiabe to care homes. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxv

28

29 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Scientific summary Background In Engand, ong-term continuing care for oder peope is principay provided by independenty owned care homes. The care home market is diverse. Across the NHS there are numerous approaches to heath-care provision for this sector, incuding Genera Medica Services (GMS) provided by oca genera practitioner (GP) practices, inked community services, outreach cinics, care home speciaist nurses or support teams, pharmacist-ed services, designated NHS hospita beds and enhanced payment schemes for GPs. The recurrent issue is how to embed and sustain productive patterns of working between heath-care services and providers of ong-term care. The organisationa fux in the NHS and the way in which oca contexts infuence services means that it is unikey that a singe mode of heath service deivery can promote effective working for a care homes. Rather, there wi be key features or expanatory mechanisms, aready manifest within severa modes, that are potentiay appicabe more widey. The research questions were as foows. What is the range of heath service deivery modes designed to maintain care home residents outside hospita? What features (in reaist evauation terms mechanisms ) of these deivery modes are the active ingredients associated with positive outcomes for care home residents? (Modes may incude GMS-inked community services, outreach cinics, community matrons, speciaist nurses or care home support teams, pharmacist-ed services, designated NHS hospita beds and enhanced payment schemes for GPs.) How are these features/mechanisms associated with key outcomes, incuding medication use; use of out-of-hours services; resident, reatives and staff satisfaction; unpanned hospita admissions [incuding accident and emergency (A&E)]; and ength of hospita stay? How are these features/mechanisms associated with costs to the NHS and from a societa perspective? What configuration of these features/mechanisms woud be recommended to promote continuity of care at a reasonabe cost for oder peope resident in care homes? Methods This reaist evauation was organised in two phases. Phase 1 deveoped a theoretica understanding and working propositions of how different contexts and mechanisms infuence how the NHS works with care homes, with reference to five outcomes: (1) medication use and review; (2) use of out-of-hours services; (3) hospita admissions, incuding emergency department attendances and ength of hospita stay; (4) resource use; and (5) user satisfaction. To deveop a preiminary understanding of what supported good heath-care provision to care homes, we competed a scoping of the iterature, which incuded a review of reviews and a survey of types of service provision to care homes. We aso interviewed NHS and oca authority commissioners, providers of services to care homes, representatives from the reguator, care home managers and residents and their famiies. We used these data to deveop theoretica propositions that were further tested in the iterature to expain why an intervention may be effective in some situations and not others. We searched eectronic databases and reated grey iterature. Finay, the findings were reviewed with an externa advisory group. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxvii

30 SCIENTIFIC SUMMARY Phase 2 was a mixed-methods ongitudina case study design. It aimed to deveop further a theory-based expanation of the inter-reationships between the different contexts and mechanisms identified from phase 1 in achieving improved outcomes for residents, the NHS and care home staff. We purposivey recruited 12 care homes from three geographicay disparate study sites. Each organised heath-care support to care homes differenty. This defined the case. Site 1 had invested in care home speciaist teams with expertise in care of oder peope, site 2 had inked care homes to specific GP practices and provided extra funding to support GP invovement and site 3 had imited extra provision for care homes apart from two inked speciaist nursing posts. The care home managers in site 3 had a competed a eadership programme. We tracked the care residents received for 12 months and interviewed care home staff, residents, famiy and visiting heath-care professionas (HCPs) about how they provided and received care, what they perceived was important in supporting residents and how they worked together to achieve good care in reation to our five outcomes of interest. At the end of data coection we conducted an onine survey with care home staff to assess their satisfaction with the heath-care services received. Resuts In phase 1 the review of service provision to care homes incuded 15 surveys of service provision to care homes that had been pubished since 2008 and six reviews on heath-care interventions to care homes. We found imited agreement in the intervention iterature about outcomes, how these shoud be defined or what quaity of care and ife for care home residents ooked ike. The review of surveys found that the variation in the organisation, provision and funding of heath services, both generaist and speciaist, to care homes coud not be expained by resident need or care home type. The wide variabiity in the provision of services to care homes and the widespread ack of denta services signaed that erratic and inadequate care for residents was a persistent feature of heath-care provision to residents in care homes. The 58 stakehoder interviews provided overapping accounts of what was necessary to achieve good heath care. These incuded education and training of care home staff, access to cinica expertise, the use of incentives and sanctions to achieve minimum standards of care, the vaue of champions and designated workers working in and with care homes and the importance of activities that buit robust working reationships between the two sectors. Combining this with the review evidence, and an initia scoping of the iterature, we refined these into propositions to test against the wider evidence. The reaist review findings ed us to propose that it is activities that support and sustain reationa working between care home staff and visiting HCPs that expain the observed differences in how heath-care interventions are accepted and embedded into care home practice. Contextua factors such as financia incentives or sanctions, agreed protocos, cinica expertise and structured approaches to assessment and care panning coud trigger activities that support reationa working. However, these were unikey to be sufficient to achieve change if they did not ead to visiting HCPs and care home staff working together to identify, pan and impement care home-appropriate protocos for care. This expanatory theory was the starting point and putative expanation of what enabed heath-care services to work we with care homes that we sought to test and refine in phase 2. In phase 2 the three sites organised heath care to care homes in different ways. Site 1 emphasised speciaised care of oder peope, working in partnership with care homes. This was characterised by mutipe mutidiscipinary teams that either worked excusivey with care homes or had expicit responsibiity for care homes as part of their work. A nurse-ed care home service had been in pace for 15 years, which incuded the case management of new residents. Forma and informa systems for team-to-team referras about specific residents incuded access to a speciaist dementia outreach team. xxviii NIHR Journas Library

31 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Site 2 emphasised incentives and sanctions and service deivery was characterised by a focus on GPs as co-ordinators of heath-care services provided to care homes. Specific GP practices received extra payments to work with care homes and homes were asked to register their residents with one of these. Structured training for care homes was being introduced to equip staff with the knowedge and skis to provide care for residents with compex needs and reduce unpanned hospitaisations. Competion of training meant that a care home was eigibe to receive additiona payments. Site 2 had some eements of speciaised services for oder peope. There was a nurse speciaist in paiative care designated to care home residents, and there were two other services avaiabe to care homes (but not specificay targeted towards them) a team of nurses and therapists, and a dementia advice and support service. The overa emphasis of provision was on services for individua residents. Site 3 heath-care provision was characterised by services that did not differentiate between oder peope iving in their own homes and those iving in care homes. Individua expert practitioners with competencies reevant to the management of care home residents, for exampe tissue viabiity and cardiac nurse speciaists, received referras through separate routes. For some but not a the care homes there was one care home nurse speciaist to respond to acute deteriorations in residents to prevent admission to hospita, and one dementia care speciaist nurse. A the care home managers had received eadership training from a charity focused on deivering positive change in care homes for oder peope. In tota, 242 residents were recruited across the three sites and 181 interviews were competed with residents, reatives, HCPs and care home staff. The resident cohort was representative of UK care homes generay in terms of the prescribing rates seen. Across the three sites, 83 participants were ost to the study through death and three were transferred to other care settings. Most residents had infrequent use of many types of heath services; GPs were the most heaviy used group, with over 90% of residents having some eve of GP contact in each site. For the most part there was no compeing difference in service use, or costs, between sites. Site 3 might have been expected to have been substantiay cheaper, given that the cohort recruited here was substantiay ess dependent; however, this was not in fact the case. It aso had a greater number of secondary care non-admitted contacts, as we as a trend towards higher costs associated with hospita admissions. This may indicate a tendency to refer residents into hospita, rather than provide care in situ. Site 1 was expected to be substantiay more expensive because of routiney using more speciaist care, but this was not the case. A descriptive anaysis of unpanned admissions found that 39 residents were hospitaised at some point during the 12-month data coection period, just 16% of the tota number of residents recruited to the study. The ength of stay ranged between one night (n = 17) and 47 nights for one case invoving a dementia-reated menta heath assessment, with 22 residents being hospitaised for more than five nights in one episode. These support the findings from the quantitative anaysis of a greater reiance upon secondary care in site 3 and they aso highight the tendency for patients to stay much onger in hospita in this site. Over the study period, there were 366, 261 and 266 medication changes representing 0.40, 0.44 and 0.49 changes per resident per month in sites 1, 2 and 3, respectivey. There were no consistent trends in antichoinergic burden scores, antibiotic or opioid prescribing. Heath-care practitioners across a three sites identified common issues with medication management, incuding concerns about care home staff knowedge of pharmacoogy, difficuties of prescribing for wound management, the chaenges of mutipe prescribers visiting care homes and the importance of access to emergency end-of-ife medication. When GPs hed reguar cinics in the care homes, there were few or no references to difficuties in securing prescriptions and reviews were conducted more frequenty. There were aso higher eves of care home staff satisfaction with access to heath care in those sites where GP cinics were offered. A ack of pharmacist invovement in medication reviews was highighted as a gap in service provision across the sites. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxix

32 SCIENTIFIC SUMMARY A HCPs identified avoidance of unnecessary hospitaisations as an important part of their invovement with care homes. However, apart from the GPs, a of the care home services worked office hours, and out-of-hours service provision did not aways fit around the needs of oder peope iving with dementia in care homes. An exampe across a three sites was that residents had to be admitted to the emergency department at the weekend if they needed a psychiatric opinion. Some out-of-hours services were perceived as having negative attitudes towards care homes and staff and were sometimes described as ignoring care pans put in pace by teams providing more routine support to care homes during the working week. Care home residents were perceived to be a ow priority for out-of-hours and emergency services. There were differences between the sites in how care home and NHS staff described working together, how care homes were represented as providers of care to oder peope, and the abiity of services to engage with the issues and care needs of peope iving and dying with dementia. Across a sites, six contexts were identified as being key to how care was organised and operationaised. These were the system for referras, avaiabiity of dedicated heath services for care homes, team working, the use of case management, care home-based training and the ength of time NHS services and practitioners had worked with care homes and each other. The synthesis of the two phases of work drew on both positive and negative exampes of NHS support to care homes to test and deveop an expanatory theory of what works when and in what circumstances, which has the foowing eements. Reationa working Supporting (incentivising) the right mix of peope to be invoved in the design of heath-care provision to care homes supports reationa working. Having discussions before setting up a service and using shared protocos, guidance and reguar meetings (context) prompts co-design and aignment of heath-care provision with the goas of care home staff and a shared view about what needs to be done. This creates opportunities for joint review and anticipation of residents needs, incuding medication and retaining residents with compex care needs in the care home (outcome). At an institutiona eve, the case studies suggested that activities that inked NHS services around the care home as we as with the care home were important. The organisation and funding of NHS services to care homes in the three sites refected a continuum of association that in part showed how reationa patterns of working had deveoped over time. Thus, the focus and content of work-based decisions were faciitated and/or inhibited by reationships between visiting HCPs and care home staff. These acted as a source of infuence on the nature and expression of heath-reated interests and vaues in conjunction with individuas differences and ength of association. How the different services were organised around the care homes affected the eve of horizonta integration achieved. An expicit (funded) commitment to spend time working with care homes was more ikey to foster reationships and confidence that residents coud access services as needed. This was especiay true when HCPs working with care homes were inked to other NHS services and their care home work was recognised by these services as important. These patterns of working and visiting created naturay occurring opportunities to meet and discuss care, and nurtured a mutua appreciation of the chaenges both NHS and care home staff faced each day. There was some evidence that it fostered access to a wider array of services, freed up GPs to focus on GMS tasks and enabed an approximation of care/case management, even when such roes were not made expicit. We found itte evidence, however, apart from adjusting times of visiting and improving access, of NHS services organising services to accommodate care home staff or residents priorities. Where individua HCPs invoved other services on a resident-by-resident basis, the frequency and intensity of their invovement were at their discretion and was often shaped by the demands of their wider caseoad. xxx NIHR Journas Library

33 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Importance of genera practitioners The invovement of the GP was important, even if other services had absorbed some of their activities such as medication review, responsive care and case management. Services that provided intensive care home support, through a mode of reationa working, sti needed inks to GPs, for diagnosis, urgent care and discussions about unresoved issues of care. This was aso reated to how the working reationships between secondary care, care home staff and visiting NHS services were organised. Investment in care home-specific work as part of a system of care Commissioning severa NHS services to work with care homes on a reguar and ongoing basis creates a network of expertise in the care of oder peope (context) and increases the confidence of NHS staff and their abiity to refer residents and review care to adapt patterns of service deivery (mechanism). This can improve residents access to care and reduce demand on urgent and emergency care services (outcomes). Where there is a narrow focus on care homes as a drain on NHS resources, commissioners and practitioners focus on short-term interventions and measuring outcomes in terms of what had not happened and how resources had not been used. This does not foster reationa approaches to working together. At a practitioner eve, forma acknowedgement that working with care homes was important and vaued work had a egitimising function that gave NHS staff permission to engage with care homes. When heath-care provision is funded to work with care homes on a reguar basis and services have deveoped over time (contexts), and practitioners see this as a egitimate and manageabe use of their time and skis, staff and services are more ikey to deveop ways of working that seek to ink residents with other services and work with care home staff to resove probems (mechanisms). This can ead to improved access to NHS services, crises avoided and care home staff and resident satisfaction with heath-care provision (outcomes). Access to age-appropriate care (dementia) Phase 1 identified access to age-appropriate cinica assessment and care as an important context. Phase 2 supported this inasmuch as it found that pain, pressure ucer prevaence, medication use and comorbidities were predictors of increased heath service utiisation among care home residents. Access to NHS expertise in dementia care is particuary important. We found that the greater the severity of cognitive impairment, the ess ikey it was that a resident woud see a primary care professiona. The presence of dementia compicated care provision and not a services coud easiy dea with this compexity. Quaitative accounts from NHS staff described how difficut they found visiting residents with dementia, notaby where there was no ready access to speciaist dementia services. If NHS and care home staff have access to dementia expertise when addressing residents behaviours that they find chaenging (context) then they have confidence and skis in providing care (mechanism) that reduces the need for antipsychotic prescribing and the distress of residents (outcome). Concusions A theory of commissioning for heath-care provision to care homes proposes that NHS services are more ikey to work we with care homes when payments and roe specification endorse staff working with care homes at an institutiona eve as we as with individua residents. Integra to such endorsement is a recognition of the vaue of supporting activities that, over time, enabe NHS staff and care home staff to co-design how they work together to improve residents heath care. Commissioning arrangements shoud aso consider how services are organised around the care home. This worked we when it incuded expertise in dementia care, the GP as part of the care deivery team and Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxxi

34 SCIENTIFIC SUMMARY access to a wider array of services, and enabed an approximation of care/case management, even when such roes were not made expicit. Impications for practice There are mutipe ways that the NHS works with care homes and it is unikey that there is one right way of working or mode of service deivery. The foowing impications are directy reated to the different eements of the programme theory. When commissioning and panning NHS service provision to care homes it is important that residents in care homes have access to heath care that is equitabe and equivaent to those received by oder peope iving at home. We found that service provision to care homes is often ad hoc and reactive and that some services (e.g. denta heath care, speech and anguage therapy) were either not offered to a care homes or imited in scope. Heath-care professionas work with care homes shoud be formay recognised by NHS managers as key to the support of integrated working for oder peope. Recognised referra inks with other community and hospita services are more ikey to support continuity of care and management of acute episodes in the care home. Where a care home service is a stand-aone service or an adjunct to an existing roe without protected time, practitioners can strugge to co-ordinate residents care and invove NHS services when needed. Investment and incentives to NHS services and practitioners working with care homes shoud be structured to support joint working and panning before services are changed or modified. Where funding and sanctions are designed to reduce inappropriate demand on secondary care and other NHS services this can have the unintended consequence of focusing on faiure. The study found that funding to support care home teams and GPs to have more time to earn how to work with care homes and identify shared priorities and training needs was more ikey to faciitate co-operation, affirm best practice and motivate staff to find shared heath-care soutions. Care home providers referra guidance needs to fit with NHS referra protocos together with opportunities for diaogue where they are uncertain about how to identify different NHS services. The study found that care home staff were often unsure who to invove when they were concerned about a resident. Estabished reationships that had deveoped over time between care home staff and HCPs were aso observed to faciitate appropriate referras that in turn heped to reinforce best practice. Care home-based training needs to incude a care home staff working with residents, not just the nurses or senior carers, and support them to work with the NHS and communicate with famiy carers. New care home staff in particuar need support from NHS staff when working with residents and understanding their heath-care needs. The study findings suggested that when training incuded a members of the workforce (e.g. catering staff and junior staff) there was more ikey to be engagement at an organisationa eve and sustained impementation of service improvements. Genera practitioners need to pay a centra roe in residents heath care. How their work compements other care home-focused services shoud be specified and agreed between a those invoved in assessing and treating residents, and making referras. Reguar GP cinics or patterns of visiting that were predictabe were associated with higher eves of care home staff satisfaction with heath care, fewer medication-reated probems and more frequent medication reviews. This was particuary true when there was an opportunity to discuss care provision across the care home and not just individua residents heath care. Dementia expertise needs to be integra to reguar service provision, not part of a separate service. The study found that both care home and NHS staff coud benefit from ongoing access to training and resources to equip them to support residents iving with dementia. Care home staff pay a vita roe in managing and monitoring residents medication, but may need further training and support in this area. The study found that this was an aspect of care that was of particuar concern to both residents and their reatives. xxxii NIHR Journas Library

35 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Recommendations for future research Our recommendations for future research reate both to aspects of research methods and to a number of research questions to further evauate and expicate our programme theory. We concude from the findings that there is imited vaue in further descriptive work on NHS heath-care service provision to care homes that is not inked to an understanding of how the services work with care home staff to improve care home residents heath-reated outcomes. There is an urgent need for research that can deveop and refine a minimum data set for residents that can ink with heath and socia care patient/cient data systems. This study found imited evidence of care home residents, staff or famiies infuencing or shaping how or what kind of heath-care support was provided. Further research is needed that can buid on the principes of reationa working and co-design to test different ways of supporting their meaningfu participation. We found very itte evidence of how famiy members contribute to or monitor the heath care that their reatives receive. There is a need for further research to understand how their knowedge of the resident and their insights might inform care. Research on how training and deveopment in dementia care across the NHS and socia care workforce (and not just care home staff) can improve the quaity of care of peope iving and dying with dementia. Funding Funding for this study was provided by the Heath Services and Deivery Research programme of the Nationa Institute for Heath Research. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxxiii

36

37 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Chapter 1 Background Introduction Oder peope iving in care homes are some of the odest and fraiest in society. They have entered a care home because they can no onger ive in their own homes. 1 Care homes provide 24-hour persona care, and some care homes provide nursing care; however, residents sti rey on primary heath-care services for access to medica care and referra to speciaist services. The reationship between the NHS and care homes is a symbiotic one. 2 5 Care homes, as independent providers, are the main providers of ong-term care for oder peope and, increasingy, respite and end-of-ife care. The majority of care home residents have cognitive impairment, mutipe morbidities and compex care needs defined by high functiona dependency and unpredictabe cinica trajectories. 6 8 In this context, good heath outcomes depend on effective day-to-day socia care and vice versa. Despite this, how the NHS works with care homes is variabe and often inequitabe. 9,10 A number of different modes of care provision have deveoped to address the identified inequity. This study considers what eements or characteristics of these different services support residents heath and maintain efficient and effective working between NHS services and care homes. Care homes Approximatey 433,000 oder or physicay disabed peope ive in care homes in the UK, with 90% of residentia and nursing care services now deivered by independent providers. A care home can offer persona care and 24-hour support (previousy caed residentia homes ), on-site nursing in addition to this (previousy caed nursing homes ) or both types of care (sometimes referred to as dua registered homes ). Care home residents account for 4% of the popuation aged 65 years or oder. 11,12 There are over three times as many care home beds as there are acute hospita beds in the UK, and approximatey 10% of care home residents receive funding from the NHS. For the majority of residents their care is either sef-funded, paid for from the state socia care budget or via a mix of state socia care funding with top-up from residents or their famiies. 12,13 Care homes are heterogeneous in terms of how these different funding sources make up their income and in how they structure themseves as businesses. A report described the sector as a highy poarised marketpace. 12 Providers that focus on private payers are reativey financiay secure, whie those reiant on pubic funding are vunerabe to the government s austerity measures, financia osses and threat of cosure. 12 The Burstow Commission on Residentia Care 14 found that ony one in four peope woud consider moving into a care home if they became fraier in ater ife, whie 43% said that they woud definitey not move. Care homes were represented by many as an accommodation of ast resort. The Commission argued that negative media coverage of care homes, despite many exampes of innovative high-quaity care for peope with compex needs and dementia, has an impact on how staff and managers fee about their jobs and how their work is vaued by wider society. Most care home residents are femae, over 85 years od and in the ast years of ife. The majority of care home residents have dementia, are in receipt of seven or more medications and a significant proportion ive with depression, mobiity probems, incontinence and pain. 6,9,15 18 They are a popuation that needs access to heath care and ongoing review. The common perception that care homes are a probem to the Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 1

38 BACKGROUND NHS is open to chaenge. In addition to being the main providers of ong-term care for oder peope, care homes provide mutipe services to the NHS, incuding respite care, intermediate care and re-enabement services. Admission to emergency departments and acute hospitas from care homes may, contrary to the pubic narrative, be as much a consequence of how primary and emergency heath-care services respond to cas for hep from the care home sector as they are a refection of care decisions by care home staff. 19 Residents cose proximity to the end of ife, however, provides an opportunity to estabish advanced care pans enabing care homes to pay an active roe in responding to heath concerns in situ, avoiding emergency hospita admissions as a consequence. 20 Heath-care provision to care homes There is a ack of shared understanding about what represents an idea package of care that shoud be provided by the NHS to care homes Aspects of care reated to the management of heath probems are often undertaken by care home staff, whether or not they are quaified nurses. 25 These incude non-pharmacoogica management of behavioura and psychoogica symptoms of dementia, monitoring the impact of pharmacoogica and non-pharmacoogica therapies, doing routine dressings and administering compex drug regimens. These arrangements are usuay informay negotiated and vary between different homes and parts of the country. As a consequence, the extent to which the opportunity and financia costs of such heath-care interventions are borne by the care home sector vary between regions. Effective working between the heath and residentia care sectors is fundamenta to residents quaity of ife and may infuence how often residents are admitted to hospita and how ong they stay there once admitted. But modes of service deivery to care homes are many and i defined. 26,27 Services at the interface between care homes and the NHS often have differing goas and funding sources, and operate in diverse ways. Athough most regard integrated working as a vita objective, definitions of integrated care differ and few interventions to improve heath-care deivery have been deveoped in coaboration with care home staff, residents and their famiies. Primary care services are frequenty deivered from a distance and are reiant on how care home staff interpret residents heath status. Inherent tensions can deveop when NHS services favour modes of care that focus on diagnosis, treatment and episodic invovement, whie care home providers prioritise ongoing support and reationships that foster continuous review of care. 19 How to estabish effective integrated working, and the modes of service deivery that coud faciitate this, remain uncear. Rationae for the research We have described a heterogeneous care home market and a range of context-sensitive variabes that shape how services are provided. Cumuativey, these make it unikey that a singe mode of heath service deivery can promote effective working for a care homes and at a times. If there are generaisabe patterns that underpin effective modes of care, it is more ikey that these wi be at the eve of recurrent features or expanatory mechanisms aready manifested within mutipe service modes and potentiay appicabe across mutipe modes in the future. As Pawson et a. 28 have noted, much that is effective in heath-care deivery is submerged, routine and taken for granted. Identifying these features and making them expicit is key to deivering effective care. Aims and objectives This study set out to identify, map and test the features or expanatory mechanisms of existing approaches to heath-care provision to care homes in reation to five key outcomes: (1) medication use and review; (2) use of out-of-hours services; (3) hospita admissions, incuding emergency department attendances and ength of hospita stay; (4) resource use; and (5) user satisfaction. 2 NIHR Journas Library

39 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 The overa aim of this study was to use a theory-driven reaist evauation approach 29 to identify ways in which the deivery of existing NHS services to care homes may be optimised for the ongoing benefit of residents, reatives and staff, and the best use of NHS resources. It addressed the foowing research questions. What is the range of heath service deivery modes designed to maintain care home residents outside hospita? What features (in reaist evauation terms: mechanisms) of these deivery modes are the active ingredients associated with positive outcomes for care home residents? How are these features/mechanisms associated with key outcomes, incuding medication use; use of out-of-hours services; resident, carer and staff satisfaction; unpanned hospita admissions (incuding A&E); and ength of hospita stay? How are these features/mechanisms associated with costs to the NHS and from a societa perspective? What configuration of these features/mechanisms woud be recommended to promote continuity of care at a reasonabe cost for oder peope resident in care homes? Structure of the report Chapter 1 describes the background and the rationae for the study. Chapter 2 describes the research approach and methods, providing detai about the study design, data coection and anaysis. Chapter 3 presents the findings of the review of surveys of heath-care provision to care homes and the review of reviews. This is foowed by the reaist synthesis of heath-care provision to care homes in Chapter 4. Chapter 5 introduces phase 2, with detai about the case study sites recruitment, participant characteristics and the organisation of heath care in each site. Chapter 6 summarises the case study findings on care home residents service use and reated costs, medication use and staff satisfaction. Chapter 7 revisits the findings of phase 1 and, based on phase 2 findings, presents context mechanism outcomes (CMOs), which capture how heath-care services work (or not) with care homes. Chapter 8 discusses the findings and their impications for commissioning and the organisation and provision of NHS services to care homes. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 3

40

41 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Chapter 2 Research approach and methods This study buit on earier descriptive work that had mapped how the NHS works with care homes without on-site nursing provision. 2 This chapter provides a brief overview of how the study was organised and managed. It aso gives the rationae for using reaist-driven approaches to evidence synthesis and evauation in order to answer the research questions and to move beyond descriptive accounts of the NHS working with care homes. It describes the two phases of the study, data coection and anaysis and, finay, notes the changes that were made to the originay funded protoco. This chapter is compemented by two pubished protocos on phases 1 and 2. 1,30 Study organisation and management The study was overseen by a management group made up of the researchers and the study research team, which met four times a year, and a Study Steering Committee (SSC) that met twice a year. The overa roe of the SSC was to ensure that the study was conducted in ine with the protoco, and that the design, execution and findings were vaid and appropriate for care home residents, reatives and the organisations invoved in their care. A ist of members and their reevant expertise is provided in the appendices (see Appendix 1). Specificay, the study steering group were asked to do the foowing: provide expert advice and guidance on a aspects of the study; individua members provided expertise for the different study phases ensure that the project was running according to the time schedue address any identified risks within the study and ensure that the appropriate procedures were in pace to miitate against these contribute to a discussion of any issues arising from either the conduct or anaysis of the study debate the emergent theoretica propositions from phase 1 of what supported heath-care working with care homes and the emergent findings from phase 2 read and comment on any reports and other reevant study outcomes act as a ink between the project and other reated research studies, NHS and charitabe organisations interested in the way that care homes work together with the NHS. Pubic invovement in research Pubic invovement in research (PIR) was integrated into the study from project design and management to dissemination. This was achieved through PIR review of the study design and research process, PIR support with resident recruitment and feedback on emergent findings presented at the SSC meetings and the anaysis workshop. Pubic invovement in research members with direct experience of visiting cose reatives and friends over ong periods of time (years) in care homes were recruited through two estabished university patient and pubic invovement groups. One member had supported recruitment of care home residents in a previous study. Her roe had been to spend time taking with those residents who wanted more time to tak about their invovement in the study. User invovement in the study design and research process Members of the PIR group at the University of Hertfordshire (John Wimott and Marion Cowe) and the University of Nottingham (Kate Sartain and Michae Osborn) were invoved in the deveopment of the study Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 5

42 RESEARCH APPROACH AND METHODS proposa and were aso asked to review resident information bookets, summary and consent forms. PIR members aso attended the SSC meetings. Users as participants in recruitment with care home residents Pubic invovement in research members who had an honorary contract with the university were invoved in the support of resident recruitment in the care homes. PIR members were aso invoved in the deveopment of a study information video for care home residents. Anaysis workshop A member of the PIR group (Kate Sartain) activey participated in a 2-day anaysis workshop together with the study management group where the emergent findings from the study sites were discussed. Reaist methods Reaist methods are based on a theory-driven approach to evidence review and evauation that argues that reaity is objective and knowabe but interpreted through cognition and senses. To expain why interventions work, these methods seek to identify the underying mechanisms that can eucidate how different outcomes are obtained and how contexts infuenced this process. 28,29,31 We defined heath-care provision to care homes as a series of compex socia processes invoving mutipe contributors over extended periods of time, where uptake and use of resources can vary widey depending on residents needs, organisationa structures and oca resources. Thinking of service deivery to care homes in this way enabed us to consider the heterogeneity of approaches used and consider the mutipicity of conditions in which they are enacted to provide an expanatory account of how one approach may work, when, for whom and why. It aowed us to go beyond descriptive accounts of the organisation of care, and the perceived barriers to and enabers of this, to provide pausibe, evidenced expanations of observed outcomes and the mechanisms associated with these, whie acknowedging and expaining the infuence of context. We conceptuaised different approaches to heath-care provision to care homes as programmes that can be deconstructed to understand how key eements or factors in their working (mechanisms) may trigger a change or effect (outcome), and which contextua conditions or resources (context) are necessary to sustain changes. Box 1 describes how context (C), mechanism (M), outcomes (O) and programme theory as the anaytica toos of reaist approaches were operationaised for the purposes of this study. This reaist evauation incuded a reaist synthesis as part of the process of deveoping and refining programme theory. Integrating different forms of knowedge (using both primary and secondary sources) to expain compex phenomena in this way is consistent with a reaist understanding of research. It enabes us to identify those contextua factors that are necessary across a range of interventions to trigger the desired mechanisms. Method Phase 1 This was designed to address questions 1 and 2 as outined in Chapter 1, which were: 1. What is the range of heath service deivery modes designed to maintain care home residents outside hospita? 2. What features (in reaist evauation terms: mechanisms) of these deivery modes are the active ingredients associated with positive outcomes for care home residents? 6 NIHR Journas Library

43 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 BOX 1 Definition of context, mechanism, outcomes and programme theory Context (C) Context can be broady understood as any condition that triggers and/or modifies the behaviour of a mechanism, that is, the backdrop conditions (which may change over time). For exampe, education and quaifications of care home staff, history of working reationships between visiting HCPs and care home staff and residents functiona abiities. Mechanism (M) A mechanism is the generative force that eads to outcomes. Often denotes the reasoning (cognitive or emotiona) of the various actors, that is, care home staff, residents, reatives and visiting HCPs. Identifying the mechanisms goes beyond describing what happened to theorising why it happened, for whom, and under what circumstances. Outcomes (O) Intervention outcomes, for exampe a reduction in episodes of unpanned hospita admissions, medication management, staff confidence and costs. Programme theory Specifies what mechanisms are associated with which outcomes and what features of the context wi affect whether or not those mechanisms operate. The programme theory encapsuates ideas about what needs to be changed or improved in how NHS services work with care home staff, and what needs to be in pace to achieve an improvement in residents heath and organisations use of resources. HCP, heath-care professiona. Pawson R, Greenhagh T, Harvey G, Washe K. Reaist review a new method of systematic review designed for compex poicy interventions. J Heath Serv Res Poicy 2005;10: Wong G, Greenhagh T, Westhorp G, Pawson R. Reaist methods in medica education research: what are they and what can they contribute? Med Educ 2012;46: Phase 1 competed a reaist review of existing evidence to deveop a theoretica understanding and working propositions of how different contexts and mechanisms infuence how the NHS works with care homes, paying specific attention to five outcomes of interest that coud then be refined and tested in the case studies that comprised phase 2. The five outcomes of interest were identified by the research team as consistent with service priorities across the NHS, care home and oca authority organisations. These were agreed through consutation with the SSC and the stakehoder organisations that they represented. These were as foows. Admission to hospita, incuding emergency department attendances and ength of hospita stay The extent to which residents are enabed to receive care in situ in the care home can refect both care home staff confidence and how they are abe to access services, support and guidance from heath-care services. Repeated admissions, particuary in the context of ambuatory care-sensitive conditions or towards the end of ife where they might be regarded as inappropriate, can be avoided where proactive coaborative advanced cinica panning is embedded within systems of care. 19,20,34 36 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 7

44 RESEARCH APPROACH AND METHODS Often the decision to admit an oder person to hospita is appropriate and cannot be avoided. However, the ength of hospita stay is infuenced by how easy it is to discharge the oder person to the care home, which in turn is infuenced by the reationship between care homes and primary care services, and the reationship of both with secondary care. 37 Use of out-of-hours services Use of out-of-hours services can be an indication of the eve of anticipatory care, joint panning, day-to-day NHS support received and care home staff capacity, confidence and abiity to dea with residents unexpected heath-care needs. For care homes, the quaity of advice and support they receive out of hours, often over the teephone, can infuence decisions to support a person in the care home or ca an ambuance. 38 Medication use and review The majority of care home residents take seven or more medications. 1,17 Evidence suggests that residents are vunerabe to prescribing and administration errors and that review of medication, using agreed criteria, 17 can improve the quaity of prescribing and medication use. Reguar review can aso highight other issues and act as a focus for making proactive decisions about care. User satisfaction Oder peope, incuding those with cognitive probems, can express what is important to them in their heath care, their preferences for who ese is invoved in discussing heath-care decisions and who shoud take responsibiity for the day-to-day management of their heath care. 39 Satisfaction with care in this setting needs to incude the mutipe perspectives of residents, famiy members and care home staff as recipients of heath-care services. 40 The reaist review took an iterative three-stage approach and was structured in ine with Reaist And MEta-narrative Evidence Syntheses: Evoving Standards (RAMESES) guidance on the organisation and reporting of reaist syntheses. 31 First, scoping searches and stakehoder interviews were used to identify sources of poicy, egisative and professiona thinking that coud hep to expain how heath-care services and care homes work with each other. Second, the findings were used to deveop theoretica propositions that coud be tested using the iterature on heath-care provision to care homes, in order to expain why an intervention may be effective in some situations and not others. Third, the findings were reviewed with our study steering group. We have pubished the phase 1 protoco. 30 Concept mining, scoping of the evidence and deveopment of programme theories To gain a preiminary understanding of what supported good heath-care provision to care homes, we conducted a series of stakehoder interviews with key informants invoved in the commissioning, provision and reguation of heath care to care homes, as we as recipients of care (residents and reatives). This was foowed by a review of surveys of heath-care services provided to care homes, a review of reviews on care home interventions and a suppementary scoping iterature review to begin to identify further the underying assumptions and theories of what supported effective working in care homes. Stakehoder interviews The purpose of the interviews was to hep inform and refine the focus of the evidence review, carify terms, identify key headings or theory areas and inked questions that shoud be asked in the deveopment of data extraction forms in the evidence review. 41,42 A more detaied account of the method and findings is pubished esewhere NIHR Journas Library

45 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 The interviews expored a number of areas of uncertainty. The priority that the NHS paces on cost management, appropriate use of resources and service efficiency is we known. There is, however, ess carity about the eve of evidence commissioners require in order to make judgements about services to care homes and how to measure effectiveness when working in and with care homes. It is aso uncear how contexts of care (e.g. history of provision, size of care homes, eadership, care homes with on-site nursing and those without) infuence demand on NHS services. Finay, it is uncear what care homes and their representatives, residents and reatives recognise as constituting effective heath care. For the purposes of this study, a stakehoder was defined as someone who had the reevant experience or knowedge to be abe to express the view of the group or organisation that they represented. 44 Consequenty, we seected individuas who either had responsibiity for the commissioning, organisation or monitoring of NHS provision to care homes, or direct experience as care home residents. The interviews addressed current patterns of commissioning and provision, exampes of success and faiure, how continuity of care was achieved, processes that supported integrated working and the anticipated impact of poicy change in rapidy changing heath and socia care economies. Recruitment To capture regiona, historica and organisationa differences, we identified a purposive sampe of NHS and oca authority commissioners, senior managers from care home organisations and the Care Quaity Commission (CQC). Reevant organisations were approached and invited to nominate peope we coud approach to interview. We aso interviewed a sma sampe of care home managers and residents who were invited to take part through My Home Life, an organisation that works with care homes to promote best practice. The extended time required to secure resident stakehoder interviews imited the number who participated and, foowing discussion with the SSC, we suppemented these interviews with a secondary data anaysis on 34 resident interviews from an earier study, ooking specificay at how they described what constituted good heath care. 2 Interviews Interviews were conducted face to face uness a participant requested a teephone interview. Participants were asked to provide a stakehoder view, in other words to use their experience and expertise, for exampe, as a care home manager, to inform what a good service shoud ook ike, rather than to provide a soey persona account. To faciitate this, the interview prompts addressed current patterns of commissioning and provision. Prompts for residents focused on what they beieved good heath care to care homes shoud comprise to inform and test our understanding of the processes that characterise how heath care is provided to care homes and how these work. Interviews asked about exampes of success and faiure, how continuity of care was achieved, what good working between NHS services and care homes ooked ike, and the mechanisms of particuar service modes necessary to achieve the desired outcomes. A interviews were recorded and fuy transcribed. To organise and structure the anaysis, data were entered into NVivo version 11 (QSR Internationa, Warrington, UK). The secondary anaysis of the resident interviews enabed us to consider what their descriptions reveaed about being heath-care recipients and what they identified as important. The interview eement of the study was reviewed and supported by the University of Hertfordshire Ethics Committee (reference number NMSCC/12/12/2/A). Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 9

46 RESEARCH APPROACH AND METHODS Anaysis Data were initiay mapped against the interview prompts. There were three stages to the anaysis. First, there was a process of famiiarisation, decontextuaisation and segmenting of data into separate and defined categories that were cose to how participants had described the issues. Second, there was a comparison within and between categories and identification of preoccupations, differences and themes. Third, there was identification of reationships and emergent hypotheses about how the favoured approaches worked, and what was necessary to support their impementation. Scoping of the pubished evidence To provide an overview of current provision of heath care to care homes we reviewed pubished sources using a review of surveys of the range and type of services provided to care homes, a review of existing systematic reviews and a suppementary scoping review to ensure that the iterature had been adequatey summarised. We had initiay proposed to survey the care home fied in the UK to understand the current range of provision to the sector. The review of surveys was suggested as an aternative approach by the SSC based on the assertion that the care home fied had been subject to a number of arge nationa surveys in the period immediatey prior to our period of research but that there had been a systematic faiure to coate these surveys or to consider how they informed each other to estabish an overview of care provision in care homes. To be eigibe for incusion in the review of surveys, pubications had to focus on heath-care deivery to care homes in the UK and had to have been competed since The review of existing systematic reviews was added to further enabe us to capture a range of approaches to service provision for care homes that may not have been identified in the review of surveys. A review of reviews was chosen as a way of approaching the pubished iterature based on the assumption that detaied summaries of the incuded studies woud be an efficient way of identifying the buk of potentiay reevant studies. It aowed us to examine reevant studies in a consistent way. Literature pubished since 2006 was incuded. This was a pragmatic decision to capture iterature that was ikey to be reevant to current systems of heath-care provision to care homes. As a fina step, a further scoping review was undertaken to ensure that no key iterature considering modes of heath-care deivery to care homes had been overooked in the review of surveys or review of reviews. Box 2 summarises the search terms, databases and e-networks used in phase 1 (i.e. the review of surveys, review of reviews and scoping of the iterature). Reaist review approaches are iterative, so these searches were refined, expanded and repeated as we tested emergent ideas about what supported heath-care services to achieve the outcomes of interest. Database searches were suppemented by onine searches conducted on the websites of prominent care home research groups, vountary sector providers of care homes, other care home organisations and their representative and professiona organisations. The websites of NHS strategic heath authorities were searched to identify care home initiatives referred to in their annua reports (up to March 2013). Citations yieded from the above searches were downoaded into and organised using EndNote [Carivate Anaytics (formery Thomson Reuters), Phiadephia, PA, USA] bibiographic software. A papers were independenty screened by two members of the research team. For the review of surveys, data extraction was structured to capture forms of NHS service provision for care homes in Engand in terms of frequency, ocation, focus and purpose and, where possibe, funding. The review of reviews and scoping review aso extracted data about the structure and function of the different types of service provision to care homes as we as considering in greater detai how services were 10 NIHR Journas Library

47 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 BOX 2 Search terms and incusion/excusion criteria for scoping of the iterature Search terms Care homes heath care survey, residentia care heath care survey, nursing homes heath care survey, oder peope heath care homes survey, oder peope heath residentia care survey, oder peope heath nursing homes survey, heath service provision care homes survey, heath service provision nursing homes survey, heath service provision residentia care homes survey, ong term care heath care survey and ong term care heath service provision survey. Eectronic databases MEDLINE (PubMed), CINAHL, BNI, EMBASE, PsycINFO, DH Data and The King s Fund were searched. In addition, we contacted care home-reated interest groups and used atera search techniques, such as checking reference ists of reevant papers and using the cited by option on WoS, Googe Schoar (Googe Inc., Mountain View, CA, USA) and Scopus, and the reated artices option on PubMed and WoS. E-networks E-networks requests for information were made to ECCA (now known as Care Engand), Care Home Providers Aiance, My Home Life Network, Nationa Care Home Research and Deveopment Forum, the PCRN, cinica study groups of the NIHR DeNDRoN and NIHR Age and Ageing network. Incusion criteria Pubications post 2006 of any research design, unpubished and grey iterature, poicy documents and information reported in speciaist conferences. Studies reevant to UK systems of heath care that addressed one or more of the outcomes of interest. Studies that were not UK-based but where there was transferabe earning reevant to the UK modes of service provision were incuded. Excusion criteria Studies where the heath-care provision to care homes was very different from UK modes of care were treated with caution or excuded, for exampe where medica support is in house (as in the Netherands) or the eve of care woud be coser to hospita-eve provision (as can be the case in the USA). Studies were excuded if the focus of the intervention or project ony invoved care home staff and/or a research team, that is, there was no input from visiting HCPs. BNI, British Nursing Index; CINAHL, Cumuative Index to Nursing and Aied Heath Literature; DeNDRoN, Dementias and Neurodegenerative Diseases Research Network; DH, Department of Heath; ECCA, Engand Community Care Association; HCP, heath-care professiona; NIHR, Nationa Institute for Heath Research; PCRN, Primary Care Research Network; WoS, Web of Science. deveoped, who was invoved and how they had affected, or considered, the outcomes of interest for our study (admissions to hospita, ength of stay, out-of-hours service use, medication use and user satisfaction). Because of substantia heterogeneity in the studies reviewed we did not poo studies in a meta-anaysis. Instead a narrative summary of findings was competed. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 11

48 RESEARCH APPROACH AND METHODS The combined findings of the above were used to deveop propositions about possibe CMO configurations and inked if then statements that were debated and refined within the team about what might support heath-care provision to care home residents. Theory refinement and testing This step invoved taking the theoretica propositions and possibe CMO configurations derived from the interviews, review of surveys and review of reviews that captured the emergent programme theories of how heath-care services worked with care homes. Detaied reading of the iterature from the earier stages of the review was accompanied by atera searches of references retrieved from artice bibiographies, driven by emerging theoretica constructs where it was cear that additiona data were required from underpinning research studies. This more in-depth consutation of the iterature was used to ook for data that supported, refuted or augmented the possibe CMO configurations identified in the earier reviews. Anaysis focused on interventions that drew on theories about the assessment of frai oder peope in the ast years of ife, system-driven quaity improvement schemes in primary care and theories of integrated working that emphasised reationa, participatory and context-sensitive approaches in care home settings (see Appendices 2 and 3). In keeping with reaist enquiry methods, equa consideration was given to negative and positive outcomes and inconsistencies in accounts of what works, when and with what outcomes. We retained the incusion and excusion criteria used for the initia scoping. Quaity assessment was based on the opportunities for earning and testing emergent theory. Thus practitioner accounts of innovation were considered as evidence aongside empirica research. Four reviewers (CG, SLD, MZ and MH) independenty screened tites and abstracts to identify reevant documents, which were retrieved and assessed according to the incusion criteria. A incuded papers were read by Caire Goodman and one of the three reviewers. Data extraction focused on how heath care was organised, funded, provided and deivered, how the underying assumptions and theoretica framework (if identified within a particuar study or group of studies) were articuated, and whether or not this fitted with the focus of our review in terms of the underying theory and the impact of the intervention on the outcomes of interest. Our approach drew on Rycroft-Maone et a. s 42 approach to data extraction in reaist synthesis that questions the integrity of each theory, considers the competing theories as expanations to why certain outcomes are achieved in simiar and different settings and compares the stated theory with observed practice. Anaysis and synthesis A reaist anaysis of data adheres to a generative expanation for causation and ooks for recurrent patterns of outcomes and their associated mechanisms and contexts (CMO configurations). 31 As the review progressed, the discussion focused on particuar papers and sources that offered competing accounts of why or how an intervention was chosen and why it had, or had not, worked. We concentrated on what appeared to be recurrent patterns of contexts and outcomes in the data (demi-reguarities) and then sought to expain these through the means (mechanisms) by which they occurred. The review s preiminary findings were presented to the study advisory group for further discussion and chaenge. This iterative discussion process compared the stated theory with the evidence reviewed. We discussed how and why different mechanisms were triggered by the different approaches to providing heath care to care homes. The findings were then used to structure the recruitment and samping approach for testing in phase NIHR Journas Library

49 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Phase 2 The case study phase of the project addressed research questions 3 to 5: How are these features/mechanisms associated with key outcomes, incuding medication use; use of out-of-hours services; resident, carer and staff satisfaction; unpanned hospita admissions (incuding to an A&E department); and ength of hospita stay? How are these features/mechanisms associated with costs to the NHS and from a societa perspective? What configuration of these features/mechanisms woud be recommended to promote continuity of care at reasonabe cost for oder peope resident in care homes? Phase 1 findings indicated that we shoud target service deivery modes that acknowedge and support the interactiona nature of decision-making between care home staff and visiting heath-care professionas (HCPs), for exampe by supporting increased contact from NHS practitioners, structured meetings and joint review of residents needs. The expanatory theory of interest and supporting CMO configurations, with the potentia to expain why some or a of the outcomes were achieved (or not), was one that specified what needed to be in pace to trigger, support and sustain mechanisms that generated trust, mutua obigation, recognition of how care homes worked and a common purpose. The review suggested that particuar activities within different service modes were important contextua factors (or possiby mechanisms). These incuded education, training and ongoing support of care home staff; empoyment of HCPs to work with care homes; opportunities for reguar review and discussions between care home staff and professionas; and the aocation of resources to increase the frequency of visits by and invovement of primary care service staff. A case study approach was chosen to faciitate a detaied description of processes of care and a comparison of the deivery of heath care over a sustained period of time to care homes and their residents, across three geographicay discrete sites. Specificay, we aimed to identify three sites where heath-care provision had been designed to refect some or a of the contexts identified in the review and particuary those that might support reationa working. Ethics approva The phase 2 case study was reviewed and given a favourabe opinion by the Socia Care Research Ethics Committee on 29 January 2014 (Ethics Committee reference number 13/IEC08/0048). Samping and recruitment Initiay we proposed to seect one Cinica Commissioning Group (CCG) area as an exampe of usua care, that is, one with an approach to commissioning and care deivery for care homes in its area where there was itte or no differentiation between commissioning of services provided to peope iving at home and those in care homes. It became cear, however, that nationa preoccupations with unpanned hospita admissions meant that it was unikey that a site woud not have any intervention or initiative operating that invoved care homes. We therefore approached and recruited a site where the main route for access to medica and speciaist care was through the genera practitioner (GP) and the Genera Medica Services (GMS) contract, but the county had aso invested in care home manager eadership training (site E beow). Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 13

50 RESEARCH APPROACH AND METHODS We identified six CCGs/geographica areas within Engand that were each operating a distinctive approach to deivering heath care in care homes and within 2 hours traveing distance of our two research centres. 1. Site A: CCG investment in care home speciaist provision, which incuded care home speciaist teams, inked speciaist dementia care, fas prevention teams and invovement of community geriatricians. 2. Site B: CCG provision of financia payments to specific GP practices to work with care homes and deiver on specific areas of care compemented by the commissioning of training and education in care homes for residents with compex care needs. The initiative required care homes to register with one practice, CCG investment in a mutidiscipinary out-of-hospita team, a 24/7 resource covering heath and socia care to avoid admission to hospita, supported eary discharge and home-based rehabiitation. The team incuded care homes in its remit and had access to beds with care within a care home environment where a patient needed added intensity of care. 3. Site C: CCG investment in community matron support for care homes and a series of topic-specific initiatives to improve medication management and access to end-of-ife care, and to prevent and reduce pressure ucers in care homes. 4. Site D: CCG investment in support to care homes that focused on the creation of a singe team for care homes ed by a community matron, set up to work cosey with GP practices. 5. Site E: geographica area that had a ong history of innovation in care home working, for exampe pioneered intermediate care beds in care homes and where mutipe care homes were activey invoved in the Enabing Research in Care Homes (ENRICH 45 ) network. 6. Site F: geographica area where there had been training and investment across the county in care home manager eadership training and deveopment. CCG investment in care homes was based on GMS contracts with inked services designed to reduce unpanned admissions from across the community. In four sites (A, B, C and D) the service proposition to care homes was deimited by the geographica boundaries of the oca CCG, as the modes of care had been specificay commissioned by CCGs in response to the chaenge of providing heath care to care homes. To fuy understand the service proposition in these areas, it was necessary to ensure both the permission and the engagement of the oca CCG to ensure that the necessary access to sites and staff stakehoders woud be supported. These sites were therefore approached for recruitment by the research team contacting the CCG and the commissioners with responsibiity for care homes and one via the organisation that had organised the training and deveopment programme for care home staff in the county. Two CCGs (C and D) expressed an initia interest in participating and received information about the study, and the researchers had preiminary teephone conversations with commissioner/site representatives. In one area the chairperson of the CCG decided against participation and in the other site the research team decided not to pursue the coaboration, as the CCG was sti in the eary stages of introducing changes as to how it worked with care homes. In site E interest was expressed from specific care home managers; despite this interest, geographica proximity and participation in the ENRICH 45 network, the team decided that site F offered more opportunities for earning. At the ast site (F), the reevant contextua factor investment in a eadership and management framework was not geographicay bounded within a singe CCG s footprint because the service was deivered as a county-wide initiative by a nationa charity. The approach to the oca care home eadership and management network was therefore made through the nationa charity, rather than the mutipe CCGs that commissioned services with which the care homes might be required to interface. The managers comprising the network agreed, unanimousy, to participate in the study. When the sites were confirmed, 72 care homes that met our incusion criteria were invited to participate in the study. 14 NIHR Journas Library

51 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Seecting and recruiting homes for invovement in the study Area A (investment in speciaist care home provision) was recruited to the study and identified as study site 1, area B (provision of financia incentives to GPs) was identified as study site 2 and area F (care home eadership and management framework) as study site 3; four care homes were recruited in each site. A sites were in Engand and were ocated in the Midands and the east of Engand (incuding the east coast). Athough the identification of service deivery modes was theory driven, based on the findings of phase 1 we focused, as far as possibe, on typica homes, that is, those with 25 beds or more (the median size of care homes with and without nursing provision is 25 beds and 48 beds, respectivey) and those identified as having contact with a range of NHS services comparabe to the common patterns of service deivery identified in phase 1. We aimed to recruit care homes from a range of ownership categories, incuding arge corporate providers, more ocaised singe-home providers or sma chains and third-sector charitaby funded homes. We did not specificay seek to recruit either NHS- or oca authority-funded homes as these now represent exceptiona modes of funding care. 12 Our excusion criteria were care homes with speciaist registration for acoho and drug abuse or earning difficuties; those with bed numbers outside the interquartie range; those whose manager had been in post for < 6 months; and those providing speciaist care services commissioned by the NHS. From the remaining homes, in sites 1 and 2, a homes that had contact with the services of interest were sent a etter inviting them to participate in the study. This was foowed up by a teephone ca from the study researcher to give them further information and to set up a meeting with managers who were interested in participating. In site 3, initia contact with care homes was made by the charitabe organisation that provided eadership training. Here the detais of interested managers, foowing their consent to be contacted, were passed on to the researcher to communicate directy. In a sites, meetings were arranged with managers to give them further information and to answer any queries about the study and what it invoved. From those wiing to participate in the study in principe, care homes were seected to incude those with and without on-site nursing and registration for dementia care. To enhance opportunities for comparison we aimed, as far as possibe, to match the first four care homes recruited in site 1 with the remaining eight care homes in the other two sites, based on resident popuation, staffing ratios and geographica proximity to a NHS acute hospita providing secondary care. Support for recruitment and participation was achieved in one of the sites through coaboration with the oca Cinica Research Network, which recruited participating homes to be part of the ENRICH 45 network aongside the research undertaken as part of this study. Recruiting residents from participant homes The chaenges of recruiting oder peope to research in care homes are we documented. 46,47 Based on previous studies, and with the support of the ENRICH 45 network and experienced PIR members, we aimed to achieve the maximum possibe recruitment of residents. Where residents acked menta capacity to give consent and had no contactabe persona consutee, we empoyed a robust protoco using nominated consutees to boost recruitment. Those residents attending for respite care ony, or those who were identified by care home staff as terminay i (i.e. in the ast weeks of ife) or too i to participate were excuded. Sampe size Based on three areas and the purposive samping framework outined, we expected to recruit a resident sampe of based on % recruitment. Our target number of care home staff was 60 (five per home) to Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 15

52 RESEARCH APPROACH AND METHODS refect a range of seniority and ski and depending on GP attachment and modes of service deivery between two and three GPs, two and three NHS nurses (district nurses/speciaist nurses) and two and three therapists per home, representing a maximum of 168 participants. Where possibe we aimed to interview the chairperson or members of the participating CCG (between three and five) and the Heath and Webeing Boards (n = 3) about current and projected patterns of service deivery to care homes. Conducting the case studies The ongitudina mixed-methods design enabed us to do two things: (1) to track the resource use of residents, particuary their use of emergency and out-of-hours services and (2) to understand how over time the different expressions of reationa working between NHS and care staff were achieved and with what outcomes. It aso enabed us to study if particuar residents (e.g. those who frequenty used resources) benefited more or ess from the different mechanisms of care provision and whether or not there were differences in responsiveness and fexibiity as residents needs changed over time. Data coection with each of the study care homes was conducted over a period of 12 months foowing the baseine data coection. Baseine data coection Baseine descriptors for residents were coected from their care home records to provide the basis for a comparison of the popuation studied within the four care homes across the three study sites. Foowing pioting with care home staff, we used a modified version of the internationa Resident Assessment Instrument (interrai) items for use in assisted iving faciities, to record information on residents cinica and functiona status 48 for modified interrai-assisted iving. 49 This too was amended to aow for differences in terminoogy between Austraia where the too was deveoped and the UK and to remove sections that were not reevant to care home residents in UK care homes. It combined assessment items reating to cinica characteristics with activities of daiy iving (ADL) and cognitive function that reate to staff care time. 50 interrai is a standardised assessment instrument for oder aduts with fraity, 49 widey used outside the UK and internationay vaidated, that provided the study with data of cross-nationa comparabiity. The too comes with a number of vaidated protocos that automaticay generate subscores for a number of cinica syndromes, common diagnoses and patterns of dependency. Based on the experience of the SHELTER study, 51 which coected data from 4156 care home residents across eight countries using a version of the interrai for ong-term care faciities (LTCFs) and incuded 507 residents across nine faciities in the UK, it was beieved to be feasibe for the interrai to be competed by care home staff (incuding care assistants) foowing training from the research team. Data on medication use at baseine were coected from Medication Administration Record (MAR) sheets. Thereafter, monthy changes to medications (additions, subtractions, substitutions) were coected from the MAR sheets and annotated onghand into the study database. Descriptive case studies of continuing care as deivered to care home residents The case studies provided descriptive data on resident characteristics and resource use, and quaitative data about how heath-care provision to care homes was seen as supporting (or not supporting) reationa working between the NHS and care homes. Case descriptions were buit iterativey, taking account of audio-recorded interviews with residents and famiy members, care home staff, heath and socia care commissioners, GPs, NHS nurses and aied heath professionas. Observation of NHS care deivery occurred 16 NIHR Journas Library

53 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 when researchers were in the care homes; this incuded, for exampe, observation of meetings between NHS and care home staff and their contact with residents. These were documented as fiedwork notes and incuded in the quaitative anaysis as memos. Care home poicies and procedures that focused on residents heath care were aso reviewed. Participants were interviewed at east once over the 12 months. Interviews were semistructured. Interview schedues were initiay focused to further inform and enabe iteration of the CMO reationships and mid-range theory emerging from phase 1 of the study and were modified over time as these conceptua frameworks evoved. Schedues drew on work on continuity of care and work on integrative processes, which previousy highighted that service provision can ony be meaningfuy understood from the eve of the patient or, in this case, the resident. A interviews were recorded and fuy transcribed. Care home staff were recruited as vounteers from among the broader workforce of the care homes. Posters were paced in staff areas of the care homes. In addition, staff were made aware of the study and the opportunity to participate in focus groups/interviews when researchers were visiting homes to support routine coection of resident-eve outcome measures. Managers payed a supporting roe in heping to recruit staff but were advised not to coerce them to participate. Heath-care professionas were identified for interview on the basis that they were frequent visitors to the care homes, as estabished through the monthy service use reviews of participating residents and iaison with the care home ink staff. Where required, permission was sought from service managers for the researchers to contact individua HCPs inviting them to participate in the study. To maximise recruitment from this group, teephone interviews were conducted with those staff unabe to meet face to face, rather than risk osing their reevant perspectives from the study atogether. Genera practitioners who had a roe in providing residents with the care homes were identified through discussions with care home staff and were then contacted individuay to request participation in a one-to-one interview. Where this faied to recruit participants, GPs were approached coectivey via the CCG to take part in a focus group discussion on working with care homes. Informed by the findings of the reaist synthesis, interviews focused on the experience of providing and receiving heath care in care homes. For residents and carers, we focused on what was important in reation to satisfaction with the services and how they saw reationships with care home staff and heath-care practitioners contributing to this. For care home staff we focused on their satisfaction with the heath-care services provided to the care homes and what they viewed as the priorities for NHS services when supporting residents heath care. Researchers provided feedback about how residents had been found to use services and used these to eicit care home staff views and experiences of working together with heath-care practitioners. Face-to-face interviews with care home managers at this stage aimed to capture any changes in the way that heath services were provided over the data coection period and to compare and contrast their satisfaction with, and perceptions of, heath service provision with those of their staff. Heath-care staff were asked to consider the research team s understanding of how they worked with care home staff, and other HCPs, NHS priorities for residents care and how satisfied they were in working with the care home. For GPs, we focused on how they worked together with care homes to provide care for residents, their eve and type of contact with other NHS HCPs and their priorities for care, with a greater focus on their roe in medication management. A fina set of interviews and focus groups shared detais from the process anaysis and the emergent conceptua frameworks regarding the outcomes of interest and asked care home managers to consider the extent to which these resonated with their experiences. Care home staff satisfaction surveys (staff outcomes) To suppement quaitative data on staff satisfaction, we conducted a survey to take account of staff members overa satisfaction with continuing heath-care services. We used the Quaity-Work-Competence (QWC) questionnaire as the basis of this, deveoped and vaidated by Hasson and Arnetz 55 as a mechanism Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 17

54 RESEARCH APPROACH AND METHODS for coecting data on care home staff competence, work stress, strain and satisfaction. The key area of interest, in keeping with the outcomes of interests stated at the start of the study and refecting the focus of our programme of work specificay around heath care and the contextua and mechanistic factors required to support good heath outcomes, was the extent to which staff were satisfied with the heath services provided within and between regions. We therefore ony used the subset of QWC questions focused around quaity of care. Resource use outcome measures A bespoke pro forma for coecting service use data was deveoped with the participating care homes to coect data on the community services that visited participating residents. These discussions were informed by the review of surveys undertaken in phase 1 that reported the range and variabiity in services accessed by care home residents. Foowing initia discussions, the specific services to be recorded for residents on a monthy basis were coated in a pro forma, which incuded the main service use outcomes of interest namey out-of-hours services, unpanned hospita admissions (incuding A&E) and ength of hospita stay. Each care home was visited by a researcher on a monthy basis at east, but usuay more frequenty. This heped to maintain working reationships with the care homes to verify and support up-to-date contemporaneous competion of forms. Two designated members of the care home staff (study ink staff) were identified who had responsibiity for supporting resident and reative recruitment, day-to-day data coection on resource use, informing us of key events in the care home (e.g. CQC inspections, staff changes, etc.) and iaising with NHS services (see Appendix 5). These data were checked and the detais were carified by researchers from care home records and in discussion with the care home staff. We had panned to cross-check the service use data obtained from the reviews of residents care home notes with data extracted from their medica notes, incuding information on hospita admission and ength of stay, out-of-hours and emergency ambuance service use and referras to other heath-care services in the preceding 12 months. If this was not acceptabe we aimed to do a 10% reiabiity check with residents GP records. At the end of data coection and despite mutipe attempts, incuding support from the Cinica Research Network, we were unabe to access resident data from GP notes. This refected rea difficuties in recruiting GP coeagues to a parts of the research study. Anaysis and synthesis Quaitative data were anaysed in stages (see aso Figure 1). Stage 1 The initia data anaysis commenced with the participant interviews, on the basis that an understanding of the provision and structure of heath-care services to care homes woud faciitate interpretation of the data from the other participant groups. The first eve of anaysis focused on the participants transcripts. The data were anaysed both inductivey and deductivey, initiay at the care home eve, foowed by a within-case anaysis of the three study sites. A interview and focus group transcripts were entered into NVivo. Each transcript was coded thematicay according to the responses given to the interview questions together with other variabes incuding roe, remit and the use of shared documentation. At the second eve of anaysis, each theme or topic was interrogated to identify the features and structure of heath-care deivery to the care homes and the way that the services were organised. This incuded fied notes and memos that were compied during the data coection process. Anaysis at the care home eve was foowed by a within-case anaysis in which each study site, and the four care homes within it, constituted a case. 18 NIHR Journas Library

55 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Stage 1: inductive Interview data CH eve Cross-sectiona anaysis First eve: thematic Second eve: interrogate theme/topic in view of the specific CH context and variabes; incude fied-notes and memos Deveop anaytic memos Begin to popuate the fieds beow (as per Gordon et a ) CH Topic Topic Topic... 1 Anaytic memos (+quotes and refs) FIGURE 1 Data synthesis. CH, care home; MHL, My Home Life. Stage 2: deductive Takes the underying assumptions made (hypotheses) for each study site as its starting point, for exampe: S1: age appropriate S2: governance/incentive mode S3: reationa (MHL) Site eve Within-case anaysis Map codes (and anaytic memos deveoped during the inductive stage) on a matrix to identify common features, shared features and codes/findings that deviate from the theoretica assumptions that undergird each study site (or case) This matrix constitutes the basis of the framework anaysis depicting CMO reationships Stage 3 Discuss, cross-check and fine tune the anaysis. Case/Site C M O Do findings ine up with quantitative data? Did assumptions hod? Create new hypotheses to expain deviance? Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 19

56 RESEARCH APPROACH AND METHODS Stage 2 In stage 2, deductive data anaysis was driven by the theoretica assumptions on which each study site had been seected for the study from the phase 1 findings and possibe within- and cross-case CMO configurations, that is, sites were oosey characterised as study site 1, an age-appropriate mode of care, study site 2, a governance mode and study site 3, a reationa mode of working. 57 Drawing on the anaysis from stage 1, the characteristics of each mode were identified as we as the common features and differences across the sites, together with any data that did not rest on the assumptions underpinning the modes. Coding of the transcripts was aso conducted for the study outcomes, medication use and management, satisfaction, unpanned hospitaisation incuding ength of stay, out-of-hours service use and A&E use. In ine with the findings from the reaist review, exampes of reationa working between HCPs and care home staff were identified and anaysed to highight their associated features. The resuting data were used to provide a detaied description of the intervention, the outcomes at different eves, context conditions and mechanisms in order to faciitate the identification of the emerging CMOs. 58 Quantitative data anaysis In order to assess the economic outcomes for each of the sites, it was necessary to appy costs to the service use frequency data, coected by month from the patient records. Unit costs were identified from nationa and pubished sources. These unit costs were mutipied by the frequency of events for each of the resource use items and a new variabe was generated containing the costs, at an individua resident eve for that resource item. The frequency variabes and costs were then aggregated over time and resource type. The resuting data set contained a set of eight frequency variabes of resource use and the corresponding eight variabes containing the summative tota costs for that resource type, at an individua resident eve. Two tota cost variabes were aso generated: the summative tota costs at an individua resident eve, incuding a eight heath resource types, and a costs, except hospita admissions, which were separated out because inpatient episodes were infrequent events, but costy when they occurred. Missing data were assessed across a resource use items by way of frequency tabes. When a positive response had been recorded for one or more heath resource variabes for a particuar time point, nested by individua resident, any missing vaues for other resource items were assumed to be zero vaues. When this rue ed to missing data for one or more time points for an individua, the observations for the resident were counted as missing and the resident was dropped from the compete-case anaysis. Using this approach, the percentage of compete cases at each month was cacuated. Beyond 6 months, when 85% of resident records were compete, the frequency of missing data increased to > 20% (and was 45% at 12 months). Hence 6 months was seected as the primary end point for the anaysis. To verify that inadvertent bias was not introduced by conducting the economic anayses based on those residents with compete data at 6 months, we compared the distribution of baseine variabes for a residents recruited to the cohort as a whoe and those incuded in the economic anaysis, and no statisticay significant difference in baseine variabes was detected. Considering associations between baseine variabes, costs and outcome variabes The mean and standard deviation (SD) for each of the summary categories of service use and costs and tota costs, broken down by site, were computed. Pairwise comparison between sites for heath resource use was conducted using Pearson chi-squared tests; sampe t-tests were used for site comparisons of costs. 20 NIHR Journas Library

57 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 We used Poisson regression to expore whether or not the site was a significant predictor of service use and tota costs after the interrai scores and the derived variabes that we had deveoped for this study (cognitive impairment, number of comorbidities and medication count), as we as the interaction between the site and these variabes, were entered in the regression equation. Poisson regression is a technique that can be used with infrequent count data, such as the service use data. We first entered each of the interrai cinica syndrome variabes [ADL hierarchy scae (sadlh), short ADL hierarchy scae (sadlsf), cognitive performance scae (scps), communication scae (scomm), cinica syndrome for pain (spain_1) and pressure ucer risk scae (spurs)], as we as each of the derived variabes (cognitive impairment, number of comorbidities and medication count) in a Poisson regression equation with each of the service use variabes separatey. As an exampe, a regression equation woud use sadlh to predict primary care contacts, GP contacts, out-of-hours contacts, community contacts, A&E visits, ambuance use, secondary care number of admissions, secondary care duration and secondary care non-admissions separatey. Then, another regression equation woud use sadlsf to predict each of the service use variabes separatey. As this was an exporatory anaysis, we seected predictors based on whether or not they were significant in the univariate anaysis at a p-vaue of < Medication anaysis Medication anaysis focused on tota medication, antibiotic and opioid counts and antichoinergic burden (ACB) scoring using the ACB scae produced by Aging Brain Care [ products/acb_scae_-_ega_size.pdf (accessed October 2017)]. This was based on the guidance of the study steering group that identified an expansive ist of prescribing decisions that coud indicate optima or suboptima prescribing in the care home setting ranging from number of antibiotics, painkiers and antihypertensives to more comprehensive indices incuding the STOPP/START criteria 59 and Medication Appropriateness Indices. 60 The eventua recommendation of this group was to use the ACB scae as a proxy measure that incorporated antipsychotic and antihypertensive prescribing rates, and to count opioids and antibiotics separatey. ACB scoring and the aocation of medications to antibiotic and opioid categories were conducted by a consutant geriatrician (ALG). For baseine medication data, counts and distributions were summarised using means (SDs) and medians (range) for parametric and non-parametric data, respectivey. Differences between sites were considered using the anaysis of variance (ANOVA) statistica method and Kruska Wais one-way ANOVA for parametric and non-parametric data, respectivey. Opioid and antibiotic prescription were treated as dichotomous variabes (present/absent) because the majority of residents were taking ony one of each of these medications and were thus compared between sites using Pearson s chi-squared test. Foow-up data comprised tota drug, antibiotic and opioid counts, and ACB scores for each month of foow-up. These were used to cacuate tota drugs/resident, antibiotics/resident, opioid/resident and ACB/resident, which were potted as ine graphs with 95% confidence intervas (CIs) to enabe comparison in prescribing trends over time. Staff satisfaction Staff satisfaction questionnaires were anaysed simiary to medication data, with counts and distributions summarised using means (SDs) and medians (range) for parametric and non-parametric data, respectivey. Differences between sites were considered using ANOVA and Kruska Wais one-way ANOVA for parametric and non-parametric data, respectivey. Categorica variabes were compared between sites using Pearson s chi-squared test. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 21

58 RESEARCH APPROACH AND METHODS Anaysis workshop A 2-day anaysis workshop was used to feed back emergent findings from the three sites to the research management team together with a member of the PIR group (Kate Sartain), in order to discuss the resuts from the quantitative data and the quaitative data and evidence supporting different CMO configurations both within and across the case study sites. To enabe comparison, home-, resident- and staff-eve data were anaysed and reported on a site-by-site basis. A matrix 61 was generated, with the rows representing sites and coumns organised to refect both the key propositions deveoped from phase 1 and the data generated from resource use. This was used to faciitate quaitative cross-case anaysis, taking account of simiarities and differences between and across the three sites. Attention was paid to what the data reveaed about the inter-reationships between the mechanism and context of care and how these inked to the outcomes of interest. 52,54 Anaysis was iterative and refected the anaytic stages foowed in phase 1; it focused on what was reveaed about the actua intervention or mechanism, the observed outcomes, the context conditions and underying mechanisms. This was compared with the theoretica propositions from phase 1 to estabish the conditions under which the mechanisms work (or not) and their transferabiity across different settings. 22 NIHR Journas Library

59 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Chapter 3 Resuts from the review of surveys and the review of reviews Introduction This chapter presents the findings that address the first question of the OPTIMAL study: What is the range of heath service deivery modes designed to maintain care home residents outside hospita? The first haf of the chapter concerns the anaysis from the review of pubished surveys of heath-care provision to care homes. The second haf provides a summary of the review of pubished reviews findings about the focus and priorities of heath-care research in and with care homes. The chapter provides an overview of the organisation of heath care for care homes with a particuar, but not excusive, focus on UK services. A paper on the review of surveys has been pubished esewhere. 4 Review of surveys To compement the searches of the databases (summarised in Chapter 2), forward citations and onine searches of NHS websites and e-mai requests, we reviewed the websites of eight academic centres known for their work in care home research, four charities, seven care home provider representative bodies and 12 NHS, socia care and professiona organisation sites with responsibiities for care homes. Data extraction and anaysis were competed in Sixteen surveys competed since 2008 were identified and fifteen were incuded. Five focused on GP service provision to care homes, whie aso coecting data on speciaist services. Ten focused on speciaist services to care homes or were topic specific, for exampe, focusing on dementia services or end-of-ife care. One survey 62 considered the care home nurses work environment, staffing eves, quaity of care, meeting residents needs and financia pressures on the home, but this was not incuded as there were no data on externay provided heath-care provision/nursing support to care homes. Ony two surveys incuded residents, one of which aso incuded reatives of residents who were unabe to participate because of cognitive impairment. The main methods of data coection were posta or onine surveys, athough some used face-to-face interviews with care home residents and teephone interviews with GPs. The surveys are summarised in Tabes 1 and 2. Tabe 1 shows, in order of pubication date, those surveys that focused on GP services, and Tabe 2 ists those focusing on speciaist services or topics. Genera practitioners were seen in most studies as key to the provision of good-quaity heath care for care home residents, incuding end-of-ife care. There was no consensus, however, about how GP and other primary care services shoud be organised in reation to the care homes or what they shoud do. For exampe, some GPs carried out reguar medication reviews (6-monthy or yeary), whie some did post-admission assessments. Invariaby, the care homes surveyed worked with mutipe practices and mutipe GPs the argest number of practices visiting one care home was 30 athough some had a singe designated GP. Consutation arrangements were aso variabe, with weeky cinics or visits being made ony on request. This unevenness of provision was mirrored in famiy and residents views: one survey found that ony 56% reported good access to, and support from, GPs, with 55% of staff aso reporting that residents got enough support from GPs. 72 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 23

60 24 NIHR Journas Library TABLE 1 Summary of generaist provision to care homes from five surveys Author, tite; year (home type) Aims Survey detais Sampe size/response GP services Other services 1. Morris J, Patients Association/University Coege London/BGS Cinica Quaity Steering Group, 2008, persona communication, interna report (nursing homes in Engand) Do care home staff and GPs get enough information about new residents? Do GPs and care home staff fee supported by primary and secondary care? Face-to-face or teephone interviews on one occasion with 11 care home managers and six GPs who worked with them using standardised questionnaires for care home managers and GPs, incuding some overapping questions. Service-reated questions incuded the foowing services: GPs, tissue viabiity, menta heath, end-of-ife and paiative care, geriatrician, od age psychiatry, audioogy, ophthamoogy, podiatry, physiotherapy, occupationa therapy and community pharmacist Eeven care homes seected to refect a range of care homes in terms of size, ocation and residents The care homes ranged in size from beds From nursing home interviews (n = 9), seven out of nine care homes had a singe designated GP five did weeky cinics, one visited daiy and the other 2-weeky From GP interviews (n = 6), three out of six did either yeary or 6-monthy medication reviews A care homes had access to an ophthamoogist or optician, tissue viabiity support and support with menta heath/behavioura probems According to GPs, five care homes had access to paiative care support Four homes had access to audioogy and podiatry, two of which were provided by the care home organisation Three care homes had access to the geriatrician and od age psychiatrist; the others had no or ad hoc access Three had access to a physiotherapist whom they empoyed directy No care homes had access to occupationa therapy services or a community pharmacist RESULTS FROM THE REVIEW OF SURVEYS AND THE REVIEW OF REVIEWS

61 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. Author, tite; year (home type) Aims Survey detais Sampe size/response GP services Other services 2. Gadman and Chikura, Medica Crises in Oder Peope; (care homes with and without on-site nursing) To conduct a review of current service provision to edery residents in 252 care homes across the county Posta survey. Data were coected on 20 services, incuding fas, GP, pharmacist, physiotherapy, occupationa therapy, end of ife, menta heath, DN, podiatry, community geriatrician, nurse practitioner, dietitian, community matron, ong-term conditions, tissue viabiity, continence, dementia, optometrist, SALT, stroke rehabiitation One hundred and eighteen responses (47% response rate) A homes aowed their residents to register with the practice of their choice (one care home was served by up to 16 practices) Most visits were on request, GPs offered reguar surgeries, others found it hard to get visits Forty-two per cent of care homes did not have reguar GP visits Ninety-seven per cent of care homes had access to pharmacy, 92% to a DN and 89% to a dietitian Most services avaiabe on request rather than routiney with the exception of pharmacists The services avaiabe to the east number of care homes incuded nurse practitioner (34%), community geriatrician (42%; 9% of care homes had reguar visits from the community geriatrician) and ong-term conditions team (43%) Twenty-three per cent of care homes coud not access SALT, physiotherapy or occupationa therapy services An exampe of specific care home services: a nurse-ed team that worked cosey with care homes to iaise with NHS services and offer training and support to care homes, medication reviews continued DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO

62 26 NIHR Journas Library TABLE 1 Summary of generaist provision to care homes from five surveys (continued) Author, tite; year (home type) Aims Survey detais Sampe size/response GP services Other services 3. Gage et a., Integrated working between residentia care homes and primary care: a survey of care homes in Engand; (care homes without on-site nursing) 4. Quince, Low Expectations: Attitudes on Choice, Care and Community for Peope with Dementia in Care Homes; (care homes with and without on-site nursing in Engand, Waes and Northern Ireand) The APPROACH study survey 1: to estabish the extent of integrated working between care homes and primary and community heath and socia services To expore attitudes on choice, care and community for peope with dementia in care homes A sef-competion, onine questionnaire of open and cosed questions designed by the research team to estabish the primary heath-care service provision to care homes and their experience of integrated working with those services Interviewed staff and observed care provided to 386 residents a Sent to a random sampe of residentia care homes in Engand in 2009 (n = 621) with more than 25 beds Ninety-three out of 587 care homes responded (a 15.8% response rate) Twenty-seven care homes with nursing: 144 aduts aged > 65 years were observed or interviewed Twenty-seven care homes without nursing: 153 residents aged > 65 years were observed or interviewed and 90 staff were interviewed A care homes received GP services 81% worked with more than one practice Consutation arrangements varied from weeky GP cinics to as required. Seven (8%) paid a retainer to the GP but these were seen as unfair Variabiity in GP services received: 44% had schedued surgeries or visits; uncear who pays for GP services. Fiftythree per cent said the NHS paid for GP services Thirty-three per cent of GPs did not do the assessment post resident admission More than 90% of homes reported using DNs and opticians Other frequenty accessed services (> 80%) incuded CPNs, podiatrists Between haf and threequarters of homes reported visits from continence nurses, pharmacists, dentists, hearing services and od-age psychiatrists Difficuty accessing speciaist services was a consistent theme Ony addressed GP provision RESULTS FROM THE REVIEW OF SURVEYS AND THE REVIEW OF REVIEWS

63 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. Author, tite; year (home type) Aims Survey detais Sampe size/response GP services Other services 5. Carter, Faiing the Frai: A Chaotic Approach to Commissioning Heathcare Services for Care Homes; (report to the British Geriatrics Society b ) To estabish what heath-care services are commissioned by PCTs for oder peope iving in care homes Onine survey by CQC to seek information about heath services provided to care homes and oder peope iving in the community in a PCTs in Engand (152 PCTs in Engand) The survey focused on nine key services: geriatricians, psychiatry, dietetics, occupationa therapy, physiotherapy, podiatry, continence, fas prevention and tissue viabiity PCTs were aso asked about additiona services provided by GPs to care homes and payment for enhanced services to care homes. The CQC focused on seven activities that GPs coud perform in care homes: heath assessments on admission, speciaist assessments, reguar visits, support with end-of-ife care panning, genera support, iaison with other services and additiona medication reviews The survey noted that some responses provided minima information. c Do not know and missing answers were treated as negative answers It was not aways possibe in the anaysis to separate the findings for care homes from those that appied to a oder peope iving in the community. The focus of the anaysis was on commissioning intentions Fifty-one per cent (n = 77) of PCTs had enhanced service agreements with GPs for their work in care homes Sixty-seven per cent of PCTs did not think that care homes needed additiona medication reviews Seventy-seven per cent of PCTs provided at east one activity considered to be an enhanced activity for care homes There were significant variations in speciaist provision to oder peope with 52 different possibe combinations identified Forty-three per cent (n = 65) of PCTs provided a the services that the CQC considers appropriate for a oder peope Sixty per cent (n = 91) of PCTs provided a geriatrician service to a oder peope Most speciaist services made visits on request Schedued visits were most ikey to be offered by continence services, podiatry, dietetics and psychiatry APPROACH, Anaysis and Perspectives of integrated working in PRimary care Organisations And Care Homes; CPN, community psychiatric nurse; DN, district nurse; PCT, primary care trust; SALT, speech and anguage therapist. a Incuded care homes for peope with earning disabiities. b Data coected from Engish primary care trusts (i.e. primary care trusts, not care homes). c Primary care trusts had a mandatory obigation to respond. DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO

64 28 NIHR Journas Library TABLE 2 Summary of data extracted from topic-specific surveys Author, tite; year (home type) Aims Survey detais Sampe size/response rate GP services Other services 1. Steves et a., Geriatricians and care homes: perspectives from geriatric medicine departments and primary care trusts; (a types of care homes in Engand) To test concordance with the Roya Coege of Physicians, Roya Coege of Nursing and the British Geriatrics Society s 2000 guideines on cinica practice for care homes Compementary surveys for PCTs and GMDs 2006 to the ead cinician in each GMD in Engand and the ead nurse in each PCT in Engand Responses from 109 of the 167 (65%) GMDs in Engand and 141 of the 303 (47%) PCTs N/A Seventeen (15.7%) GMDs specificay aocated sessions to care home work, mosty with nursing homes Some PCTs funding geriatrician invovement in care homes (18.4%), but 52% of PCTs (n = 74) required either geriatrician s invovement in the admissions process (20%) and/or the support of ongoing care of care home residents (40%) Twenty-four per cent (26/109) of GMDs gave ongoing input to care homes Most PCTs had a standardised assessment for admission to care homes with assessment of need for nursing mainy competed by a socia worker and/or nurse RESULTS FROM THE REVIEW OF SURVEYS AND THE REVIEW OF REVIEWS 2. Monaghan and Morgan, Ora heath poicy and access to dentistry in care homes; (a types of care homes, Waes) To expore the factors that may faciitate or impede access to denta care and arrangements within care homes in Waes A care homes in Waes. Posta survey with 10% random sampe interviewed. Questions focused on new residents, denta assessment and access to routine and emergency care, denta care faciities, ora awareness, hygiene practice, diet and nutrition Six hundred and seventy-three care homes without on-site nursing, 88 with nursing and 186 dua. Eighty-one per cent response rate, 957 out of 1185 N/A Managers reported more difficuty in accessing routine denta care than emergency denta care of the reported findings. Twenty-four per cent of care homes reported aways having probems accessing routine denta care. Eighteen per cent of care homes reported aways having probems accessing emergency denta care

65 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. Author, tite; year (home type) Aims Survey detais Sampe size/response rate GP services Other services 3. Darton, Study of Care Home Residents and Reatives Expectations and Experiences; (a care homes, Engand) 4. Seymour et a., Do nursing homes for oder peope have the support they need to provide end-of-ife care? A Mixed-Methods Enquiry in Engand; (nursing homes, Northern Engand) To examine oder peope s expectations and experiences of iving in a care home and to coect information from reatives about choosing a care home To identify key factors in the wider heath and socia care system infuencing the quaity of end-of-ife care provided in nursing homes Focus was on residents decision to move into a care home and the difference between their expectations and experiences. Some data were coected on heath service use before and after the move. An initia interview was conducted with a foow-up 3 months ater A posta survey to 180 nursing home managers. It incuded questions about the profie of deaths in the homes, access to externa support and barriers to and perceived priorities for improving end-of-ife care. A mixed-methods study incuding two quaitative case studies comprising interviews with seven care home staff and ten stakehoders nominated by them Random sampe of 150 care homes approached in six regions of Engand, ocated via the CQC website. Sixty-seven per cent response rate: 605 out of 900 homes recruited. Sixty-nine residents participated and 33 reatives from 46 care homes There was a 46% response rate (82/180) Since admission, 80% of residents had had a consutation with a GP or a practice nurse and 30% had been to hospita Reatives reported that 92% had had a consutation with a GP or a practice nurse and 46% had been to hospita Most externa support for end-of-ife care was provided by GPs of the 72% (n = 59) of care home managers who responded to this question, 97% (n = 58) of them reported that they received some or aot of support from GPs The majority of responders 93% (n = 76) reported iaising with between 1 and 11 practices (mean, n = 5) and a range of 1 34 individua GPs (mean, n = 12) Variabiity of support from GPs with end-ofife care highighted Prior to admission over one-third of residents received chiropody services but few received other services. After moving in, twice as many receive chiropody. Few residents received occupationa therapy and none received speech therapy. Reatives reported that residents were no more ikey to receive other therapy services than before admission. Reatives reported that 79% of residents had hospita treatment prior to moving in and that the medica or nursing care provided in the home was of a higher standard Eighty per cent of care homes received support from speciaist nurses, 51% from DNs (n = 30) and 54% from speciaist paiative care nurses One-quarter of homes did not request hep from speciaist paiative care teams (24%) Sixty-six per cent of responding care home managers reported that they accessed paiative care support via a direct advice ine to the oca hospice or Macmian nurses Support for residents with cancer was viewed as better than support given to those with dementia continued DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO

66 30 NIHR Journas Library TABLE 2 Summary of data extracted from topic-specific surveys (continued) Author, tite; year (home type) Aims Survey detais Sampe size/response rate GP services Other services 5. Briggs et a., Standards of medica care for nursing home residents in Europe; (nursing homes, in 25 European countries) 6. British Denta Association, Dentistry in Care Homes Research; (Engand, Waes, Northern Ireand and Scotand) To investigate whether or not 25 countries in Europe have guideines to formaise the medica care deivered to oder peope iving in nursing homes To investigate care home residents denta care, incuding access to dentists, care home staff input and knowedge Survey e-maied to representatives of 25 European geriatric medicine societies asking if their heath service or professiona group: 1. Required geriatric medicine training for doctors working in nursing homes? 2. Had written medica standards for nursing home care? 3. Had a nursing home doctor society? If yes, did it have written medica care standards for nursing homes? Semistructured in-depth quaitative teephone interviews with a core framework of topics were conducted with managers from 13 care homes and an onine survey was sent to 39 cinica directors who reported to deiver services to care homes One hundred per cent response rate from 25 geriatric medicine societies in 25 European countries Homes chosen that covered a range of sizes, ownership, ocation and resident needs No information on how homes were recruited and how many decined to take part The Netherands was the ony country where the nationa GP society had written medica care standards for nursing homes N/A Five out of 25 (20%) heath services required specific training in geriatric medicine for doctors working in nursing homes Four out of 25 (16%) geriatric medicine societies had written medica care standards for nursing homes four out of 25 countries had a nursing home doctor society and one had pubished medica care standards for residents Haf of the care home managers reported that their residents received reguar check-ups. Homes were eveny spit between those that used high-street dentists and those that used saaried primary care dentists Managers reported a ack of information about NHS providers and those wiing to provide domiciiary care RESULTS FROM THE REVIEW OF SURVEYS AND THE REVIEW OF REVIEWS A purposive sampe of 39 respondents who competed the annua survey of cinica directors and indicated that they provided dentistry to care homes, were approached to take part in an onine survey. Twenty-six responded, a 67% response rate Homes with saaried dentists were more ikey to have reguar check-ups and to receive domiciiary care

67 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. Author, tite; year (home type) Aims Survey detais Sampe size/response rate GP services Other services 7. CQC, Heath Care in Care Homes. A Specia Review of the Provision of Heath Care to Those in Care Homes; (a homes, Engand) 8. Morgan et a., Waes Care Home Denta Survey ; (no detais of homes, Waes) Provides new evidence on the key issues affecting oder peope with dementia iving in care homes To investigate any unmet denta care needs in a sampe of care home residents Three questionnaire surveys distributed to the Azheimer s Society members (and care homes contact detais obtained from reguators, and through some Azheimer s Society staff and onine. No detais of how they were recruited) YouGov/Azheimer s Society 2012 survey of UK aduts regarding dementia and care homes 65 (n = 2060 aduts) Suppemented the Adut Denta Heath Survey 2009: common ora heath conditions and their impact on the popuation 74 to compare with oder peope iving at home Cinica data coected by dentists and questionnaire data on service use by denta nurses. Excuded residents who coud not consent Reatives of oder peope 1139 responses Care home staff) 647 responses from a direct maiing to 300 care homes in Engand, Waes and Northern Ireand YouGov: 34 out of 2060 responses Twenty-eight care homes randomy seected and five residents in each randomy seected to take part. No detais on response rate and sampe size; approximatey 708 residents were examined and/or questioned about denta care, unknown how many participated Reatives: 56% (n = 637) of respondents said access to and support from GPs was good Care home staff: 55% (n = 354) reported that the resident got enough support from the GP N/A Large numbers of respondents did not know about access to denta services. Care home staff views on support from dentists was mixed, with ony 23% (n = 259) saying access was good; 44% (n = 286) said that residents got enough support from the dentist Thirty-six per cent (n = 408) of reatives were positive about access to, and support from, other heathcare services Forty-three per cent (n = 281) of care home staff were positive about residents getting enough support from other heath services The majority of residents woud ony attend the dentist when having troube. Residents with their own teeth were much ess ikey to report reguar denta check-ups (19%) than oder peope iving at home continued DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO

68 32 NIHR Journas Library TABLE 2 Summary of data extracted from topic-specific surveys (continued) Author, tite; year (home type) Aims Survey detais Sampe size/response rate GP services Other services 9. Morris J, Patients Association/University Coege London/BGS Cinica Steering Group, 2013, persona communication; nursing homes, care home organisation specific 10. Leemrijse et a., The avaiabiity and use of aied heath care in care homes in the Midands, UK; (a care home types) To identify good practice and barriers to the deivery of an integrated approach to end-of-ife care in 10 Barchester homes To estabish the access to, and use of, services provided by aied heath professionas to care homes in Oxfordshire and Warwickshire Face-to-face or teephone interviews with 10 care home managers and eight GPs Focus was on how GPs work with care homes in reation to end-of-ife care Cross-sectiona posta survey on use of service, frequency of use, referra mechanisms, funding and most common probems service sought for Incuded in the survey were physiotherapy, occupationa therapy, chiropody, dietetics, optometry, speech and anguage therapy, compementary (aternative) therapy, hearing services and socia activity organisation services Ten care home managers, eight GPs (10 approached) Ninety-five per cent (115/121) response rate from care homes Nine out of 10 care homes had attached GPs; one care home worked with mutipe GPs Four out of eight care homes worked cosey with the GP; the other four had probems getting them to visit DN, district nurse; GMD, geriatric medicine department; N/A, not appicabe; OT, occupationa therapist; PCT, primary care trust; SALT, speech and anguage therapist. N/A Seven care homes had access to paiative care services Three care homes had access to the DN Two care homes had access to the od age psychiatrist One care home had access to the geriatrician The majority of care homes had access to chiropody (91%), optician (86%), audioogy (63%) and physiotherapist (65%). Less than haf had access to an OT (41%), dietitian (44%), SALT (39%). One-third used an aternative therapist and socia activities organiser. Sources of funding for services were variabe with up to 15 variations and a high proportion of aied heath care was privatey funded. Referra mechanisms were compex, with care homes uncertain how to refer residents to the NHS and socia services RESULTS FROM THE REVIEW OF SURVEYS AND THE REVIEW OF REVIEWS

69 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 There was imited information in the surveys about how wider NHS provision was organised for care homes, athough some information on geriatrician services was reported in the survey by Steves et a., 21 and on denta care by the British Denta Association s survey. 71 Reports that focused on speciaist or topic-specific service provision for care homes are summarised in Tabe Nurses with different areas of speciaism visited the homes; eight types were identified. Community psychiatric nurses (CPNs), however, were not mentioned and overa menta heath services were under-represented. District nurses were the most frequenty mentioned group, but nursing care coud be organised as nurse-ed teams or nurse speciaists dedicated to working with care homes, fas prevention services, continence care, tissue viabiity, Parkinson s disease nurse and paiative care nurse speciaists. Access to speciaist services and genera dentistry services was reported as probematic for some care homes but not a. Athough most surveys reported frequency of contact and range of services, there was minima information about how services were organised, who was seen and how often or whether or not the quaity or range of provision of the care was assessed. In the surveys reviewed, it was not possibe to differentiate between services that care homes coud theoreticay have access to and what was actuay being deivered to care homes. This was the case even in surveys that incuded care home-specific services. Moving into a care home did not improve access to heath-care services, 68 athough some nursing home residents were more ikey to see a geriatrician. The variation in the organisation, provision and funding of heath services, both generaist and speciaist, to care homes coud not be expained by resident need or care home type. Two consistent findings emerged: first, wide variabiity in the provision of services to care homes and, second, widespread ack of denta services. Both signa inadequate care for residents. The surveys reviewed were heterogeneous and their quaity, athough not formay assessed, was variabe. Consequenty, caution shoud be exercised in generaising these resuts. There have been severa surveys pubished between 2007 and 2015 and, athough their findings are variabe, they a point to a picture in which there is itte agreement between commissioners and providers as to how services to care homes shoud be organised. Litte has changed in this regard since the first nationa survey in The findings indicate that there is imited vaue in further descriptive work on NHS heath-care service provision to care homes that is not inked to an understanding of how the services work with care home staff to improve care home residents heath-reated outcomes. Review of reviews The review of reviews compemented the review of surveys in addressing the first research question about the range and type of provision with the aim of estabishing an evidence base for existing approaches to service deivery. We identified 13 systematic and narrative reviews that focused on care homes and heath-care provision. Seven were excuded: four because they focused on care home working and heath-care provision without reference to working with externa heath-care provision, and three because they provided overviews of heath-care provision to care homes, but referenced rather than discussed reevant research studies. Six reviews were incuded, which comprised two Cochrane reviews; 85,86 one scoping review on the provision of ora heath care to care homes; 87 one annotated bibiography of research in care homes that deveoped a thematic review of modes for improving care in residentia care homes; 26 one on predictors of hospitaisation from US nursing homes; 88 and a systematic review of quaitative research 89 on resident accounts of iving we in care homes that incuded studies reevant to the experience of care. Tabe 3 provides a summary of the incuded studies. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 33

70 34 NIHR Journas Library TABLE 3 Review of reviews ist of incuded studies Tite and type of review Primary aim/objective Incuded studies 1. Interventions for improving paiative care for oder peope iving in nursing care homes; 86 systematic review 2. Interventions to optimise prescribing for oder peope in care homes; 85 systematic review To determine effectiveness of muticomponent paiative care service deivery interventions for residents of care homes for oder peope To determine the effect of interventions to optimise prescribing for oder peope iving in care homes Two RCTs and one controed before-and-after tria Seven hundred and thirty-five participants A based in the USA (studies graded as poor quaity and at risk of bias) Eight studies: six custer RCTs and two patient RCTs Interventions/modes of service provision Interventions differed in the three studies: 1. identification of residents ikey to benefit from speciaist paiative care with onward referra 2. deveopment of cross-discipinary paiative care training and preparation pus onward referra 3. creation of specia unit within the care home Diverse, muticomponent interventions Medication review part of seven studies Mutidiscipinary case conferencing three studies Two studies, education for care home staff Reevant outcomes considered Use of hospice care, hospita admissions Length of hospita stay Pace of death Famiy ratings of quaity of care Residents symptoms: pain, discomfort, distress, compications Documentation, for exampe ACP Adverse drug outcomes Hospita admissions Mortaity Quaity of ife Medication-reated probems Findings reevant to service deivery to care homes Improved quaity of care as assessed by famiy. Improved residents comfort, but did not address behaviours associated with dementia or physica compications A interventions reied on effective communication between cinicians and care home staff and training being provided Authors comments suggested effective communication was ikey to be essentia but not sufficient to achieve change Interventions ed to the identification of medicationreated probems and some evidence that medication appropriateness was improved. No effect on adverse drug events, hospita admission and mortaity. Equivoca findings on costs. Need for a consensus on what are important resident-reated outcomes RESULTS FROM THE REVIEW OF SURVEYS AND THE REVIEW OF REVIEWS Residents (n = 7653) in 262 care homes in six countries Decision support technoogy, one study Medication appropriateness Overa quaity rated as ow or very ow Majority invoved mutidiscipinary working with pharmacists competing the review Medicine costs

71 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. Tite and type of review Primary aim/objective Incuded studies 3. Predictors of nursing home hospitaization: a review of the iterature; 88 iterature review To review the evidence for the association between the decision to hospitaise and factors reated to the residents wefare and preferences, the providers attitudes and the financia impications of hospitaisation Reviewed papers from 1980 to 2006 Fifty-nine studies (incuding two RCTS) examining predictors of nursing home hospitaisations a Interventions/modes of service provision Resident-eve data coected prospectivey and retrospectivey: studies using nursing home data (n = 27), hospita data (n = 10), mutipe data sources (n = 19), interviews and survey (n = 3) Reevant outcomes considered Hospitaisations Use of the emergency department Findings reevant to service deivery to care homes Variabiity in how hospitaisations are defined in the iterature: preventabe/ avoidabe/discretionary Age and specific heath conditions are associated with admissions, for exampe congestive heart faiure, respiratory infections; however, severe cognitive impairment is not associated with admissions and the authors ask if this refects a reuctance to treat peope with dementia Past hospitaisations associated with future hospitaisations Patient preferences were an infuence on referra to hospita Presence of nurse practitioners/ physician assistant may reduce hospitaisations (ess cear about access to a physician) Equivoca findings about nursing staffing eves Access to hospice care reduced hospitaisations in some studies How residents were funded = negative incentive to reduce hospitaisation for some state-funded residents continued DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO

72 36 NIHR Journas Library TABLE 3 Review of reviews ist of incuded studies (continued) Tite and type of review Primary aim/objective Incuded studies 4. In-reach speciaist nursing teams for residentia care homes; 27 iterature review To bring together avaiabe evidence reevant to various approaches to improving care in residentia care homes Thematic review Interventions/modes of service provision Annotated bibiography under seven themes Residents and reatives views on care (73 papers) Cinica areas (107 papers) Medication in care homes (34 papers) Medica input into care homes (21 papers) Nursing care in care homes (66 papers) Reevant outcomes considered Narrative account Notes absence of resource use outcomes Findings reevant to service deivery to care homes Evidence mainy reated to nursing homes Debate about the reationship between quaity of care and quaity of ife in nursing and residentia homes. Measures of socia care, as we as cinica care, needed Need for better management of medication in nursing homes Medica cover is suboptima. GP workoad shoud be more proactive RESULTS FROM THE REVIEW OF SURVEYS AND THE REVIEW OF REVIEWS Hospita Admissions (43 papers) Medicine appropriateness outcomes The quaity of interinstitutiona transfers and patient safety across settings is important Modes of care improvement in care homes (113 papers) Partnership working between DNs and care home staff intermittent, with ess evidence on therapist input to care homes

73 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. Tite and type of review Primary aim/objective Incuded studies 5. Living we in care homes: a systematic review of quaitative studies; 89 systematic review 6. A scoping review and research synthesis on financing and reguating ora care in ong-term care faciities; 87 scoping review To conduct a systematic quaitative review of care home ife How is ora heath care for frai eders financed and reguated in LTCFs? Systematic review Interventions/modes of service provision Thematic accounts of residents and reatives views b Reevant outcomes considered Maintaining independence Findings reevant to service deivery to care homes Main focus was on the care received in the care home and not from visiting HCPs Thirty-one studies Safety Continuity of care and ess rigid time schedues and routines were important to residents. This has impications for how they experienced care Scoping/reaist review with stakehoder invovement Sixty-eight papers Different systems: financing systems for ora heath care pubic funding, insurance systems, managed care and contractua agreements Reguations Fee-for-service or saaried appointments Professiona segregation between dentistry and medicine Carers technica knowedge and competency in nursing Access to ora heath care in LTCFs Attitudes of staff and caring practices were aso important Inadequate reguation of ora heath care in ong-term care a reason for ack of provision Uncertainty of treatment needs Despite government-sponsored incentives for dentists, very few dentists work with specia popuations How denta services are financed and organised affects access to care Portabe denta equipment possibe, but not iked by dentists MDS systems are not used to compete assessments of ora heath and not prioritised by staff in ong-term care ACP, advanced care panning; DN, district nurse; MDS, minimum data set; RCT, randomised controed tria. a North America ony. b About iving in a care home; very itte focus on heath-care provision. Four key themes: (1) acceptance and adaptation, (2) connectedness with others, (3) a homeike environment and (4) caring practices. DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO

74 RESULTS FROM THE REVIEW OF SURVEYS AND THE REVIEW OF REVIEWS A the incuded reviews highighted the absence of agreement in the studies about outcomes and how these were defined (e.g. what is the difference between a discretionary hospitaisation and an avoidabe hospitaisation?). How interventions in the different studies coud be transated into improvements in resident-reated outcomes, and particuary quaity of ife, were aso either not considered in detai or discussed. Findings on the prescribing of medication had simiar findings. The reviews that focused on care home-specific service provision 26 and topic-specific services such as paiative care, 86,90 ora heath care 87 and prescribing of medication 85 aso confirmed what the review of surveys found, namey that heath-care support and provision were erratic and access to care was often imited and constrained by funding and how imited resources were aocated. Interventions were characterised by their heterogeneity of approach. For exampe, the muticomponent nature of paiative care and medication-reated interventions often invoved education of care home staff and structured and informa approaches to communication between cinicians and care home staff. This meant that it was uncear which eements of the intervention were essentia to either supporting or triggering change in the observed outcomes. One review in its discussion of the evidence observed that... none of the studies attempted to disentange the back box effect, that is to understand the effects of the contributing components. 85 Two reviews considered reatives and residents accounts/views of care. 26,89 Bradshaw et a. s 89 review of residents and reatives accounts of what was important to them focused mainy on ife within the care home and reationships with care home staff rather than the care provided by HCPs. Athough it is possibe that residents do not differentiate between different professionas, their insights remain reevant. For exampe, residents identified continuity of care, and that nursing staff were technicay competent and knowedgeabe, as important. Interestingy, and reevant for the organisation of heath care from outside the care home, the routines of the care home were aso identified as important and the abiity to be fexibe in how and when care was schedued. The review by Szczepura et a. 27 identified US work that estabished that high catheter use, poor skin care and residents ow participation in organised activities are associated with negative outcomes for residents and that this was improved when staffing eves were higher and nurse turnover was ower. Bradshaw et a. 89 concuded that a key theme was how care was provided (summarised as caring practices) and that this was possiby predicated on the resource constraints such as ack of staff, avaiabiity of training and supervision, which was aso suggested by Szczepura et a. s 27 review. These are issues that ink to how visiting heath-care services work with care homes and whether or not they emphasise support and education of care home staff as some or part of their roe and responsibiities. The review that focused on hospitaisations from nursing homes imited its scope to North American studies. 88 As aready noted, this has impications for transferabiity to the UK setting. However, it did provide a usefu overview of resident characteristics that are more or ess ikey to infuence hospitaisations. The majority of incuded studies coud draw on arge minimum data sets about residents characteristics and resource use. Resident numbers in the incuded studies ranged from 67 to 36,702, and, when incuded, from 1 to 527 nursing homes. The reviewers were abe to identify with confidence the associations between age, particuar conditions (e.g. pneumonia) and hospitaisations and raised some interesting questions about why residents with cognitive impairment had ower service use than those who were cognitivey intact. Summary In summary, the review of surveys and the review of reviews provided a compementary and comprehensive commentary on the erratic provision of services to care homes. The ack of consensus on what needs to be in pace to support care homes was compounded by imited evidence about how to measure effectiveness 38 NIHR Journas Library

75 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 and whether certain approaches to providing heath care were more or ess effective or acceptabe to care home residents and staff. The muticomponent nature of interventions and the impact of contextua factors such as oca history, funding and staffing on outcomes were consistenty highighted. More recent work emphasised the ack of residents access to ora heath and dentistry services. The findings reinforced the vaue of taking a theory-driven approach to try and understand the muticomponent nature of provision, and the peope, structures and organisations that (possiby) need to be in pace for heath-care provision to be effective. It provided a patform for a theory-driven review of the evidence and deveopment of programme theories for testing in the case study phase. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 39

76

77 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Chapter 4 Reaist synthesis Introduction The reaist synthesis was undertaken to investigate what happens when different modes of heath-care deivery attempt to achieve improved outcomes for care home residents, and to deveop a programme theory for further testing in the study. Specificay, the reaist review addressed the second question of the project: What features (mechanisms) of heath service deivery modes are the active ingredients defined as being associated with positive outcomes for care home residents? Our previous findings from the review of surveys and review of reviews estabished the range and type of approaches to heath-care provision and how outcomes were defined in the iterature (see Figure 1). For the reaist review we focused on research that yieded information on one or more of five outcomes for residents: medication use; use of out-of-hours services; hospita admissions, incuding emergency department attendances; ength of hospita stay; and user satisfaction. This chapter provides an account of the findings from the three stages of the synthesis. The review protoco 30 and a detaied account of the stakehoder interviews from phase 1 91 are pubished esewhere. Stage 1 stakehoder invovement Stakehoders were interviewed to expore their perspectives on heath-care provision (Tabe 4). Twenty-one peope were interviewed as representative of the views and experiences of care home organisations, residents, the reguator, and commissioners of heath and socia care for care homes. We were ony abe to interview three residents face to face. The resident data were therefore suppemented with a secondary data anaysis (ed by CV) of 34 residents interviews from an earier care home study in which residents had been asked to discuss what they thought about the heath care they received. 2 A the stakehoders were asked to consider what good heath-care provision might ook ike. The insights provided by participants drew on observations of good practice, hearsay and persona experience. There was a consensus that therueofthumb shoud be that residents have equivaent access to heath care to those who ive at home, specificay, access to a GP when needed. Stakehoders aspirations for heath care were anaysed TABLE 4 Stakehoder interviews Roe Number of stakehoders Care home organisation owner/representatives 7 Residents representatives 4 CQC (reguator) 4 Commissioners of heath and socia care for care homes: CCGs (heath) and oca authority (socia care) 6 Care home residents (34, secondary data anaysis) 37 Tota 58 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 41

78 REALIST SYNTHESIS and interpreted to identify their impicit theories of change, concerning how to impement evidence-based practice, manage risk in primary heath care and achieve integrated working between heath and socia care. The foowing summarises what the different stakehoder groups highighted as important. Care home organisation owner and care home representatives The narratives found in the care home managers and their representatives accounts stressed the importance of responsive access to care and services that refected the residents needs. These views were suppemented by the frequenty stated beief that services to care homes were rationed. This was attributed to ageism and a misunderstanding of the knowedge and skis of care home staff. Good care was therefore characterised by both the frequency and the quaity of contact with the GP and inked community services that the care home received: Good means for me, for exampe, GPs who proactivey go into care homes, who have good reationships with the staff, who have the proactive work but who are aso prepared to be reactive, who might have good inks to their genera hospita as we so that they know when to admit and that they re comprehensive in their approach to wider admitting. They communicate we with the staff at the care home as we. Care home representative organisation stakehoder Activities that supported continuity of the contact with heath-care providers, mutua respect and shared opportunities for earning appeared to be key. This was seen as important in deveoping a common understanding of when and how to invove services if care home personne were concerned about a resident s heath. Exampes of such activities incuded joint training events and care home staff being consuted by NHS providers. Mutua confidence was buit on previous experience of having resoved probems together. One care home manager described how residents needs were met and hospita admissions avoided because she knew that GPs woud respond when asked, woud isten to the home s assessment of a resident s needs and woud support care home staff to provide care. Athough the GP was identified as the inchpin, it was the quaity of the association with the care home that was emphasised as crucia. Residents and residents representatives accounts Residents accounts presented them as being in the centre of a fow of HCPs (doctors, nurses, opticians, podiatrists, dentists), who a visited them at various times, perceived as ad hoc and unco-ordinated. Residents compared this experience with the service they had received whie iving at home. They fet that some provision was not avaiabe in the home, or that they now had to purchase it (e.g. podiatry and dentistry). They were uncear about the organisation of heath care and did not understand the roes and responsibiities of visiting HCPs. The ack of a persona reationship with visiting heath-care staff was aso highighted, for instance district nurses and GPs being too busy to tak or being task focused and not engaging with the resident. Residents did identify the key roe of care home staff in iaising on their behaf with care home staff, knowing their heath-care needs and deciding when to refer to a GP. One resident with a eg wound saw that her needs were secondary to the needs of other, possiby more unwe, residents. She described a protracted process of decision-making as increasingy senior staff decided if a HCP shoud be caed: We that I don t know. I just fee I m on a sort of, waiting, I m not as i as a ot of peope so I think I m just eft to tick over... this morning, I was seeing the senior nurse who comes with the others [care staff] and te her and she s had a ook and she s going to be in touch, get in touch with somebody ese who is higher up sti, who is going to ook at it this afternoon. Resident NIHR Journas Library

79 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Residents were aso aware that care home staff may not be trained adequatey to undertake the assessment and monitoring roe expected of them by residents; this created some uncertainty. This point was emphasised by a stakehoder from a resident representative organisation who inked it to the need for speciaist input into care homes: Peope are very worried about post-stroke patients not having the kind of rehab [rehabiitation] they d get if they were at home or staying onger in hospita... It s not just GPs, think about denta care... You know, you reay need peope ike geriatricians who are speciaist in the care of oder peope... the peope who speciaise in od age psychiatry aso need to have a key roe... there needs to be an a-round service pan. Stakehoder SH11 Commissioners accounts Commissioners recognised that many of the chaenges surrounding heath care for care homes came down to efficient integration of heath and socia care. Negotiations about what provision shoud be pubicy funded as heath care and what shoud come under the jurisdiction of the care home were a significant preoccupation. It coud be a jugging act, having to reconcie the different priorities of care homes and heath-care providers. Commissioners saw their priority as purchasing activities ikey to reduce demands on hospita services; specificay, opportunities for heath care and care home staff to work together. They saw the objective as pre-empting crises and reducing misuse of secondary services such as A&E, which coud arise because of a ack of confidence or cinica knowedge in care home staff. Typica activities to achieve this end incuded financia incentives for GPs, which coud be appied to more frequent visits to work more cosey with care home staff, the creation of Care Home Speciaist roes or the servicing of existing working reationships. One GP commissioner, drawing on persona experience, said that it was possibe to maintain continuity and the desired outcomes of care if there was a specific key worker in the care home for NHS staff to iaise with. Simiary, a oca authority commissioner identified the importance of nominated care home staff ( champions ) being aocated to work with visiting NHS staff. This was considered to structure their working together, which was often needed in situations of pressure and imited resources: It s getting the right peope from these particuar homes to have that spare time to come aong and get invoved... It s having champions; it s making sure that each home has their particuar champion on particuar (heath) topics and they ve got ownership of that particuar subject. Loca Authority Commissioner Stakehoder SH5 Commissioners aso emphasised the importance of audit and review of heath-care service deivery to care homes. Athough interna quaity assurance is undertaken by commissioners, independent audit is the roe of the CQC, as the reguator. The accounts of the reguator Reguator representatives, with responsibiities for reguation of services, highighted the use of incentives or performance management as a means to support residents access to heath care. They characterised good heath care as evidence based and age appropriate, as we as being continuous in terms of the reationships between residents and visiting professionas. In keeping with the roe of the CQC inspectorate, interviewees highighted the consideration of untoward incidents and suboptima care. Exampes of eder abuse, avoidabe deaths, high rates of pressure sores and unnecessary antipsychotic prescribing were given as the reasons why monitoring needed to be in pace. One stakehoder acknowedged the importance of good GP care home reationships, but saw that these need to be underwritten by expicit agreements and what was characterised as a proper system of care: It s where they ve got a proper agreed arrangement with that GP surgery around, you know, they visit at certain times of the week and they can be contacted if there are any probems. Ensuring then that peope have a propery panned package of care that is reay focused on their needs. Stakehoder SH13 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 43

80 REALIST SYNTHESIS Inferences from interviews Stakehoders paced different emphases on what was contextua and what was seen as essentia to the achievement of effective heath care for care home residents. They had different impicit theories about why they thought that practitioners responded to various approaches used. These conceptua frameworks informed how effective modes of service deivery were described. They therefore shaped our first reading of the iterature and how we eaborated ideas about CMO configurations for further review. Scoping of the iterature In parae to the interviews reported above, we competed an initia scoping of the iterature. This added to the findings from the review of surveys and the review of reviews. The database searches initiay considered care home-wide interventions and then topic-ed interventions that were inked to one or more of the five outcomes. This generated 556 records. Foowing screening and de-dupication, 64 fu-text artices were assessed (see Appendix 9 for a tabe of incuded studies by design and focus). These papers were read by Sue L Davies, Meanie Handey, Maria Zubair and Caire Goodman and grouped by their study design, topic, approach and outcomes. A reaist approach to incusion of evidence and data extraction recognises that within one document different types of data can iuminate or buid one theory, refute another or offer an aternative interpretation. This guided the seection (for incusion or excusion) and appraisa of the contribution of pieces of data within a document. Quaity was assessed based on the eve of detai the paper provided; specificay, the amount of information provided about the intervention and participants responses, acknowedgement of the underying theories or assumptions guiding the study and discussion of what the findings reveaed about heath care and care home staff working together. Quaity of the research design and method were assessed on the basis of rigour and carity. Athough we did not appy a hierarchy of evidence, descriptive accounts that focused on achievement or innovation, but provided minima information about the process, were excuded. Evidence that came from professiona opinion (e.g. bogs or commentaries or detaied accounts of an innovation) was treated as specia cases to be discussed at team meetings. Within-team discussion and preiminary mapping of the scoping resuts focused on what the studies reveaed about different possibe CMO configurations. For exampe, discussions focused on the significance of how the organisation and make-up of mutidiscipinary teams (MDTs) or singe discipinary groups working with care home teams and their patterns of contact were associated with earning opportunity uptake by care home staff. We aso expored how shared protocos were introduced into care homes and who was invoved in their impementation; what kind of feedback on HCPs performance was thought to ead to change in visiting professionas responses; the use of sanctions and financia incentives; and which groups of practitioners were expected to impement change. (Note that these searches were rerun and a more detaied data extraction competed for the detaied review of the evidence.) In ight of the stakehoder interviews, the scoping of the iterature, review of surveys and review of reviews, we deveoped six statements. These were drafted to seek to capture the possibe expanations for what needs to be in pace for the provision of heath care to care home residents and rewritten as if then statements as the basis for refinement into potentia programme theories for consideration by the SSC. Our purpose was to test each statement s pausibiity and utimatey to guide the in-depth evidence review of how heath-care services to care homes improved the heath of residents and use of services. The foowing statements posited that heath outcomes for care home residents coud be improved under certain conditions. 1. If taiored education and support for care home staff are provided by cinica experts and supported by the use of structured documentation and protocos then resident outcomes wi be improved. This wi come about through prioritising specific assessment/care activities that trigger changes in how residents care is panned, and in how care home staff recognise and frame their need for training and support from visiting cinicians. 44 NIHR Journas Library

81 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO If contracts and financia incentives are provided to GPs to provide dedicated services to care homes and monitored against prespecified process and outcome measures, then the pattern and frequency of GP contact with residents and staff wi change. This wi come about through increasing the time and opportunities for screening and review of care, and enhancing staff confidence that they can access a GP. This wi reduce demand on emergency and secondary care services. 3. If formaised recognition and ongoing faciitated support are provided to care home staff to equip them to buid reationships and work with heath service providers, this wi increase their confidence when working with visiting HCPs and enabe them to identify priorities for residents heath care with visiting HCPs. This wi come about through enhancing and vaidating the expertise of care home staff. It wi reduce demand on emergency and secondary care services. 4. If care home champions are appointed, who have expertise in quaity of care for oder peope and designated responsibiity to work with care homes then they wi faciitate continuity of support to the care home staff workforce. This wi come about through their promotion of knowedge exchange and encouragement of skis acquisition by care home staff. The resut wi be that staff are more proactive in providing age-appropriate heath care to residents. 5. If the commissioning and provision of services focuses on specific probems of od age and refects the heath-care needs frequenty experienced by care home residents (e.g. fas prevention, end-of-ife care, continence management) then the focus of services wi shift to a more individuaised pattern of care. As a resut the heath care provided wi be perceived by residents as equivaent to that which they received iving in their own homes. 6. If there is investment in the creation of interorganisationa and intersectora networks at the organisationa eve; between heath and socia care providers in the pubic and private sectors, then this wi change how different services work together. It wi come about through highighting gaps and overaps in service provision. It wi trigger conversations and panning between services about resource use and who is responsibe for providing heath care. As a resut, provision wi become more efficient. Further presentation and discussion within the team and the study steering group focused on areas of overap and fit with their experience of heath-care provision to care homes. This generated three broad programme areas from which the team then deveoped possibe CMO configurations for rigorous testing in the detaied review of the evidence: 1. system change and cross-organisationa working between care home and visiting heath-care staff 2. age-appropriate care accessed by oder peope resident in ong-term care 3. reationa approaches to promote integrated working between visiting heath-care and care home staff that emphasise interpersona skis and shared decision-making. The search strategy and derived theories are shown in Figure 2. The three programme areas are now discussed in turn. System-based quaity improvement mechanisms to improve heath-care outcomes: the use of incentives, sanctions and targets The assumptions underying system-based incentives, targets and sanctions are that they prompt behavioura change through targeting particuar professiona groups or organisations, focusing on the improvement of specific processes or outcomes, and thereby improve quaity of care and reduce inequity of provision. 92 The Quaity and Outcomes Framework (QOF), introduced for GPs in Engand in 2003, inked financia incentives to the quaity of care that is provided by practices 93 and has been described as a ever to reduce heath inequaities and reinforce evidence-based practice. 94 Based on the different theoretica perspectives about how system-based approaches might work to improve one or more of the five outcomes we posited a possibe CMO configuration to test and refine the evidence reviewed on the use of incentives, sanctions and targets (Box 3); it made expicit how we understood the intervention. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 45

82 REALIST SYNTHESIS Survey of surveys of heath-care provision in care homes Review of reviews of care home interventions Outcomes of interest Medication use Out-of-hours service use Hospita admissions Length of hospita stay User satisfaction Stage 2 Stage 1 Scoping search resuts (n = 556) Searches of care home interventions and topic-ed interventions reporting outcomes of interest Date restricted to post 2006 Databases: PubMed, CINAHL, The Cochrane Library, DARE, HTA Database, NHS EED, Scopus, Soc Abs, ASSIA, BibioMap, Sirius, OpenGrey, Socia Care Onine, the Nationa Research Register Archive, NIHR, Googe, Googe Schoar Theory 1 Stage 2 search resuts System-based quaity improvement Language restricted to Engish Date restriction: excuded pre 2006 Databases: PubMed, CINAHL, The Cochrane Library, DARE, HTA Database, NHS EED, Scopus, Soc Abs, ASSIA, BibioMap, Sirius, OpenGrey, Socia Care Onine, the Nationa Research Register Archive, NIHR, Googe, Googe Schoar Latera searches Tota resuts (n = 687) Theory 2 Fu texts taken forward for eigibiity, n = 99 Age-appropriate care Stakehoder interviews (n = 58) Interviews, n = 24 Secondary anaysis, n = 34 Theory 3 Grey iterature search Reationa approaches to promote integrated working The foowing organisation databases were searched: MyHomeLife Network Nationa Care Home Research and Deveopment Forum DeNDRoN CRNs Care Engand Nationa Care Forum Residents and Reatives Association Papers taken forward for in-depth review (n = 64) FIGURE 2 Search strategy for stages 1 and 2 of the reaist review. ASSIA, Appied Socia Sciences Index and Abstracts; CINAHL, Cumuative Index to Nursing and Aied Heath Literature; CRN, Cinica Research Network; DARE, Database of Abstracts of Reviews of Effects; DeNDRoN, Dementia and Neurodegenerative Diseases Research Network; HTA, Heath Technoogy Assessment; NHS EED, NHS Economic Evauation Database; NIHR, Nationa Institute for Heath Research; Soc Abs, Socioogica Abstracts. 46 NIHR Journas Library

83 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 BOX 3 Possibe CMO configuration to expain how incentives and sanctions paid to primary care can improve heath care in care homes Background There is intermittent and unpredictabe contact by GPs visiting residents in care homes; encounters with primary care are usuay unpanned and in response to an urgent need and this affects the proactive identification of residents heath-care needs, access to, and quaity of, care and frequency of acute episodes of i heath. Context GPs are provided with a range of incentives and admonitions (intervention) to visit reguary (context) and undertake structured resident assessments (context) in key areas of care (context) (e.g. medication review) and to provide the care home with support and advice in addition to individua patient visits. Mechanisms GPs are motivated to engage with the care home staff. The incentives and sanctions prompt them to be more present in the care homes, compete reviews of care home residents and work with care home staff to pan care and identify residents in need of additiona support and care. Outcomes Both care home staff and primary care professionas are more confident and satisfied working with each other. Residents have access to services required. There are improvements in care, specificay medication management and reduced use of out-of-hours emergency services and avoidabe secondary care. For GPs working with care homes, rewards inked to particuar cinica activities are used as incentives to define and increase the ength and frequency of their visits in order to achieve the desired outcomes of continuity of contact and proactive approaches to patient care 95 together with improved medication management. 85 The iterature suggests that additiona payments to GPs and pharmacists to do specific activities can improve monitoring of medication use. 96 However, the use of payments or sanctions aone to trigger GP invovement in resident assessment and review did not appear to be sufficient. 94,95,97 Three additiona contextua factors were identified from the evidence reviewed: the need for an accountabiity structure; the need to identify the named professiona responsibe for deivering a specified intervention; and care home-sensitive protocos that addressed the high number of residents iving with dementia The need to consider those residents at particuar risk and aso care home staff needs for ongoing support and training were aso fagged as important. Generay, the iterature woud appear to support the view that, athough incentives can improve the process of care and productivity (e.g. better adherence to protocos and care pathways), the evidence is imited about their impact on patient outcomes. 96,101 Charesworth et a. 96 argued that: Incentive schemes can ony work if the organisations and cinicians whose behaviour they are trying to change understand what is required [our emphasis]. Too often, the incentives are burred or inconsistent. In part, this is a resut of the compexity of the current system. Charesworth et a., 96 p. 14 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 47

84 REALIST SYNTHESIS The main pay-for-performance approach in UK primary care (QOF) aows practices to excude patients for reasons such as extreme fraity or evidence of decine. Moreover, individuas with dementia in care homes achieve ower quaity indicators in the QOF pay-for-performance system than their community-dweing counterparts. 100 This arguaby creates an impicit expectation that QOF incentives are ess reevant to care homes. 100 Indeed, as QOF indicators focus on very specific aspects of disease management, care home residents as a discrete popuation may not be recognised by GPs as a priority group in need of identification and active management. Therefore payments aone may not be sufficient to motivate GPs to make care home residents a priority, nor to address issues of accessibiity, appropriateness or system co-ordination. 102 Incentives work when they are aigned with what GPs beieve they shoud be doing. The mechanism is encouragement to practice what they aready recognise as doctors work: They [incentives] work best when a the ducks are ined up in a row: financia, organisationa, and professiona incentives, then the incentives are providing encouragement [our emphasis] to do the things that doctors beieve they shoud be doing anyway. Roand 103 One sma study focused on eary identification and support of frai oder peope and the impementation of anticipatory care. It audited cases of residents admissions to hospita as a trigger to identify and discuss with GPs the factors infuencing hospita admissions from care homes. As a resut of the discussion, the authors reported a change in GP behaviour, with an increase in care home visit rates and a reduction in overa hospita admissions. 97 However, audit and review of residents admitted to hospita had no impact on the numbers of hospita admissions sought by care home staff. The authors suggested that care home staff needed greater support from visiting HCPs and invovement in anticipatory panning for residents at risk of hospita admission particuary where there was no on-site nursing provision. This was the ony study we found that expored, and reported how, the mechanism of providing feedback on GP performance coud infuence how GPs worked with care homes. Other studies suggested that forma notification to GPs of the need to improve care or guidance on good practice (prescribing) did not provoke change. 104,105 A possibe expanation is that feedback on medication management does not have the same impact as aerts about unpanned hospita admissions that are recognised as avoidabe and costy. The urgency of the issue to the heath service, as opposed to its impact on individua residents or care homes, may be the contextua factor that infuences when audit and feedback mechanisms trigger increased engagement with care homes by NHS services. We found no evidence that targeted payments aone were a trigger to change practitioner behaviour or proactive assessment of care home residents heath-care and medication needs. One US study found that financia payments, when paid directy to care homes rather than to HCPs, improved resident outcomes, but this was for specific projects identified by care home staff. The incentive was to introduce new approaches to care, not to ensure that heath care was provided. 106 Age-appropriate care accessed by oder peope resident in ong-term care There is evidence that systematic approaches to the assessment and management of oder peope can reduce mortaity and improve function These interventions rey on the invovement of cinicians with expertise in the care of frai oder peope and their abiity to work with others to impement care pans (Box 4). An increasing body of work has deveoped interventions for care home residents that have focused on specific processes such as assessment, targeted interventions and protoco-based care. Objectives incude comprehensive assessment, 110 depression, dementia, 114 fas prevention, 115 nutrition, 39, recovery from stroke, 119 medication management, 120 end-of-ife care, tissue viabiity, 124 ora hygiene 125 and occupationa therapy. 126,127 Most of these interventions were muticomponent, but had in common the detaied assessment of residents functiona abiities and the teaching of new skis to care home staff to improve residents heath and we-being. 48 NIHR Journas Library

85 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 BOX 4 Possibe CMO configuration to expain how provision of expert practitioners in od age care can improve heath care in care homes Background Care homes have unpredictabe access to heath-care services, the majority of staff are not cinicay quaified and residents are frai and in the ast years of ife with compex heath and socia care needs. Context Experts in care of oder peope (intervention) visit care homes reguary (context) to compensate for known deficits in knowedge and skis (context) and faciitate (context) in-house earning and review. Mechanisms Care home staff fee supported and trained in how to recognise and manage symptoms and provide care to frai oder peope. They are motivated to consut with visiting HCPs and earn new skis because of the faciitation and ongoing expert support they receive. Outcomes Care home staff are more confident and skied in ooking after care home residents and specific areas of care. Residents function is improved or maintained. Staff have higher eves of job satisfaction. Care homes are ess ikey to use emergency and out-of-hours services for residents symptoms that are non-urgent. Most, but not a, interventions were appreciated by care home staff, often with reports of increased staff confidence that coud have acted as a feedback oop and a potentia additiona mechanism to infuence the improvement of residents heath. However, the positive response of staff was as ikey to have been a refection of care home staff s previousy imited experience of professiona support and encouragement. This arguaby suggests that the mechanism that triggered a change in staff (or not) was the process of working together and receiving cinica support. The underying assumption of many of the studies was that the aocation of professiona (biomedica) expertise, education and training of staff, together with proactive identification of peope at risk, woud ead to improved heath outcomes. Overa, this assumption was not supported. Our inference is that these were important contextua factors necessary for change, but not the key mechanisms that provided the generative force to ater resident outcomes. Severa contextua factors have been suggested that may inhibit care homes and/or residents abiity to engage with interventions, but these remain argey untested. Putative factors incude care home size and ownership, staff turnover, percentage of residents who have been resident in the care home for ess than 12 months, and the absence of additiona triggers or mechanisms such as the invovement of care home eadership, staff quaifications and the duration of programmes. 113 Two studies on end-of-ife training programmes found that the manager s ength of empoyment was positivey associated with use of advanced care panning documentation, improved staff satisfaction and reduced hospita deaths. Low staff turnover was aso impicated as an important contextua factor. 121,128 One study with a positive outcome appears to have been successfu because of particuar contextua factors. The key differences between the intervention process described in this study and that of the others reviewed was that it was a singe, time-specific intervention that coud be co-ordinated by one member of staff per care home. It was a simpe intervention with a quantifiabe outcome in which the proposed heath benefits to both staff and residents were cear for staff and residents. 129 Researchers 129 tested the effectiveness of an infuenza vaccine programme for care home staff (not residents) to prevent death, morbidity and heath Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 49

86 REALIST SYNTHESIS service use. The mechanisms of interest within the programme were the identification of a key ink worker within the care home and the deveopment of taiored processes to encourage vaccination uptake by care home staff. These were supported by a care home poicy for immunisation. It achieved significanty ower mortaity of residents in intervention homes than that of residents in contro homes. An expert practitioner appeared to be important as a resource that enabed the ink worker in the care home to impement the immunisation process that generated the positive outcomes. Reationa approaches to promote integrated working between visiting heath-care professionas and care home staff that emphasise interpersona skis and shared decision-making The competing priorities of heath and socia care staff, inherent power imbaances between quaified and unquaified staff, staff turnover and the difficuties HCPs have in understanding the predominanty private care home environment are we-documented barriers to effective coaboration between visiting HCPs and care home staff. 15,84,118,130,131 Reationa working draws on theories that emphasise strategies that co-ordinate and support shared probem-soving (and not baming). Working reationships are grounded in common goas, shared knowedge and mutua respect In the extraction of data in this stage of the review, reationa working was characterised as those activities and processes that emphasise shared decision-making, panning and earning, and continuity of contact between staff from different sectors (Box 5). BOX 5 Possibe CMO configuration to expain how an intervention designed to improve reationa working achieves improved outcomes for care home residents and staff invoved Background The expertise of care home staff in providing care for oder peope with fraity and/or dementia is sedom recognised by visiting HCPs. Heath-care interventions, emphasising physica heath, do not fit we with care home priorities of providing a homey setting and working practices that seek to baance positive risk-taking with patient safety. Working patterns to faciitate in-reach from numerous heath professionas are difficut to accommodate by care home staff with imited resources who want to achieve a more personaised environment for residents. Context Modes of care (interventions) that introduce opportunities for joint priority setting between care home and NHS personne (context) and processes that support ongoing discussion and review of residents heath-care needs between care homes and visiting HCPs (context). Mechanisms Identification of key personne in the care home to work with visiting HCPs triggers a response whereby staff are motivated to deveop shared priorities for care and a sense of common purpose because their views are vaued. They deveop approaches that fit with the care home working patterns and incorporate care home staff knowedge. Priorities are jointy agreed, enacted and reviewed. Outcomes Care home staff and visiting HCPs are motivated to work together and improve care for residents in agreed areas of practice. Residents function is improved or maintained, staff experience job satisfaction and the care homes are ess ikey to use emergency and out-of-hours services. 50 NIHR Journas Library

87 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 The organisation of care between the resident, their reatives, care home staff and visiting HCPs requires more than the one-on-one encounter between cinician and patient. It is a negotiated process over time, within a changing environment. Over time, there may be individua and organisationa changes in who has responsibiity for providing and/or paying for care, and changes in the arrangements for commissioning heath and socia eements of care. Roes and responsibiities for a resident s care can shift as a consequence of an acute heath event and/or a gradua shift in need from socia to heath care as compex ong-term conditions progress, and/or as part of a transition to end-of-ife care. 5,135 Three contextua factors refecting aspects of reationa working were identified as important for triggering activities and processes that were ikey to ead to improved outcomes. These were important whether or not the intervention being reported had an expicit focus on working with care homes coaborativey. They were (1) the active invovement of care staff in impementing an initiative; (2) a nominated ink person; and (3) some fexibiity in how an initiative coud be impemented. Most of the heath-care interventions reviewed were muticomponent (e.g. competion of education and training programmes, improved documentation of residents care). These were more ikey to have positive uptake and promising outcomes when they focused on a concern of mutua interest to care home and heath-care staff and/or residents and famiy, for exampe end-of-ife care that avoided unpanned hospita admissions and enabed the person to die in the care home fitted with care home staff views that they were the person s proxy famiy. The care home was the person s home, and being with strangers (hospita staff) at the end of ife was distressing for residents. 26,122 Where the initiative was identified as a priority, based on a review of resident need, but not recognised by staff as such (particuary where it added to their workoad) it was unikey to be impemented or sustained. 113,136 As one informant concerning a study that introduced a therapy-ed intervention to reduce depression observed: At times it was difficut to expain our remit to staff. We had itte time to change attitudes of some staff to issues of mobiity; making it hard to faciitate a change in practice. Eard et a., 113 p. 4 This reates to who the HCPs worked with and their roe in care deivery. Having a nominated ink person in the care home, particuary when this person coud pay a coaborative roe in reviewing, panning and supporting care, was hepfu. 129 There was evidence of improved outcomes where care home staff had fexibiity in how an intervention was impemented. 104,129,137 This was particuary the case when there was access to expert faciitation and support. Emphasis on preparatory work, structured assessment of a care home s readiness to participate, coaborative and bottom-up approaches, shared earning and the deveopment of a common understanding between care home staff and heath-care providers were key mechanisms for improvement and invovement of care home staff in the intervention. 124,131,138 In one study this invoved deveoping an intervention with care home managers that buit on previous staff earning in end-of-ife care, it was an iterative and refective process that invoved day and night staff and sought to address care home-specific issues such as supporting peope with dementia: We think this success (reduction in hospita deaths, improvement in quaity of ife for residents with dementia) is reated to the training addressing staff fears and probems [our emphasis] as we as increasing knowedge. Livingston et a., 121 p The invovement of care home staff, particuary senior staff, and other psychoogica and contextua factors that coud be characterised coectivey as a care home s readiness for change had a positive impact on the uptake of innovation. 118,139 Bamford et a. 118 found that, whie some changes coud be Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 51

88 REALIST SYNTHESIS achieved in staff understanding of nutrition, the impementation of nutrition guideines in care homes foundered because: It proved difficut to buid coective understanding of and commitment to the study resuting in inconsistent impementation... Managers commitment to the nutrition guideines did not extend to using scarce resources to faciitate impementation. Bamford et a., 118 p. 10 This finding was resonant with mutipe references in the reviewed texts to the probabe infuence of the eadership and cuture of particuar care homes on heath-care outcomes and staff satisfaction. The iterature on the attachment of a given GP practice to a care home did not support the goa of improved working reationships and resident outcomes. In fact there was evidence that such a system did not in practice ead to continuity of support. It coud have the unintended consequence of rationing care because GPs set reguar, but fixed, times for their avaiabiity. 2 There was aso evidence that one-practiceper-home arrangements coud effectivey trap providers in dysfunctiona reationships, providing an adverse context for appropriate heath-care deivery. 5 Ongoing support from a cinician or team with reevant expertise was nevertheess important, depending on how this was deivered. Where the faciitator or ead cinician was abe to be present and responsive to the needs of particuar residents as they arose, and to engage staff in action earning that focused on issues of interest to them, there were higher eves of staff engagement and fideity with training 26,104,137,140,141 than in interventions where the cinician input was episodic or task focused. 116,120,137,142 The mechanism identified here was one in which the HCP worked with staff as the bridge to connect between interventions to improve heath care of residents over time in a way that coud be incorporated into existing patterns of working. Discussion The reaist synthesis in phase 1 has identified recurring themes and emergent patterns or demi-reguarities that underine the importance of how HCPs introduce and provide heath-care support to care homes. The way in which they work with care home staff, residents and their famiies, and the duration of this reationa working, appear to be important, regardess of the specific heath issue targeted. Broad mechanisms within a programme that can hep deiver appropriate heath care to care home residents are those activities that ensure that an intervention is specific to the care home, aigning with the goas and priorities of care home staff. They shoud not be adapted from other care settings and patient groups, but from the outset they shoud focus on activities that aim to buid reationships between care home staff and visiting HCPs. Contextua eements that shape the achievement of these outcomes and hep to sustain participation have been identified as: care home readiness to work with heath-care staff (e.g. care home eadership and previous history of coaboration) avaiabiity of structured assessment and care pans invovement of a HCP to support change and reinforce earning organisationa endorsement financia remuneration staff incentives. This is consistent with what is known generay about integrated working. 143 From the evidence reviewed the reevance and usefuness of the heath-care interventions and, utimatey, their impact were diminished in situations in which there was either itte evidence of prior coaborations or faiure to engage in a period of exporation and preparation that coud shape how HCPs 52 NIHR Journas Library

89 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 and care home staff coud work together. It highights the eves (structura, service and persona) at which care homes and the NHS have to work together to achieve the desired outcomes. Interventions ater context, that is, they attempt to change the care home environment so that the correct mechanisms are triggered to generate the desired outcome. Interventions do not in themseves have causa powers. When care home and visiting NHS staff beieved that they were working together and that there was common ground, change was more ikey to occur. An exampe of this was interventions that introduced new knowedge by inking it with existing ways of working, using care panning and ongoing conversations to find a way to reconcie the innovation with competing priorities in the care home. This was achieved by connecting new knowedge with existing practice and knowedge, using processes such as care panning and ongoing conversations to reconcie competing priorities in the care home. These findings resonate with internationa studies on the impementation of evidence-based care in residentia care faciities and on working with care home staff to improve residents we-being. 144,145 A review on the use of advanced care panning that incuded care homes 146 argued that no amount of faciitation or structured toos is sufficient to reduce the effects of those things that undermine them. Interventions that were feasibe to be deivered within time-pressured environments, the mechanisms of which support diaogue, experimentation and coaboration, and aow the system to evove and seforganise over time, were most ikey to be effective. Financia incentives or sanctions, agreed protocos, continuity of contact and evidence-based approaches to assessment and care panning provided the necessary equipment or resources to enabe those mechanisms to achieve improved resident and staff outcomes. Concusion We drew together a disparate iterature on care home residents access to heath care. The interpretation of the possibe CMO configurations was constrained by the ack of detai of the processes at work in the various interventions and by studies focus on staff satisfaction and confidence, rather than resident priorities, observed changes in practice or measurabe changes in resident outcomes. Previous review and survey work has demonstrated the compexity of the setting, the paucity of evidence and the shortcomings and inadequacies of either care home providers or heath-care providers. 4,5 In reaist terms, even when the desired outcomes are not achieved there is an opportunity to earn from the evidence and deveop a theoretica understanding of what needs to be in pace. This conceptua mode for further deveopment in phase 2 proposes that interventions (regardess of their use of sanctions and incentives, speciaist practitioners or care home-specific resources) are more ikey to achieve the outcomes of interest when they trigger the engagement of care home staff from the outset and create opportunities for heath-care and care home staff to work together and structure the intervention to fit with the priorities and working practices of the care home. This principe became the basis for the identification and seection of the case study sites in phase 2 and further refinement of the proposed programme theory of how and why NHS services work with care homes to achieve the five outcomes of interest. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 53

90

91 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Chapter 5 Phase 2 case studies: comparative description of the study sites Introduction Together, Chapters 5, 6 and 7 provide the basis for the deveopment of an expanatory account of how NHS support to care homes works, for whom, in what circumstances and with what outcomes. This addresses research questions 3 5: How are the features/mechanisms (as identified in phase 1 of the research) associated with key outcomes, incuding medication use; use of out-of-hours services; resident, carer and staff satisfaction; unpanned hospita admissions (incuding A&E); and ength of hospita stay? How are these features/mechanisms associated with costs of care from a NHS perspective? What configuration of these features/mechanisms woud be recommended to promote continuity of care at a reasonabe cost for oder peope resident in care homes? Chapter 5 summarises recruitment of care homes to the study and provides a detaied description of the phase 2 case studies on a site-by-site basis. It provides a within- and cross-case narrative of NHS provision to care homes that focuses on the different activities and responses that reated to the areas of interest identified in phase 1, namey how the services were provided and received, and how they were thought to work. This provides the necessary detai to be abe to understand how service modes infuenced the outcomes of interest, which is discussed in greater detai in Chapter 6. Detais of the protoco are pubished esewhere. 56 The chapter is divided into two sections. The first section describes the case study sites, recruitment and participant detais. The second section describes the services provided at each study site in greater detai. Case study sites, recruitment and participant detais Study sites 1, 2 and 3 covered three geographicay discrete areas in Engand comprising an inner area of a major city, a suburban area and a coasta area, respectivey (Tabe 5). Sites 1 and 2 were each ocated within a singe CCG area, but in site 3 recruitment was based on a county and around care homes engagement with the MyHomeLife eadership and management programme (see Chapter 2). TABLE 5 Study site characteristics Characteristic Site Popuation Life expectancy (years): femae (Engand, 82.8) Aged > 85 years, % (Engand, 2.3) Dementia prevaence (Engand, 4.27) Nursing home patients, % (0.5) Number of care homes Number of GP practices 1 342, , a 885, Source: Pubic Heath Engand. 147 Contains pubic sector information icensed under the Open Government Licence v3.0. a Site 3 represents data from across three CCGs encompassed within the sampe; the number of care homes comes from those enroed with the MyHomeLife programme. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 55

92 PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES Seventy-two care homes were contacted across the three sites. Care home recruitment is summarised in Tabe 5. Care homes in site 1 were recruited first, in accordance with the purposive samping framework outined in Chapter 2. Foowing the confirmation of care homes in site 1, care homes with simiar characteristics in terms of size, ownership and registration were identified in the other two sites. Care homes in site 2 were eigibe if they were incuded in the site s GP care home scheme that reimbursed specific practices to work with designated care homes. Care homes in site 3 were eigibe if they participated in the oca MyHomeLife network. For each site, the e-mai or etter invitation was foowed by a teephone ca and then a face-to-face meeting. After this, managers coud express an interest in participating in the study. This process is outined in Tabe 6 whie the factors that faciitated and inhibited care home recruitment are given in Tabe 7. At a three sites the managers who had confirmed their intention to participate signed an agreement outining their commitment to the study and their understanding of what taking part invoved, and nominated two staff members to take on the roe of care home ink staff members with the OPTIMAL study. The characteristics of the individua care homes and a summary of the services accessed by a the residents in the care homes at baseine are summarised in Tabes 8 and 9. Changes in service provision and care home-specific changes across the study The study took pace against a recent backdrop of the reorganisation of primary care trusts into CCGs. In sites 1 and 2, the majority of NHS services were provided by a singe trust, apart from menta heath services. In site 3, most services were provided by a singe trust, apart from menta heath services and dietetics, even though the services were commissioned through three separate CCGs. In site 2, compex care payments were beginning to be introduced for particuar residents with mutipe comorbidities inked to staff having competed a specific training programme. This had not been impemented in the study care homes by the cose of the study. TABLE 6 Care home recruitment process Site Number of care homes contacted Number of face-to-face meetings with managers Number of care homes interested Tota number of care homes recruited Time to recruit care homes (months) TABLE 7 Factors that faciitated, and inhibited, the care home recruitment process Factors Faciitating care home recruitment Inhibiting care home recruitment Care home having pre-existing invovement with the MyHomeLife project Previous invovement in research, for exampe through the ENRICH Network Cinician known to care home to introduce the study University being in cose proximity to care home No previous reationship with care home cod caing E-mai invitations being identified as spam by the care home server Lengthy permissions process for arge care home providers up to 6 weeks Staff workoads made research participation ow priority Staff turnover 56 NIHR Journas Library

93 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 TABLE 8 Care home characteristics at baseine Care home characteristic Care home ID Short ID Description Number of rooms, n Type of beds Specia registration Site 1, care home 1 S1CH1 City; privatey owned; converted house; singe care home provider Singe rooms, 19 Residentia Od age and dementia Site 1, care home 2 S1CH2 City; not for profit; purpose-buit two units across two foors Singe rooms, 62 Residentia Dementia, od age and physica disabiity Site 1, care home 3 S1CH3 City; privatey owned; purpose buit Singe rooms, 77 Dua registered; 47 residentia; 30 nursing Od age and dementia Site 1, care home 4 S1CH4 City; privatey owned; purpose buit Singe rooms, 40 Residentia Od age and dementia Site 2, care home 1 S2CH1 Town; privatey owned; converted house Singe rooms, 19 Residentia Od age and dementia Site 2, care home 2 S2CH2 Town; not for profit; purpose buit on two foors, spit into separate units Singe rooms, 51 Residentia Od age and dementia Site 2, care home 3 S2CH3 Town; not for profit; purpose buit on two foors and spit into five different units Singe rooms, 60 Residentia Od age, dementia and physica disabiity Site 2, care home 4 S2CH4 Town; not for profit; modern, purpose buit on two foors spit into five separate units, three of which took part in the study Singe rooms, 93 Nursing Od age, dementia and physica disabiity Site 3, care home 1 S3CH1 Town; for profit; arge corporate provider; modern, purpose buit on two foors spit into two units Singe beds, 50 Residentia Od age and dementia Site 3, care home 2 S3CH2 Town; for profit; sma chain provider; converted house Rooms, 20 (18 singe and two shared) Residentia Od age and dementia Site 3, care home 3 S3CH3 Town; for profit; sma chain provider; converted house Rooms, 34 Residentia Od age and dementia Site 3, care home 4 S3CH4 Town; for profit; privatey owned; converted house Rooms, 51 (50 singe and one shared) Dua registered, nursing and residentia Od age, dementia, physica disabiity, menta heath and sensory impairment Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 57

94 58 NIHR Journas Library TABLE 9 Heath-care services accessed by care homes at baseine Care home S1CH S1CH S1CH S1CH S2CH S2CH S2CH Service DN CHNP SW OOH DOT PT OT Fas team TVN CPN CNS DNS PCN PDNS ACNS Geriatrician Dietitian Chiropody Optician Dentist Psychiatrist Psychoogist Audioogy SALT Pheb a ++ a S2CH S3CH a a a +++ a a PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES S3CH S3CH S3CH4 +/ a , accessed rarey; ++, accessed often; +++, accessed frequenty; +/, used ony by those in residentia home beds; ACNS, anticoaguant nurse speciaist; CHNP, care home nurse practitioner; CNS, continence nurse speciaist; DN, district nurse; DNS, diabetic nurse speciaist; DOT, Dementia Outreach Service/Community Menta Heath Team; OOH, out of hours; OT, occupationa therapist; PCN, paiative care nurse; PDNS, Parkinson s disease nurse speciaist; pheb, phebotomy; PT, physiotherapist; SALT, speech and anguage therapist; SW, socia work; TVN, tissue viabiity nurse speciaist. a Private fee-for-service arrangement.

95 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 There were some changes in how services were structured over the course of the case studies. In site 2, personaised care pans were introduced by GPs for residents in three of the care homes. In site 3, there was a 6-month piot study for a sma number of GP practices to hod monthy cinics in care homes to review residents heath and medication that invoved care homes 2 and 3. The other care home in site 3 witnessed an increased tendency for speciaist nurses to substitute for GP visits over the course of the study as ocums repaced reguar GPs who had retired from the practice, or who were on sick or maternity eave. Domiciiary denta services were withdrawn from two care homes in both sites 2 and 3 and, subsequenty, repaced with another (domiciiary) service in site 2, but not in care homes 2 and 3 in site 3. District nursing teams were reorganised for care homes in sites 2 and 3. One care home that was dua registered at the start of the study became a nursing home. This change of registration was used as the basis of a decision that district nurses woud no onger visit the remaining residentia care residents. These changes are summarised in Tabe 10 aongside a summary of care home staffing changes over the period of the study. These findings are compatibe with previous studies describing high eves of staff turnover in the care home sector. One consequence was difficuty in maintaining the study ink roe. We had reimbursed this at market rates, emphasising the importance of continuity; however, staff were frequenty required to prioritise care deivery over research. Resident recruitment and retention Identifying care homes and recruiting eigibe residents to participate was a protracted process. Recruitment commenced in site 1, foowed by site 3 and then site 2. Site 1 took 7 months to recruit 93 residents, site 2 took 8 months to recruit 92 residents and site 3 took 5 months to recruit 57 residents. Site 2 was the ast to be incuded to aow sufficient time for a newy estabished GP care home scheme to be embedded. After excusions, 472 residents were eigibe for recruitment across the sites; the overa recruitment rate was 55% for site 1, 52% for site 2 and 46% for site 3 (see Tabes 11 13). In sites 1 and 3, where the care homes with nursing beds were based in homes dua registered for persona care, the percentages of residents recruited who were in nursing beds were comparabe at 13% and 14%, respectivey. In site 2, where the fourth care home was nursing ony, the proportion of residents in nursing beds was higher at 39%. In tota, 242 residents were recruited across the three sites; 93 in site 1, 92 in site 2 and 57 in site 3. Thirty per cent of the sampe (n = 73) was made up of those residents with capacity to consent and these residents were recruited using standard consent procedures without the need to refer to consutees. The majority (n = 169) of residents were recruited via a consutee process, as outined in Chapter 2. Nominated consutees, who were independent to the study, were appointed to act on behaf of eight residents. In site 3, it was not possibe to identify a nominated consutee. Tabe 11 gives the breakdown of recruitment numbers incuding number of beds, those excuded and those with and without capacity to consent. Athough the financia incentives paid to the care homes to compensate for the staff time invoved in the study appeared to faciitate the recruitment of care homes at an institutiona eve, it did not appear to have an impact on resident recruitment and data coection. Apart from at one care home, funding was not used to empoy extra staff to cover study invovement. Monies received by care homes were instead put aside for a variety of panned projects incuding the purchase of specia adjustabe beds for use by one or two of the oder residents in the care home, a minibus for taking residents on outings, a sensory garden, the refurbishment of a sitting room for residents with dementia and staff training on wound management. Baseine interrai data were coected for 234 residents, with medication data for 228 residents and baseine service use for 235 residents (see Tabe 13). Monthy heath service use, medication changes and reviews were coected for 12 months post baseine for 11 out of the 12 care homes. A deay between the 12th care home agreeing to take part and their avaiabiity to commence recruitment meant that monthy resident data were coected ony for 9 months post baseine. Across the three sites, 85 residents (35%) were ost to the study, predominanty through death, athough three residents were transferred to other Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 59

96 PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES TABLE 10 Summary of NHS and care home changes identified over the study period Changes in the organisation of care to and within care homes Site Changes in NHS provision to care homes An eectronic prescription process was introduced by the GP practices inked with two of the care homes. One of these homes had three different GP practices. This resuted in variabe usage of paper and eectronic prescriptions, depending on which practice the resident was registered with GPs introduced personaised care pans for care home residents in care homes 3 and 4 In care home 1, GP ocums were reported as not wanting to visit care homes. The ocums had repaced a retired GP and one on sick eave Monthy GP cinic piot was introduced for 6 months in care homes 2 and 3 Domiciiary dentist ceased visiting care homes 2 and 3 with no repacement The community dentist who visited residents in care home 4 retired and was repaced by a temporary ocum Reorganisation of district nursing teams for care homes 1 and 3 integrated with therapists New DN teams inked with care home 2 Change in the district nursing provision to care homes 3 and 4 resuting from the geographica extension of the Care Homes Team to the oca area of these homes. Existing DNs visiting homes 3 and 4 were repaced by the Care Homes Nursing Team New domiciiary denta service visiting care homes 1 and 3 To reduce needs for SALT team visits, care home staff trained for swaowing assessments Care home changes Three consecutive managers from one of the study care homes and four study ink staff from two of the study care homes eft their empoyment with these care homes. Two further study ink staff from a third care home, whie remaining in their existing posts within the same care home, withdrew from their study ink staff roe as a resut of the continuing demands of their care home workoads Five ink staff members eft in two different care homes, incuding two managers in one care home. One care home cosed a unit to refurbish it for oder peope with dementia Four care home ink staff, in three different care homes, eft before the end of data coection and one became deputy manager but remained as the dedicated staff member for the study One care home was sod out to a different independent arge chain corporate care home provider. The care home manager remained in post in this care home but was assigned additiona responsibiity for another care home owned by the same company DN, district nurse; SALT, speech and anguage therapist. 60 NIHR Journas Library

97 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 TABLE 11 Resident recruitment figures, excusions and those with/without capacity to consent Care home Number of beds Number of residents Excuded Number after excusions With capacity: agreed to participate Letters sent out Consutee (yes) Consutee (no) Tota (yes) Recruited after excusions (%) Site 1 S1CH S1CH S1CH3 47 (R) 45 (R) (R) (N) 28 (N) 12 (N) 77 (tota) 73 (tota) 30 (tota) S1CH Totas Site 2 S2CH S2CH S2CH S2CH4 93 (N) (N) 49 Totas continued Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 61

98 62 NIHR Journas Library TABLE 11 Resident recruitment figures, excusions and those with/without capacity to consent (continued) Care home Site 3 Number of beds Number of residents Excuded Number after excusions With capacity: agreed to participate Letters sent out Consutee (yes) Consutee (no) Tota (yes) S3CH S3CH S3CH S3CH4 14 (R) 14 (R) (R) (N) 30 (N) 8 (N) 53 (tota) 44 (tota) 13 (tota) Totas A sites Totas N, nursing; R, residentia. Recruited after excusions (%) PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES

99 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 care settings (Tabe 12). A compete set of service use data was coected for a residents in site 3, but in sites 1 and 2 this was incompete. Fourteen residents, two in site 1 and 12 in site 2, had missing service use data equivaent to 36 months of service use (Tabe 13). This was because the information coud not be ocated in archived notes as a resut of the variabe archiving poicies across the 12 homes, which had not been accounted for in the study design. Recruitment of interview subjects In tota, 181 individuas, incuding residents, reatives, HCPs (such as community nurses, aied heath professionas and GPs), care home staff and managers took part in individua interviews or focus groups (Tabe 14). Across the sites, 116 interviews took pace. Eeven focus groups were conducted with care home staff in 11 care homes, one with GPs in sites 2 and 3 and with commissioners (incuding GP commissioners) in sites 2 and 3. Residents with capacity to consent to interview were more difficut to recruit in site 2, where a significanty greater proportion of care home residents had cognitive impairment and acked capacity. Thirty-five resident interviews were competed in tota, but one interview was not used as the resident was unabe to focus on questions about heath and service use and no meaningfu data reevant to the study were obtained (see Tabe 14). TABLE 12 Resident retention and oss to foow-up Site Retention and oss to foow-up Care home Care home Care home Tota, Tota, n (%) n (%) Tota, n (%) A, n (%) Retention Baseine tota number of residents recruited Number of residents retained at 3 months Number of residents retained at 6 months Number of residents retained at 9 months Number of residents retained at 12 months (90) (88) (95) 219 (90) (83) (84) (86) 203 (84) (72) (70) (72) 172 (71) (62) a 37 (40) (63) 131 (54) Reasons for oss to foow-up Tota number of residents ost to foow-up at 12 months Number of residents who moved care faciity (37) (47) (37) 98 (40) (2) (1) (1) Number of residents who are deceased (34) (34) (37) 84 (35) a Two hundred and three participants were retained in the study at 6 months, but some of these had missing data points. Therefore, 195 had compete data to 6 months and were incuded in cost anaysis (see Tabe 19). Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 63

100 64 NIHR Journas Library TABLE 13 Resident data coected over 12 months from baseine incuding interrai, medication and service use Number of Number of residents Number of Number of Number of residents: month of service use Site residents recruited deceased or eft the care home interrai data coected MAR sheets coected baseine service use Tota PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES

101 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 TABLE 14 Participants incuded in interviews and focus groups broken down by site and group Site Participants Tota HCPs Residents Reatives Care home staff Care home managers GPs Stakehoders Tota In addition to the 18 interviews conducted with famiy members, four famiy members were present during resident interviews and in each case presented data additiona to those provided by the resident, adding to the interview transcript. Reatives were harder to recruit in site 2, where a significant number of residents had no identified next of kin, and site 3, where the care home staff reported that it was difficut to engage reatives in care home-reated activities. To counteract this the researcher attended residents and reatives meetings where possibe in site 2 and set up evening meetings for reatives in site 3 care homes to inform them about the study and invite them to take part in an interview. Despite this, no reatives from the meetings in site 2 expressed an interest in taking part and no famiy members attended the evening meetings in site 3. The care home staff represented a range of experience and roes incuding care provision, catering and training for care home staff. Staff were interviewed in the care home. Fewer HCPs were interviewed in site 2, which refected the fact that care home provision at this site came from a reativey sma poo of staff from the primary care organisation thus there was significant overap between the team members supporting individua homes. The breakdown of HCPs interviewed is summarised in Tabe 15. Genera practitioner interviews were the most chaenging to arrange and were not competed unti after care home data coection had stopped. Interviews and focus group discussions were time constrained and arranged either to coincide with other meetings at the CCG or as individua interviews competed at the end of the study. On two occasions, in two sites, agreement to participate was withdrawn. One GP TABLE 15 Interviews conducted with HCPs, incuding community nurses and aied heath professionas Site Community nurses Aied HCPs Tota 1 9 (incuding care home, dementia, Parkinson s disease, continence, fas and DNs) 2 6 (incuding district, paiative care, tissue viabiity, menta heath and continuing heath-care nurses) 3 9 (incuding care home, district, dementia, heart faiure and tissue viabiity nurses) 9 (incuding dentist, optician, chiropodist, fas, rehabiitation and dementia OTs, fas team, dementia and rehabiitation physiotherapists) 3 (incuding chiropodist, optician and a therapist assistant) 7 (incuding dietitian, OT, rehabiitation and stroke physiotherapists, dentist, optician and chiropodist DN, district nurse; OT, occupationa therapist. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 65

102 PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES decined an interview but did compete a written questionnaire comprising the interview prompts about how they worked with care homes. Consequenty, the focus of these interviews was not excusivey on the study care homes. They did, however, a consider care taking pace within GP practices aigned to the CCGs incuded in the case studies. The breakdown of GP research contact is provided in Tabe 16. Resident characteristics at baseine Resident characteristics by care home and site are summarised in Appendix 10. Gender and the age of the cohort were simiar between sites and were broady representative of the UK care home popuation. 148 Comorbidities were imited to those that coud be retrieved from the care home notes and are significanty ower than the numbers seen in other arge care home studies where medica records were used to count comorbidities. They serve here to act as a comparator ooking for baseine confounding between sites where no significant difference was seen, rather than to represent a definitive picture of comorbidities across the cohort. A significanty greater number of sef-funders were seen in site 2 than in site 1 (p < 0.05). Despite efforts to cosey match the care home popuations, a significanty higher number of nursing home residents were recruited in site 2 than in sites 1 or 3 (p < 0.01). Site 3 reported significanty ower eves of functiona dependency, communication difficuties and cognitive impairment, as we as significanty ower scores for the interrai summary scores for ADL, cognitive performance, communication (p < 0.01 for a) and pressure ucer risk (p = 0.05). Site 3 reported significanty ower eves of functiona dependency, communication difficuties and cognitive impairment, as we as significanty ower interrai summary scores for ADL, cognitive performance, communication (p < 0.01 for a) and pressure ucer risk (p = 0.05) we were not aware of any particuar recruitment bias that ed to this. Services provided at each study site in detai Heath-care professionas accounts of working with care homes and other HCPs were used to provide a description of service provision in each site and how it was structured. Specific attention was paid to how the different characteristics of the service modes in each site were recognised and understood by participants. Across a three sites, six aspects of the services were identified as being key to understanding how care was organised and operationaised: 1. the system for referras 2. avaiabiity of dedicated heath services for care homes 3. team working 4. the use of case management 5. care home-based training 6. ength of time HCPs had worked with care homes and each other. Engagement with care homes for each of the three sites is considered under the headings beow. TABLE 16 Summary of GP research contact across the 3 sites Sites Number of transcripts Number of GP participants 1 1 competed questionnaire 5 4 interviews 2 1 focus group 11 1 interview 3 1 focus group 5 (pus one practice manager) 66 NIHR Journas Library

103 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Site 1: age-appropriate care In site 1, HCPs pattern of engagement and working with care homes was characterised by: mutipe MDTs that either worked excusivey with care homes or had expicit responsibiity for care homes as part of their work (dementia care, fractures and fas prevention) a nurse-ed care home service that had been in pace in different iterations for more than 15 years. This incuded proactive approaches to working with care homes incuding case management of new and continuing heath-care-funded residents. forma and informa systems for team-to-team referras about specific residents, incuding access to speciaist dementia knowedge a history of innovations around working with care homes with evidence of consutations with care home managers about heath-care priorities and the need for support and training ongoing structured training and forums to support both care home staff and managers. Figure 3 gives a detaied outine of the way that services were structured in site 1 and interactions between them. Services that were care home specific and those that had significant contact with care homes or provided peer support or training for care homes are highighted in green text. The figure is drawn from 18 interviews with HCPs. It may not capture the fu picture of working reationships and practices, but it does refect practitioner understanding of what was avaiabe and how the system worked. Referra systems Most referras on behaf of residents by care home staff and HCPs to other services were made through a singe hub for both heath and socia care by teephone or eectronicay through a shared system between the services working in care homes. GP practices coud choose to use or bypass the hub. Referras to the optician, chiropodist and dentist were usuay arranged separatey by care home staff. According to the HCPs, this system worked we. The hub referra system was intended to simpify the decision for care home staff as to whom they shoud make a referra. However, in practice, this was not aways the case and there was imited evidence of triaging by resident need. In some instances care home staff had referred a resident to more than one team, generating dupicate visits and assessments, because they found knowing who to refer to uncear. HCPs from the same discipine were frequenty part of mutipe different speciaist teams that either provided input, or potentiay coud provide input, to the same care home. For exampe, if a resident had a fa, they coud be referred to a physiotherapist from either the fas team or the rehabiitation team and it was not aways cear to professionas from outside these teams which was most appropriate. This confusion was not confined to care homes but was aso found with other HCPs who were not part of the teams with care home responsibiities, incuding the GPs. Referras made by GPs that bypassed the hub coud aso cause dupication. To offset this, some teams introduced an informa system of checking with each other when picking up a referra, to ensure that the care was not being dupicated. An informa network of communication deveoped over time heped to mitigate these probems. Team working Heath-care provision in site 1 was characterised by numerous MDTs each with a specific focus of care and speciaist knowedge (e.g. fas prevention) who iaised and worked cosey with each other. These teams carried out joint visits to care homes and frequenty made referras both within their own team and across teams. Initia assessments of residents heath needs were reported to be hoistic in order to highight which aspects of care might require GP invovement and which might mandate speciaist referra. One exampe described a situation in which a resident with dementia was seen at a fracture cinic but then foowed up in situ by two physiotherapists, one each from the dementia and rehabiitation teams, who worked together to ensure the resident was assessed in the care home. This took the pace of routine outpatient fracture cinic physiotherapy foow-up and aowed access to specific expertise in both muscuoskeeta rehabiitation and modification of rehabiitation regimes to take account of cognitive impairment. Figure 3 iustrates a compex network in which there coud be mutipe teams working with a resident, each with rapid access to speciaist Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 67

104 PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES Referras through the heath and socia care hub and SystmOne GP can bypass it Care co-ordinator (one per cinica deivery group) Can make referras for HCPs Care home team 6 staff in one team visit 36 care homes (new service) Team composition Care home speciaist nurses Care home district nurses Receive referras from GP hospitas other teams direct care home referras Care home speciaist nurses Refer to dementia team fas team rehabiitation team and others New residents assessed by the speciaist nurses on caseoad for up to 8 weeks In consutation with care home manager can arrange MDT meetings with GPs and/or speciaist teams (socia workers and others) Dedicated staff member for care home staff training on insuin, catheter, end-of-ife and pressure area care Weeky meeting in the care home with the geriatrician Overview of site 1 (n = 18) Most care homes work with 1 GP ony Care home-specific teams, n = 2 MDT with significant eve of care home input (incuding training for care home staff), n = 1 1 MDT; nurse speciaists with reguar input, n = 2 AHPs with reguar ongoing input, n = 3 Quaity forum attended by representatives from teams that go into care homes to discuss incidents and safeguarding Dementia team (estabished 15 years ago) NHS Team composition Dementia nurse Physiotherapist (speciaist training) OT Support workers Consutant time cose inks Receive referras from GP ony or via other HCPs: care home (with HCP authorisation) community menta heath team nursing home team continuing care team OT, physiotherapist socia services; SALT team; fas team; DNs Refer to wi ask care home to refer to HCPs if required, for exampe fas and bone nurse community OT community physiotherapist community care team nursing home team Dementia nurses case manage residents funded by CHC Run a care home managers forum every 3 months Run an activity co-ordinators forum every 3 months Fas team (estabished 15 years ago 25 in team) High care home input Team composition Cinica speciaist Fas team nurses Fas team OT Physiotherapist Administrator (books and co-ordinates appointments) Assistant practitioners (band 4) Rehabiitation support workers (band 3) Heath promotion speciaist Receive referras from emergency department (injurious fas) GP cinics, care home team Direct care home referras Refer to dietitian, optician, tissue viabiity, stroke team Heath promotion speciaist; this is a unique educationa roe and ead roe for dementia and safeguarding Fas and bone heath training for a care homes on roing basis giving them skis to manage fas Loca enhanced practice (each care home works with ony one GP practice) Community rehabiitation team (estabished at east 7 years ago) Physiotherapists, OT, assistant practitioners GP Monthy cinics in care home Aso in between if required GP practice is remunerated for this On average sees 5 9 residents Speaks to care home manager when visiting Continence nurse speciaist Visits on request Parkinson s disease nurse speciaist Visits on request Dentist Arranges ongoing bookings with care home Visits once every 4 months Optometrist NHS and private Arranges ongoing bookings with care home Visits care home twice a year Chiropodist (private) Visits residents every 8 weeks ongoing arrangement Does not do foot care for peope who have diabetes or MRSA; these are seen by a NHS podiatrist Does not receive or make referras Urgent care team (response time 24 hours); short term ony, aim to prevent hospitaisation Does not cover care homes Other teams (referred to ony) Heath reabement team Community stroke team Neuroogica team Information/comunication A teams can access SystmOne FIGURE 3 Overview of heath services provision in site 1. Green font indicates care home-specific services, direct referra, high care home input or organised training. Circuar boxes indicate teams, square boxes indicate individua HCPs. AHP, aied heath professiona; CHC, continuing heath care; DN, district nurse; MRSA, meticiin-resistant Staphyococcus aureus; OT, occupationa therapist; SALT, speech and anguage therapist. 68 NIHR Journas Library

105 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 services. This coud ead to dupication of work and confusion for care home staff, especiay if they are given contradictory advice regarding an individua resident s care. A HCP described addressing this probem as foows:... I think one of the other difficuties they [care home staff] have is often peope [residents] end up having mutipe teams invoved, often giving sighty contradictory advice, and they re ike We what are we supposed to do? You re teing me one thing, someone is teing me another thing and you can see they re just in the midde going We what am I, whatever I do, someone s going to be annoyed with me. What we aways do in the Fas Team is we aways make sure we write in the professiona s bit of the records and then we send a recommendations etter saying This is what we ve recommended, so if another professiona from another team ooks in the care pan they can see what we ve done... S1CH2HP01 Dedicated heath services for care homes Site 1 was the ony site with HCP teams funded to work excusivey with care homes providing direct care for residents. Two teams worked specificay with care homes, incuding a nurse-ed care home team and a dementia care home-specific team. In addition, a third team had substantia input into care homes mainy for staff training and support around fas prevention in addition to working with oder peope iving at home. The nurse-ed care home team incuded district nurses and community nurses who provided on-demand nursing support to residents without on-site nursing provision and speciaist nurses whose roe was to assess a new admissions to a care homes and support the transition into the care home. They woud then oversee care of residents (case management) during the first 8 weeks in the care home. This team met weeky with a consutant community geriatrician to review residents and discuss care panning. The dementia outreach team incuded physiotherapists, occupationa therapists and menta heath nurses. They provided assessments of the care needs of individua residents iving with dementia and support to care home staff. Their work was not defined excusivey as heath and socia care. For exampe, support workers within the team had the remit to focus on residents engagement with dementia-reated activities in the care home, such as reminiscence techniques. Case management Site 1 was the ony site that had case management for new admissions to care homes. This was ed by speciaist nurses in the care home team, who worked excusivey with the care homes and thus had more time for taking to and supporting care home staff than their district nurse counterparts. Their presence meant that they were aso abe to respond to staff queries about other patients without needing a schedued appointment. The speciaist nurses from the dementia team had a case co-ordinator roe for those residents in receipt of NHS continuing heath-care funding. These residents stayed on their caseoad, which aso faciitated reguar contact and reationship buiding with the care home staff. Care home training and monitoring A programme of panned training was avaiabe to care home staff in site 1. This was deivered by staff members of the dementia, fas and NHS care home nursing teams and targeted a range of staff from care home managers to activities organisers. This was suppemented by training that was avaiabe on request and in response to care home priorities from the tissue viabiity, diabetes meitus and nutrition teams. The dementia team ran a programme of care home forums with meetings every 3 months that aimed to provide a programme of training whie aso providing an opportunity for more informa interaction, peer support and reationship-buiding. Safeguarding and reviews of the quaity of care provided in care homes were monitored and discussed at specific quaity meetings attended by the NHS service eads. The approach taken was to identify homes with particuar probems and to offer support to address any identified probems. Quaity meetings Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 69

106 PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES augmented the teaching and were seen as a vehice to highight and share good practice in care homes for future earning and practice. They were not, however, attended by representatives from the care home sector and so communicating constructive outputs from this forum were contingent upon members iaising with the care homes afterwards: What we tend to do is, apart from the incidents that are aready recorded, that have aready been safeguarded or dated, we aso have... areas of good practice that we want to share. So, homes... that we have identified as improving or have shown, you know, reay good ideas around that improved practice... that are good. So you know, praise where it s due, I aways say, because I think they get bashed so much, it s about saying, Actuay this is reay good, can we share that good practice? How have you done that? You know, can we sort of use that as a mode and you know, pass that on, and pass that knowedge on to other homes? S1CH1HP06 Prior history of working with care homes From the interview transcripts it was evident that some of the individua HCPs had ong-standing working reationships with care homes, which had a positive impact on their abiity to engage with each other. The care home teams in this site were buit on a history of projects conducted with care homes. An earier piot had provided a sma group of care homes with training from speciaist nurses, which focused on six core areas: tissue viabiity, nutrition, hydration, medicine management, end-of-ife care and continence. There was evidence that as piot projects had drawn to a cose, the NHS staff invoved had tended to move into empoyment upon the next wave of care home initiatives being deveoped within the area, thus aowing and faciitating reationship-buiding over time. Continuity of reationships were maintained over a number of years, even as Nationa Government Services projects stopped, started and were reconfigured. Site 2: incentives, sanctions and targets In site 2, HCP working with care homes was characterised by: a focus on the GP as the organiser and partner in the co-ordination of services provided to the care homes specific GP practices receiving extra payments to work with care homes; care homes were asked to register with one of these practices a nurse speciaist in paiative care who had a designated roe to work with care homes structured training for care homes commissioned by the CCG and provided by a training organisation; a focus on knowedge and skis needed to support peope with compex needs to stay in the care home and avoid hospitaisation; competion of training meant that the care home was eigibe to receive additiona payments for residents recognised as needing extra care and support. In addition, the foowing services were avaiabe to, and accessed by, care homes, but were not specificay targeted at the care home sector, nor did they have care homes as a priority within their service specification. A newy estabished, integrated team of nurses and therapists that provided wrap-around care to oder peope iving at home, which incuded care homes. Dementia-specific advice and support provided via menta heath services to oder peope iving at home, which incuded care homes. This ed to the foowing. Services were focused around individua residents, with few opportunities to meet with managers and care home staff about more generic issues affecting care across the resident cohort as a whoe. Care home staff and HCPs were uncertain about who to approach about specific probems and HCPs were simiary uncertain about how to signpost them. The service mode at site 2 is summarised in Figure NIHR Journas Library

107 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Referras Majority through GP Singe point of access SystmOne Overview of site 2 (n = 9) Care homes work with 1 3 GP practices 1 speciaist practitioner dedicated to working with care homes ony on end-of-ife care 2 speciaist nurses, nursing and menta heath teams, 2 AHPs visit on request NHS Named GPs receive additiona payments to work with care homes (1-3 practices per care home) Community nursing team One team visits 6 care homes Technica instructor for OT/physiotherapist 6 trained nurses 3 HCAs staff in 1 team (integrated with the physiotherapists and OTs) Receive referras from GP hospitas socia worker direct care home referras for DNs but through GP for therapist apart from repacing equipment Externa referras tissue viabiity nurse speciaist GP continuing heath-care assessor socia work care managers Out-of-hours team covers nursing emergencies from 5 p.m. overnight Information/communication Access SystmOne Paiative care nurse speciaist Based in the hospice team Ony works with care homes Goes into 15 care homes in site 2 Active caseoad of 5 10 residents Reguar joint visits and meetings with other HCPs Works with hospice team Speciaist nurses OT, SALT Physiotherapist Counseing Consutant time Receive referras from hospita 20% direct referras from care homes 75% (incuding nursing homes) nurse speciaist (e.g. neuroogica MDT, heart faiure CNS, respiratory CNS) famiy can aso refer Refer to other members of the hospice team, ymphoedema service, rapid response hospice at home, GP, DN, tissue viabiity CNS, dietitian, continuing nurse assessor Reguar training for care homes on end-of-ife care. Topics requested by care home staff Information/communication Paper ight but uses different eectronic system so not possibe to share notes with other HCPs Tissue viabiity nurse speciaist Receive referras from GP DNs nursing home may refer direct Refer to GP for vascuar team DNs for residentia homes and nursing homes dietitian Provides training to the top 10 care home referrers Aso provides training for care homes through sessions for care providers association Intensive menta heath team Co-ocation with team for oder peope (menta heath) Covers a care homes in site 2 Team composition CPNs OTs Works cosey with menta heath team focusing on oder peope psychiatrist Receive referras from menta heath team GP hospita ward staff socia worker memory service Care homes have to refer via GP Refer to oder peope s team, OT, dietitian, SALT Information/communication Access SystmOne Continuing heath-care assessor Works for CCG and covers nursing homes Checks nursing needs are met and funding is appropriate Referras to other services wi be via the GP GP Weeky cinics in care homes Aso in between if required GP practice is remunerated for this Some care homes sti have residents registered with GPs that visit on request Optician NHS and private Arranges ongoing bookings with care home Visits care home once a year Personaised eye care reports in resident s notes with care pan for staff to refer to Chiropodist private Visits residents every 6 8 weeks; ongoing arrangement Any referras are made through the care home staff (e.g. DN) Other teams referred to ony Neuroogica team FIGURE 4 Overview of heath service provision in site 2. Green font indicates care home-specific services, direct referra, high care home input or organised training. Circuar boxes indicate teams, square boxes indicate individua HCPs. AHP, aied heath professiona; CNS, cinica nurse speciaist; DN, district nurse; HCA, heath-care assistant; OT, occupationa therapist; SALT, speech and anguage therapist. Referra systems Heath-care professionas in site 2 had access to a variety of referra routes, incuding a singe point of access and eectronic referras through a shared system, but there was an emphasis on GPs acting as co-ordinators and conduits for such referras. There were some differences across the teams; for exampe, for one district nurse team the care home coud make referras directy through the singe point of contact, whereas for the other, which covered two of the study care homes, a referras had to go through the GP. It was not cear, even to those invoved, whether this was because of forma differences in service specification or simpy a consequence of different working practices across GP surgeries. There was evidence that once a resident was being ooked after by one member of the integrated care team, members of that team woud make direct referras from one to another, for exampe between district nurse and physiotherapist, without having to go back through the GP. There were exampes of care home residents receiving joint assessment visits when, for exampe, some GPs woud visit care homes together with members of the menta heath team or Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 71

108 PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES when the paiative care nurse woud attend together with other members of the hospice team whose input she had identified as ikey to be important or hepfu. In contrast to site 1, there were no forma or informa referra systems organised around the care homes and their residents. As in site 1, referras from the optician and chiropodist were separatey organised. Dupication occurred when, because there were no shared records, care home staff were unaware that the district nurse had aready visited. In this site there were frequent references to inappropriate referras by care homes. As the foowing quotation demonstrates, inappropriate referras were underpinned by care home staff not knowing that they coud initiate conversations and few opportunities for meaningfu diaogue between care homes and visiting HCPs; what might be regarded as an appropriate referra or how staff might differentiate between urgent and non-urgent reasons for referra were not discussed: Interviewer: Anything ese you d ike to say about working with care homes, or setting up a new service to provide heath services to them? S2CH2HCP01:... a ot of the times we do see patients at maybe inappropriate referras, and maybe because of the ack of teaching and a ack of referra knowedge [of the care home staff] they refer when they shoudn t, and maybe they shoud just ask us a question before they refer to us. So ike if someone said, Oh, can I just ask you a question, just for a bit of advice? And then we say, Yeah, that s fine, refer that person, and then we can see them, rather than referring them to us when actuay there was nothing there [no reason to visit] in the first pace. So sometimes it can be a bit of time wasting on both sides, that s it [aughs]. Team working Speciaist teams payed a part in the organisation and provision of heath services in site 2. Athough they visited care homes, they aso had responsibiities for community-dweing oder peope. Their responsibiity for care home residents and how much time they shoud aow for work in these settings were not specified. A notabe exception, and one that was coser to the patterns of visiting reported in site 1, was the input of the paiative care nurse speciaist. She worked excusivey with care homes and had reguar meetings with a hospice team and a neuroogica speciaist team as we as other HCPs who visited the care homes as part of their broader responsibiities. Her key roes were to advise on the care of residents who were dying, iaise with speciaist nurses so that the homes coud access their services more easiy and provide ongoing support and training to care home staff. Her remit was care home wide and not imited to individua resident episodes of care: Interviewer: So, asking about the eve of contact you have with the care home staff when you go in, how the visit works, you know...? S2CH4HCP01, paiative care nurse speciaist: Yeah, so I woud say I aways have contact with the care home staff. Many patients do not have fu menta capacity so care home staff hep with assessing of situation, feedback and education is given to the care home staff on drug management, symptom contro, psychoogica and spiritua support. If a new member of staff is there, I try to encourage them to work aongside me so I m teaching them how to manage without me when having End of Life discussions, advanced care panning discussions and difficut conversations. Yeah, so I hope that this faciitates education of staff and heps continuity and coordination, the abiity of the staff to communicate we with famiies of patients and MDT is vita. Access to dementia speciaist support for care home residents was identified by severa respondents as a particuar issue and there were accounts of where the ack of access to speciaist support had ed to residents receiving poor care from NHS services and, in one exampe, poice invovement the staff who discussed this case fet that proactive engagement by a speciaist team with co-ordination of the services invoved coud have avoided this outcome. Dementia training (3 hours, internet based) had recenty been made mandatory for one district nurse team but not the other. Severa HCPs identified that they fet particuary uncertain in situations where they had to respond to residents whose behaviours they found chaenging. There were exampes of HCPs who showed in their anguage a tendency to depersonaise 72 NIHR Journas Library

109 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 peope iving with dementia, referring to care homes as a dementia and assuming that aggressive behaviour was an inevitabe consequence of the iness: Not so much the residents because in care home X, a the residents are dementia so but the thing is on the nursing side we can go in and because we know how they re [residents] speciaist going to react, i.e. there s a coupe that wi ca you everything under the sun, kick and bite you and everything, but I know how to work around them so I think that s why it s a reay good thing to have the same peope going in, especiay on the dementia side because you know how that patient is going to react. Some of them wi just sit there and put their arm out, another one wi ca you everything under the sun but you don t take offence to it because you know that they don t mean what they re saying, you re just going in to do your job and ook after those patients. So, but again, I do ask for somebody to be with me [care home staff member], especiay if I know that they re going to ash out anyway. S2CH3HCP01, community nursing team member Dedicated heath services for care homes A heath services in site 2 worked with care homes as part of a broader remit to care for oder peope in the community. When care home strategies had been deveoped, the focus was on GP provision and the use of enhanced payments to increase access and frequency of visiting. Care home training and monitoring Some of the district nurses thought that care home staff training shoud be a priority, especiay in reation to catheter care and pressure area care; however, none of them was invoved in deivering any such training. The majority of the HCPs, incuding the community and menta heath nurses, focused on what they saw as the inadequacy of training and their monitoring roe in highighting safeguarding and quaity issues. There was a perception that it was not their responsibiity to provide training: There shoud be better training I think for carers [care home staff], I mean because at the end of the day some of our referras, we aren t actuay required in a home... I fee they need more updating on pressure care for sure, definitey, and their competence when it comes to pressure area and care is poor, very poor. And aso their moving and handing skis are atrocious, I mean I ve had to report quite a few times the way I ve seen peope handed in a home and then the girs that I ve worked with, they ve been ike near to tears teing me that they go on courses and this is how they ve been shown and then when I ve gone to... the manager of the home, she said No, this isn t how we have shown them, they shoud know different. S2CH3HCP01 p. 7, community nurse district nursing team The structured training that was being provided (paid for by the CCG) during the time of data coection incuded some of the HCPs working with a socia care training provider and was inked to care homes receiving extra funding. Partitioning training off in this way coud possiby have had an unintended consequence of reinforcing other HCPs beiefs that supporting staff earning was not their responsibiity. Prior history of working with care homes The paiative care home nurse speciaist had a ong history of working with the care homes. She had been a district nurse before becoming a speciaist nurse and had buit reationships with care home staff when working in that capacity. In the ast 3 years in which she had been expicity responsibe for supporting care homes, she had worked with care home staff to ro out an onine training package in end-of-ife care. She interpreted her roe as providing expertise, support and advocacy for both residents and staff: Interviewer: What does your roe invove? S2CH4HP01: So to go into the care home after a referra s been sent and to support the care home staff to support their patient, to support the famiy or carers of that patient, to provide hoistic support for that patient and their famiy, to iaise and co-ordinate the other MDT services, provide speciaist drug advice. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 73

110 PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES The paiative care nurse was, however, the exception at this site. For the other HCPs interviewed they did not have this history of association or care home focus. The reason for visiting was to see particuar residents, which meant that their invovement with the care homes and contact with care home staff fuctuated according to the needs of individua residents and was not structured to promote engagement at the care home eve. In addition, care often focused around specific care episodes with intensive periods of invovement foowed by withdrawa as things stabiised. Sustained and continuous engagement, even with individua residents, was not a feature of this way of working. Site 3: Genera Medica Services pus investment in care home eadership and reationa working within the care home Heath-care professionas working with care homes in site 3 was characterised by: a range of services that were argey reactive, unco-ordinated and did not differentiate between services provided to oder peope iving in their own homes and those in care homes individua expert practitioners with competencies reevant to management of care home residents (e.g. tissue viabiity nurse speciaist, cardiac nurse speciaist) who received referras about care home residents through separate routes and usuay operated independenty of each other evidence that some individua HCPs worked with care homes at an institutiona, rather than individua resident, eve imited forma or informa opportunities to meet with managers and care home staff beyond seeing residents. Speciaist nurse roes for care homes were integrated into district nursing teams with a remit to prevent unpanned hospitaisations and compensate for what was perceived as imited access to GP services. The service mode is summarised in Figure 5. Referra process In site 3 there were a variety of referra routes, incuding a singe point of access, eectronic referras through a shared system, GP referras and informa referras through HCPs visiting the care home. Care homes aso had access to direct ines for a number of individua practitioners incuding the care home nurse and dementia nurse speciaists, dietitian and community nurses. As with site 1, care home staff faced a chaenge as to how to decide where to refer a resident and in what circumstances. The care home nurse speciaist was designated as the first port of ca. However, this ony appied during office hours and their remit ony covered specific conditions. There were guideines for care home staff, but utimatey, their referra decision depended on a number of factors, incuding their eve of experience and the quaity of the working reationships with the HCP concerned: Interviewer: So how do they [care home staff] decide whether to contact you directy or the contact point or the GP? S3CH3HCP01, care home nurse speciaist: We, [aughs]... it s regards preference on who s on duty to be fair, I mean we, they have got information regarding the Community Matrons and amost ike a spider diagram which woud say that Community Matrons support a residentia care homes and give us an idea of the sort of things that we woud see for a the patients. Some carers [care home staff] woud aways ca a doctor, where some more senior carer woud ca us direct, I think if I ve got a reationship with the home, I ve been going for a ong time, they know me very we, then they re more ikey to ca me direct and we d go from there. 74 NIHR Journas Library

111 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. Community nursing team Visits twice a week on set days Direct referra via teephoning DN iaison service Receive referras from GP hospitas dementia nurse speciaist Refers to diabetic nurse speciaist GP dementia nurse speciaist practice nurses paiative care team Out-of-hours team covers nursing emergencies from 5 p.m. overnight Information/communication Access SystmOne Care home 1 No on-site nursing Integrated team Care home nurse speciaist Covers 10 care homes Weeky visits sees up to 10 residents Roe to prevent hospitaisation Nurse prescriber Integrated in community nursing team GP covers out of hours Works cosey with dementia nurse speciaist GP Refers to district nurses GP tissue viabiity nurse speciaist fas service (out of area) paiative care team Optician reguar service NHS and private Arranges ongoing bookings with care home Provides visua awareness training Personaised eye care reports in resident s notes with care pan for staff to refer to Community dentist Visits on request with denta nurses Direct care home referras Provides ora heath promotion training for care home staff Refers to GP, ora surgeon, genera Referras through: singe point of access SystmOne GP some direct service referras Dementia nurse speciaist New service from 2013 Not excusive to care homes but high input with them Based in secondary care but commissioned by primary care Nurse prescriber/medication reviews Supports care home staff in dementia care Prevention of unpanned hospitaisations Direct referra from care homes Works cosey with: intensive dementia service psychiatrist memory service GP (see Figure 2) Information/communication Access SystmOne Overview of site 3 (n = 16) Care homes recruited on county basis via care home organisation Care home nurse works with 3 care homes Dementia nurse speciaist has high input in one care home Dietitian has high input in two care homes Nursing home has no dedicated care home services Speciaist nurses and AHPs NHS Chiropodist private reguar service Visits a residents every 8 weeks Two chiropodists see 20 residents in 1 day Referras to DN or GP via care home staff Care home 4 Dua registration at start of study nursing home by end No district nursing services Respiratory nurse speciaist Direct care home referras Nurse prescriber Referras to GP, DNs, physiotherapists in their team, oxygen team Information/communication Access SystmOne GPs work with care homes using GMS contract (up to 6 practices per care home) Nurse practitioner may substitute for care home visits Community nursing team Covers 6 care homes Daiy visits for insuin Direct referra via teephoning DN iaison service, fax and corridor referras Co-ocated with intermediate care team Receive referras from: GP hospitas care home nurse speciaist Refers to: tissue viabiity nurse speciaist GP end-of-ife team vascuar team continence service care home nurse speciaist intermediate care team Information/communication Access SystmOne OT Integrated with physiotherapists and assistants Care home referras must be via HCP Referras from GPs, socia workers, SALT, DNs, wheechair services, hospita therapists Integrated team Heart faiure nurse speciaist Direct referras to nursing home for known residents Most referras from GP or secondary care Paperess working, shares eectronic notes with other HCPs Refers to GP, PDNS, COPD team, cardiac consutants, heathy iving team Tissue viabiity nurse speciaist Initia assessment with nursing home nurse Referras ony via other HCP Informa training ony Referras to GP, physiotherapist, vascuar team, OT, wheechair services Care homes 2 and 3 No on-site nursing Care home nurse speciaist Roe to prevent hospitaisation Nurse prescribing Integrated in community nursing team Direct care home referras by phone Some GPs prefer their nurse practitioner to see residents on their ist Receive referras from GP (attends some MDT meetings) hospitas DNs direct care home referras Refers to hospitas GP physiotherapist occupationa therapy neuroogica rehabiitation dietitian SALT Information/communication Access SystmOne Information/communication Access SystmOne Information/communication Access SystmOne Physiotherapist Based in rehabiitation team Works with assistant practitioner Care home referras via GP Referras from care home nurse speciaist hospitas, GPs, DNs Physiotherapist Based in discharge team Care home referras via GP Referras from SALT, speciaist stroke service, GPs, DNs, orthotic service Dietitian high input Secondary care based Works cosey with dietetic assistant who does 3-monthy reviews Care home-specific training and support Receive referras from GP SALT Direct care home referras Refers to care home nurse speciaist GP nurse practitioner (based with GP) SALT Information/communication Reguar care home newsetters No access to eectronic infomation Information/communication Access SystmOne FIGURE 5 Overview of heath service provision in site 3. Green font indicates care home-specific services, direct referra, high care home input or organised training. Circuar boxes indicate teams, square boxes indicate individua HCPs. AHP, aied heath professiona; COPD, chronic obstructive pumonary disease; DN, district nurse; OT, occupationa therapist; PDNS, Parkinson s disease nurse speciaist; SALT, speech and anguage therapist. DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO

112 PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES In some areas, the referra system was further compicated by care home providers giving referra guidance to their care home staff that did not fit with the NHS protocos: So sometimes... we find that, especiay the big chains with care homes, they have set guideines as to when to refer and who to refer and that might not actuay tie in with our referra guidance where we ve had referras and we get... Some care homes are tod to refer to us if somebody oses a kiogram which actuay woudn t be significant or it might be two kiograms but if you ve somebody who s actuay overweight and they ose two kiograms, you know, it woudn t be a factor, it woudn t be a referra criteria so that s something that, you know, it depends on (the ownership). S3CH2HCP01, dietitian Team working As in site 2, there was imited evidence of HCPs from different discipines or teams working together on behaf of care home residents, or of referras within and across teams. One exception to this was the dementia nurse speciaist, who worked cosey with the community menta heath team via the psychiatrist and an intensive dementia service that provided day-to-day support for care homes, as we as working to a specific care pathway for medication reviews in coaboration with the memory service (Figure 6). Dedicated heath services for care homes The care homes in site 3 had been recruited on the basis of their invovement in the MyHomeLife eadership training programme. Athough geographicay cose, they were not a within the same CCG area because mutipe CCGs had contracts with the same NHS provider organisation; the same NHS practitioner often visited more than one of the care homes in the case study. Individua nurse practitioners Dementia nurse speciaist roe and integration with other services Intensive dementia support a team day-to-day support Roe Medication reviews Support care home staff in non-pharmacoogica management of dementia-reated chaenging behaviour Prevention of unpanned hospitaisations Informa care home staff training Dementia nurse speciaist a S3CH1HCP03 Commissioned by primary care Referras to Dementia nurse does 6-monthy reviews Discharge back Memory service a Diagnosis Treatment 3-monthy medication review 1-year medication review Liaison, joint visits Care home referras Direct by teephone If unavaiabe, coeague buddy system or CMHT on duty OOH cover is GP Information sharing No shared paperwork with care home With other HCPs eectronic with consent of resident or famiy But she can t see GP information Cose working Attends MDT meetings in arge practices Psychiatrist Community menta heath team GP FIGURE 6 Dementia nurse speciaist roe and integration with other services. a, Same cinica manager. CMHT, community menta heath team; OOH, out of hours. 76 NIHR Journas Library

113 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 dedicated to work with care homes were a feature of service provision in site 3. These practitioners were co-ocated with the community nurse teams. Their service specification was oriented around compensating for care homes ack of access to GPs with an aim to preventing unnecessary hospitaisation. There was, however, evidence of a ack of carity in how these terms had been estabished with some practitioners confating protecting vunerabe oder patients from the harms of hospitaisation with rationing care on the basis of age:... Um, I think the main aim that we... We try to prevent hospita admission, so if we can do something in the home, especiay with our dementia patients that are in homes, you know, we don t want any disorientation, sending someone in, say for instance for, I don t know, because they re constipated for instance, if we can manage that in the home so they re comfortabe in their own room, IV antibiotics to prevent them going into hospita. It has to be an utter emergency for us to send them into hospita because they are, you know, edery. S3CH23HCP03, district nurse team eader The care home nurse practitioners had prescribing rights but they often deferred to the GP regarding medication reviews, especiay if the GP had initiated the medication. Not a residents were seen by the care home nurse speciaists because some GPs empoyed a nurse practitioner to visit in their pace, and preferred for them to co-ordinate care for care home residents on their ist. The nursing home at site 3 did not receive any care home-specific services. The dementia nurse speciaist was empoyed by a secondary care provider organisation. She appeared to be one of the ony practitioners with specific dementia expertise. She had a significant eve of input into care homes but aso worked with oder peope iving with dementia at home. Her roe was to support the care home staff to use non-pharmacoogica approaches to manage dementia-reated behaviours that staff found chaenging. She reviewed antipsychotic medication for care home residents as part of a structured care pathway. However, this approach was restricted by the imited time that care home staff had avaiabe to work with her or to provide one-to-one input with residents: So the strategy is reay sometimes it s aready in the care pan but what we find within the care home sometimes it s written in the care pan but it s not actuay being impemented... so it s reay about occupying the patient, providing that faciity, using distraction techniques, using other methods of music and getting to know the patient. So it s reay, it s quite a chaenge for the care homes because sometimes the staffing eve is quite ow so trying to provide some form of ike compete care for, one-to-one type of care for a patient with dementia who s quite distressed I think that s where the chaenging bits come and that s where a ot of the time you find they wi either use medication to just sort of hep the situation at that moment rather than just ooking at the whoe, somebody just spending a bit of time, you know. Sometimes it takes more than just 10 minutes; reay you re taking about an hour or so with the patient just to sort of cam situations. S3CH1HCP03, dementia nurse speciaist Care home training and monitoring Heath-care professionas offered opportunistic training on request from care home staff. The enthusiasm shown for such a roe was variabe, with some staff reporting that they woud have iked to have had the time and resources to offer more, whereas others did not see it as part of their roe. In common with the community nurses in this site, the physiotherapists and occupationa therapists ony provided training to care home staff that focused on the needs of individua residents, for exampe to demonstrate postura management or exercises. However, three services provided dedicated training to the care home staff in the care home at no cost: the dietitian, the community denta team (which offered ora heath promotion training) and the community optician (who provided visua awareness training). Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 77

114 PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES One exampe of an integrated approach was the dietetic service training for both care home and catering staff. This 6-monthy training programme focused on working together with the care home staff using a specific protoco and a standardised too. Care home staff were trained to use MUST (Manutrition Universa Screening Too; modified to incude pressure sore-reated nutritiona requirements) as a too for picking up any nutrition-reated probems in conjunction with a reated care pan and protoco to foow. This inked to information on how to refer residents to the dietetic service and what to do if a resident had swaowing probems or nausea. This was reinforced by the estabishment of a specific area on the trust website for care homes to refer to, incuding information on MUST, food and body mass index record charts and a reguar care home newsetter with features such as areas of best practice. Dementia-reated training was incuded when taking about the impact it had on residents eating behaviours with catering and care home staff:...whenwe ve done training we ve aso... had peope from the dementia care service who came aong to do part of the tak and the ast time, when we were taking to the catering staff we aso taked about things ike the coours of tabecoths because we had... And we tod them about we had a ady who had a gass that had a fower pattern on it and she d sit and try and pick the fower out, so she never drank anything out of it because she was too busy trying to get the fower and the same with if there s a pattern on the tabecoth... and we taked about, you know, cooured pates and how the food ooks on pates... so if you have white food on a white pate somebody with dementia s not going to be abe to eat it because they re not going to be abe to compute what it is. So things ike that we wi tak about. S3CH2HCP01 Prior history of working with care homes There was reference to a history of piot projects in the education of dementia-reated hospitaisations in secondary care prior to introducing the dementia nurse speciaist roe in The care home nurse speciaist roes had been impemented 10 years ago. The dietitian had previousy received funding by the medicines management group to undertake a nutrition project that informed the way in which they currenty worked with and provided training to care homes. There was evidence that these initiatives had been more stop start than those described in site 1, with ess opportunity for staff from one NHS initiative to move onto the next, with consequent oss of continuity. Genera practitioner invovement with care homes across the sites The GP accounts are set apart from the site-by-site presentation because, out of necessity, not a GP participants recruited were directy invoved with the study care homes. The GPs eventuay recruited were, however, a working directy with care homes within the case study sites. Key to the use of incentives across the three sites was how their purpose was interpreted by the GP, as the means to do work they vaued or as payment for activities that were compensating for care home staff shortcomings or firefighting to reduce the number of hospitaisations. Tabe 17 provides a summary of these accounts by site. Across a sites, GP participants consistenty stated that, if their time was aocated and resourced to work with care homes, they woud do more than respond to urgent care requests. Site 2 had an exampe of a time when they had used winter pressures funding to visit care homes 7 days a week, and GPs were sure that this had averted hospitaisations. However, such proactive working was not seen universay as a good thing. Severa GPs at site 2 stated concerns that increased input woud create dependency on GPs by care home staff. They were worried that this might reinforce reuctance among care home staff to take responsibiity for decision-making about what was urgent and what was not. This worry, that additiona funding coud seed unmanageabe eves of demand that woud persist beyond the time-imited nature of the additiona payments, was used by severa GP respondents as the rationae for focusing on individua resident contacts, rather than engaging with the care home more structuray. 78 NIHR Journas Library

115 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 TABLE 17 The GP invovement in care homes Site Characteristic GP invovement Minimum bimonthy reguar visit/cinic (most do every week or every 2 weeks) pus responsive visits to urgent cas pus exampes of monitoring at-risk residents, popping in, messaging, texting and teephone conversations Practice aocated to specific care homes. Emphasis on responsive working but visiting two or three times a week to faciitate this. One GP described fixed Friday visits as popping in before the weekend Winter pressures money = GPs proactive visiting every day, 7 days a week, emphasis maintained on responding to crisis and avoiding admissions at an individua resident eve Panned and urgent May send practice nurse/nurse practitioner in their pace Fax communication of referras to GP regarded as norma practice Monthy cinics pioted but not sustained because of a ack of funding GPs ooking after three or four care homes each How services are provided Cinic seeing 6 18 peope at a time. Operationaised differenty in different practices. One GP described setting aside 4 hours one week and shorter cinic the next, whereas another consistenty did a 1-hour cinic each week Reguar visiting, where GPs were frequenty asked to see more than one resident. A care home visit combined urgent and panned care (something that GPs strugged to accommodate within fixed duration visits) In response to request to see residents Speciaist nurses with remit to reduce hospitaisation worked in three care homes Working reationships Different demands from care home with and without nursing but estabished reationships. GPs encouraged to work with care home at an institutiona eve to estabish expicit arrangements for joint working Acknowedgement of variabiity within the sector. Certain care homes identified as not a probem, whereas others strugged to hod probems pending the next GP visit Expectation that care homes wi have some abiity to triage ca-outs and estabish which residents require more urgent medica attention History of working with care homes Owing to duration of current and preceding ocay enhanced service agreements, GPs have > 2 years experience of working with care homes Worries that coser working with care homes coud raise expectations and estabish unsustainabe patterns of working, creating dependency on GPs Past history of GPs having private retainer fee arrangements with care homes paid for at an institutiona eve in exchange for more frequent visiting In site 3, GPs described how they had provided a care home-based cinic when there was funding, but that when that funding was no onger avaiabe, it ceased. In site 1, additiona payments to GPs had been perceived as egitimising more structured engagement with care homes. This was, however, in the context of the other comprehensive care home services (see Figure 4). The GP interviews, however, did not suggest that they recognised this provision as sharing the burden; it was not cear how much the GPs coaborated with the care home teams. It was more that the site 1 GP accounts suggested that they had more time and opportunities to think more constructivey about engagement with the sector. There was cross-site recognition of the importance of having confident, we-estabished care home staff who had been there for years, which in turn infuenced how staff managed their anxieties about patient care and communicated these to GPs. None of the GPs interviewed beieved that they shoud be invoved in education or training of care home staff. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 79

116 PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES In site 1, the GPs had worked in individua 1 : 1 reations with care homes, underpinned by forma contractua arrangements, for at east 2 years, and had estabished ways of working. There was an emphasis on the persona reationship they had buit up with care homes. Site 1 had mutipe accounts of a history of working together and mutipe methods of communicating by, for exampe, using text messages and teephone cas to faciitate and support direct face-to-face contacts. The GPs interviewed were interested in care homes and gave accounts of being invoved in fund-raising for homes and setting up protocos for shared working. Interestingy, this reationship was represented as separate to the work done by the speciaist care home team and other services; indeed, there were some doubts around whether or not the care home speciaist team hindered GP work because of a perceived reuctance to share information about residents: Interviewer: You mentioned difficuties in communication with the care home nursing team? S1GP04: That s the ony thing... I mean it works perfecty fine, I mean I can see what they ve written in the notes and they can see what I ve written, but there is no one person ike. Interviewer: Not a main person you can contact? S1GP04: Yeah, I mean recenty there seems to be one nurse who s coming, I met her, and we see if it progresses to anything. Inappropriate cas were seen as the biggest threat to working reationships in sites 2 and 3. The confation of panned care and urgent (and thus unpredictabe) requests was highighted as probematic. One GP observed that it was not the residents who were chaenging, it was more that care home staff coud not (or woud not) differentiate between what was urgent and what coud wait: If it was just the residents and not the repeat ca-outs for nothing very much then I think that woud be fine that woud be workabe. Site 2 GPI01 This was acknowedged but not represented so negativey by site 1 GPs. This may be because the other care home services in site 1 absorbed some of the demands experienced by GPs in other sites. The organisation of NHS care in and around care homes was uncear to GPs in a the sites. They were aware of some but not a services (e.g. dementia outreach team or a particuar nurse speciaist), but did not work cosey with them and in some cases these were regarded as being unhepfu (e.g. aso prescribing for their patients). A cross-cutting issue when providing heath care that GPs raised was the support of peope with dementia whose behaviours care home staff did not know how to hande. Continuum of association The organisation and funding of NHS services to care homes in the three sites refected a continuum of association that in part refects how reationa patterns of working have deveoped over time. Figure 7 provides a graphic iustration of how the different services described were organised around the care homes and the eve of horizonta integration achieved. The pae green circes denote an expicit (funded) commitment to spend time working with care homes. Site 1 HCPs working with care homes aso iaised with each other in how they worked with care homes. Sites 2 and 3 individua HCPs invoved other services, but this was on a resident-by-resident basis and there was imited evidence of shared working. The green circes denote those services that visited care homes, but the frequency and intensity of their invovement were at the discretion of the HCPs invoved. This was in turn infuenced by the demands of their wider caseoad. 80 NIHR Journas Library

117 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Site 1 Site 2 Site 3 AHPs GP GP AHPs GP AHPs Nurse speciaists Nurse speciaists Care homes Rehabiitation team Care homes Nurse speciaists Care homes Dietitian Care home team Dementia team Paiative care nurse speciaist DNs DNs Care home nurse speciaist Fas team FIGURE 7 Continuum of horizonta integration of heath services with care homes across the sites. AHP, aied heath professiona; DN, district nurse. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 81

118 PHASE 2 CASE STUDIES: COMPARATIVE DESCRIPTION OF THE STUDY SITES Concusion Chapter 5 has summarised in detai a within- and cross-case narrative of NHS provision to care homes from the three case study sites, with specific reference to those activities reated to the areas of interest identified in phase 1. Resident baseine characteristics were very simiar across the three sites, indicating broad comparabiity in terms of the resident cohorts and providing justification for cross-site anaysis. There were differences in ength of association, intensity of contact and opportunities for consutation, how particuar services were organised and used, how participants worked with each other and how they interpreted their responsibiities. Chapter 6 wi now go on to report a detaied site-by-site anaysis of the outcomes of interest. 82 NIHR Journas Library

119 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Chapter 6 Phase 2 case studies: case study outcomes of interest Introduction This chapter focuses on describing the outcomes of interest, identified through the earier reaist synthesis described in Chapter 4. These were medication use; use of out-of-hours service; resident, carer and staff satisfaction; unpanned hospita admissions (incuding A&E); and ength of hospita stay. These are described across the three case study sites before a cross-case anaysis of differences and commonaities across the three sites. Service use data by site Service use data were anaysed based on compete cases at 6 months (85% of the baseine sampe) because this provided the best baance of duration of foow-up and competeness of the data set, in order to avoid confounding introduced by missing data after the 6-month cut-off point. Service use contacts are summarised by site in Tabe 18. Most residents had infrequent heath service use of many heath services over the 6-month period, resuting in high eves of no utiisation for many of the resource items when considered on their own. For this reason, services were combined into eight categories. GPs were the most heaviy utiised group; over 90% of residents in each site had some eve of genera practice contact during the 6 months. Community care contact, which was made up of community nurses, aied heath professionas and other speciaist community care contacts, was the next most utiised group with approximatey 88% of residents in sites 1 and 3, and 86% in site 2, having some eve of community care. A other services were used by reativey sma proportions of residents. There was significanty more use of primary care (other than GPs, i.e. dentists, opticians, chiropodists) in site 3, than in sites 1 and 2. Secondary care (ambuatory/no admission) was significanty higher in site 3, than in site 2. Athough more residents in site 3 had hospita admissions than in the other two sites, the differences were not statisticay significant. Costs by site Costs are summarised by site inferences and other covariates (Tabe 19). Average cost of hospita stay per resident was the singe greatest contributor to costs at a three sites and did not show statisticay significant differences between sites. The average tota cost of 6 months of heath resources per participant, excuding in-hospita stays, was 634 for site 1, 730 for site 2 and 880 for site 3; when the cost of hospita admissions was incuded, the means increased to 1160, 1190 and 2096, respectivey. Excuding hospita stays, GP and community costs accounted for most costs. GP costs were significanty higher in site 2, where financia incentives provided to GPs to increase frequency of contact were a mainstay of the service mode, than those in sites 1 and 3. Costs for dentistry, opticians and chiropody were higher in site 3 than in site 2. Athough not statisticay significant, it is notabe that costs were higher in site 3, where care modes were ess specified and structured, than in sites 1 and 2. This was argey because of the increased expense associated with hospitaisation. Site 3, however, was associated with greater access for residents to dentists, opticians and chiropody services. There are some caveats to this anaysis. In particuar, no account is taken of the on-site nursing resource in nursing homes. Residents in residentia homes without on-site nursing receive care from community or district nurses, and this was refected in the service use frequencies coected from residentia homes, and in the cost cacuations. No data were avaiabe, however, on use of on-site nurses by residents in the sampe who were iving in nursing homes. The Care of Oder Peope UK Market Report 2014/15 12 suggests a Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 83

120 84 NIHR Journas Library TABLE 18 Service contacts by site Care/service type Site, contacts per resident 1(n = 77) 2 (n = 69) 3 (n = 49) Mean (SD) Median (range) Mean (SD) Median (range) Mean (SD) p-vaue Median (range) Site 1 vs. 2 Site 1 vs. 3 Site 2 vs. 3 GP care contacts 4.06 (2.97) 4 (0 17) 6.04 (4.49) 5 (3 8) 4.45 (3.98) 4 (0 20) Community care contacts 9.74 (22.02) 5 (0 189) (51.39) 3 (1 7) (76.05) 3 (0 376) Primary care (optician, dentist, pharmacist) 0.30 (0.51) 0 (0 2) 0.39 (0.79) 0 (0 1) 0.76 (0.78) 1 (0 3) 0.40 < 0.01* 0.01* Out-of-hours care (GP or nurse) contacts 0.25 (0.91) 0 (0 7) 0.35 (0.61) 0 (0 1) 0.31 (0.74) 0 (0 4) A&E visits 0.23 (0.79) 0 (0 6) 0.14 (0.49) 0 (0 0) 0.06 (0.24) 0 (0 1) Secondary care non-admitted contacts 0.43 (0.97) 0 (0 5) 0.19 (0.69) 0 (0 0) 0.65 (1.07) 0 (0 5) * Secondary care admissions 0.25 (0.61) 0 (0 3) 0.17 (0.42) 0 (0 0) 0.33 (0.55) 0 (0 2) Ambuance use 0.35 (0.82) 0 (0 5) 0.28 (0.70) 0 (0 4) 0.35 (0.56) 0 (0 1) *, statisticay significant. PHASE 2 CASE STUDIES: CASE STUDY OUTCOMES OF INTEREST

121 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. TABLE 19 Costs by site NHS service used Site, cost ( ) per resident Mean (SD) Median (range) Mean (SD) Median (range) Mean (SD) Median (range) p-vaue Site 1 vs. 2 Site 1 vs. 3 Site 2 vs. 3 GP (GP and practice staff) 175 (132) (90 225) 270 (194) 225 ( ) 189 (174) 135 (56 270) < 0.01* * Community a 224 (384) 143 (44 281) 292 (876) 70 (17 184) 450 (1275) 79 (50 215) Primary care (other) b 18 (34) 0 (0 39) 20 (41) 0 (0 39) 40 (60) 39 (0 39) * 0.05* Out of hours 17 (62) 0 (0) 23 (40) 0 (0 0) 21 (51) 0 (0 0) A&E not admitted 32 (107) 0 (0) 20 (67) 0 (0 0) 8 (33) 0 (0 0) Secondary care not admitted 88 (231) 0 (0) 42 (225) 0 (0 0) 99 (205) 0 (0 134) Admissions (cost) 525 (1888) 0 (0) 519 (1913) 0 (0 0) 1202 (3326) 0 (0 512) Ambuance use 81 (190) 0 (0) 64 (163) 0 (0 0) 80 (130) 0 (0 231) Tota cost 1160 (2184) 492 ( ) 1190 (2250) 439 ( ) 2069 (3745) 682 ( ) Tota cost excuding hospita admissions 634 (687) 458 ( ) 730 (991) 413 ( ) 880 (1320) 493 ( ) *, statisticay significant. a Community care comprises district nurse, practitioner nurse, continence nurse, diabetes meitus nurse, continuing heath-care nurse assessor, care home nurse team, Parkinson s disease nurse, chronic obstructive pumonary disease nurse speciaist, cardiac nurse speciaist, registered menta heath nurse, CPN, menta heath team visit, nutrition speciaist nurse, anticoaguant nurse, best interests team assessment, paiative care team visit, community matron, physiotherapist, occupationa therapist, speech and anguage speciaist, dietitian, audioogist, psychotherapist, psychoogist, socia worker, chiropodist, other dementia, other phebotomy, other fas prevention, other ong-term care, other speciaist contacts and geriatrician. b Primary care comprises dentist, optician and pharmacist visits. DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO

122 PHASE 2 CASE STUDIES: CASE STUDY OUTCOMES OF INTEREST difference of 226 per week in the average fees of private nursing and residentia homes ( 821 vs. 595, respectivey), and this might be taken as an indicative cost to be added to the cost of other services used by nursing home residents. This is a particuar probem because the proportion of residents from nursing (vs. residentia homes) incuded in the study was not the same between sites. There were three and four times as many nursing home residents in site 2 as in sites 1 and 3, respectivey, thus it is possibe that the costs for community care on site 2 woud have been greater had it been possibe to account for this nursing resource use in the costing mode. A further hidden cost is that of the financia incentives paid to GPs in site 2 for enhanced services to care homes, impying that the fu economic cost of GP services in that site is higher than shown. The service frequency and cost comparisons are based on the differences in unconditiona means. The subsequent anaysis contros for baseine differences and other covariates. Regression anaysis of outcomes taking account of baseine variabes by site As described in Chapter 2, we used Poisson regression to consider whether or not a site was a significant predictor of service use. Tota costs after the baseine interrai scores and the derived interrai scaes, as we as the interaction between the site and these variabes, were taken into account. We first entered each of the baseine interrai scores and the derived variabes into a Poisson regression equation in univariate anayses using each of the categories of service use and tota costs as dependent variabes. The resuts from this anaysis are shown in Appendix 11. Using a significance eve of 0.10, we then tested interaction terms for each site and each baseine variabe or derived variabe found to be statisticay significant in the univariate anaysis. In a third and fina step, significant interaction terms were entered aong with the site and the variabes found to be significant in the univariate anaysis to determine whether or not the site was a significant determinant of differences between outcomes. Gender, age, bed type (residentia or nursing) and payment source (sef-funder vs. received state contributions) were incuded in the anaysis based upon pre hoc assumptions derived from the existing iterature about the ikey roe that they might pay in confounding outcomes. The continuous predictors were mean centred; the mean was subtracted from the score to faciitate interpretation of the main effect of the site. The outputs of this process are summarised in Tabes For GP contacts we found that site 2 had 1.40 times as many GP contacts as site 1 (p < 0.01). This remained significant after TABLE 20 Significant predictors of GP contacts: site 1 used as reference category Variabe IRR SE z p-vaue 95% CI Pearson s R 2 Site < to Site to 1.46 Gender to 1.16 Age to 1.02 Bed type to 1.17 Payment source to 1.01 spurs: site to 1.11 Comorbidities < to 1.20 spain_ < to 1.61 Medication count < to 1.09 spurs: site < to 0.99 spurs: site < to 1.31 IRR, incidence rate ratio; SE, standard error. 86 NIHR Journas Library

123 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 TABLE 21 Significant predictors of primary care contacts: site 1 used as reference category Variabe IRR SE z p-vaue 95% CI Pearson s R 2 Site to Site < to 3.44 Gender to 1.34 Age to 1.04 Bed type to 2.52 Payment source to 1.02 scps to 1.24 Cognitive impairment < to 0.91 IRR, incidence rate ratio; SE, standard error. TABLE 22 Significant predictors of out-of-hours contacts: site 1 used as reference category Variabe IRR SE z p-vaue 95% CI Pearson s R 2 Site < to Site < to Gender to 1.57 Age to 1.06 Bed type to 1.70 Payment source to 1.01 spurs to 1.26 sadlsf to 1.21 Comorbidities to 1.80 spain_ to 9.45 IRR, incidence rate ratio; SE, standard error. the effect of other associated variabes, namey comorbidities, pain and medication count, and the interaction term between site 3 and pressure ucer risk were taken into account. For primary care (optician, dentist and pharmacist) contacts, we found that site 3 had amost twice as many of these as site 1 (p < 0.05). This remained significant after the effect of other associated variabes, namey payment source and cognitive impairment, were taken into account. For out-of-hours contacts, we found that sites 2 and 3 had more than four times as many contacts as site 1 (p < 0.05 for both). Interpreting service use data from the case studies Before moving on to consider the specific substudies around unpanned hospitaisations and medication use data, and the quaitative data about the outcomes of interest, the quantitative outcome data wi be discussed. Within the broader anaysis, these contain important, abeit unsurprising, insights about service use; namey that pain, pressure ucer prevaence, medication use and comorbidities are predicted for increased heath service utiisation among care home residents (see Appendix 12). Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 87

124 PHASE 2 CASE STUDIES: CASE STUDY OUTCOMES OF INTEREST For the most part there were no compeing differences in service use or costs between sites. Site 3 might have been expected to have been substantiay cheaper, given that the cohort recruited here was substantiay ess dependent; however, this was not in fact the case. Site 1 might have been expected to have been substantiay more expensive because of markedy different modes of service provision that were more formaised and not primariy focused around the GP, but this was not the case. Site 3 used more chiropody, pharmacy and optician resource than sites 1 or 2 and spent significanty more money on doing so, athough, in rea terms, the excess expense was sma. This may have indicated better access to these services on this site, rather than excessive use of these resources. Site 3 aso had a greater number of secondary care non-admitted contacts, as we as a trend towards higher costs associated with hospita admissions. This may indicate a tendency at this site to refer residents into hospita, rather than provide care in situ. Genera practitioner attendances at site 2 cost more than in either sites 1 or 3. This is consistent with the mode of care at this site where additiona payments were designed to increase the frequency of GP attendances in care homes. Once baseine variabes (incuding the functiona dependency and comorbidities of the cohorts) were controed for, there was evidence that GPs attended more frequenty in site 2 than in site 1 and that there were more frequent out-of-hours attendances in sites 2 and 3. Site 1 was no more expensive in any domain than the other two sites. It did not see a greater number of contacts for any one service than the other two sites. Substudy of unpanned hospitaisations To gain a better understanding of the nature and type of unpanned hospita admissions, we conducted a descriptive anaysis of the data, summarised in Tabe 23 (see aso Appendices 11 and 12). In tota, 39 residents were hospitaised at some point during the 12-month data coection period, just 16% of the tota number of residents recruited to the study. The ength of stay ranged between one night (n = 17) and 47 nights for one case invoving a dementia-reated menta heath assessment, with 22 residents being hospitaised for more than five nights in one episode. These support the findings from the quantitative anaysis above of a greater reiance upon secondary care in site 3, but aso highight the tendency for patients to stay much onger in hospita in this site. The reasons for residents being hospitaised did not differ much across the three sites (Tabe 24). The most common reason was fas (n = 13), eight of which resuted in a fracture, foowed by respiratory-reated conditions (n = 6), incuding pneumonia, chest infection and breathing difficuties. No information was avaiabe in the care home record about the reason for admission or discharge diagnosis for six admissions. Substudy of medication data Medication data were avaiabe from 214 out of 239 residents. A residents from site 1 had fu medication data. Twenty-four residents from site 2 had missing data, reated to difficuties in obtaining baseine TABLE 23 Unpanned hospitaisations by site Site Number of residents hospitaised Hospitaisations as a percentage of the tota number of residents recruited Number of residents hospitaised for... 1 occasion > 1 occasion 1 night 5 nights % (n = 93) % (n = 92) % (n = 57) Tota 39 16% of a residents NIHR Journas Library

125 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 TABLE 24 Reasons for hospitaisations recorded from residents care home notes Site Reason for admission Number of admissions Reason for admission Number of admissions Reason for admission Number of admissions Fa 1 Fa 2 Fa 2 Fa and fracture 3 Fa and fracture 3 Fa and fracture 2 Respiratory conditions 2 Respiratory conditions 2 Respiratory conditions 2 UTI 3 Urinary retention 2 Urinary retention 1 Syncope 1 Osteomyeitis 1 Hypotension 2 Pyrexia 2 Pyrexia 1 Abdomina pain 1 Transient ischaemic attack 1 Chest pain 1 Costridium difficie infection 2 Chest pain 1 No information 1 Recta beed 1 Vomiting 1 Menta heath assessment 1 No information 3 Not eating or drinking 1 UTI, urinary tract infection. No information 2 medication data from one care home. One resident from site 3 was not taking any medication. Where baseine data were successfuy retrieved, our mechanism for coecting foow-up data from the care home yieded a fu foow-up data set. Thus, there were 90, 68 and 56 residents from each of sites 1, 2 and 3, respectivey, with fuy anaysabe medication data. The mean number of medications at baseine for a sites was 8.28 (SD 3.4), and 8.00 (SD 3.5), 8.24 (SD 3.6), 8.77 (SD 3.1) for sites 1, 2 and 3, respectivey, with no statisticay significant difference between sites. The range of medications was Forty-nine residents were taking opioids at baseine, with 47 of those taking a singe opioid medication. Two were taking two opioid medications. Participants were significanty ess ikey to be taking opioids at site 1 (p < 0.01). Thirty-four residents were taking antibiotics at baseine. Participants were significanty ess ikey to be taking antibiotics at site 3 (p < 0.01). The median (range) ACB score was 1 (0 14), with no statisticay significant difference in distribution of ACB scores between sites. Over the study period, there were 366 medication changes (starting, stopping or substituting a medicine) in site 1, 261 changes in site 2 and 266 in site 3. This represents 0.40 (SD 0.66), 0.44 (SD 0.84) and 0.49 (SD 0.79) changes per resident per month in sites 1, 2 and 3, respectivey. Foow-up data did not demonstrate consistent trends in antibiotic or opioid prescribing. Site 3 reported the owest ACB score throughout the study, with evidence of a rising ACB score in sites 1 and 2 (Figure 8). In summary, the cohort was representative of UK care homes generay in terms of the prescribing rates seen. For most variabes measured, the sites were not substantivey different at baseine and differences seen at baseine in antibiotic and opioid prescribing disappeared with foow-up. The impications of the tendency towards ower ACB scores in site 3 are uncear; it coud either be attributabe to the dementia speciaist nurse s invovement in the reduction of antipsychotic prescribing in the study care homes or be indicative of a different cuture of care within the care homes in site 3, consistent with their participation in the My Home Life eadership training programme. There is, however, no evidence to suggest that increased GP contact in site 2, or case management as part of wrap-around care, worked to optimise prescribing in any way over the more traditiona modes of working seen at site 3. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 89

126 PHASE 2 CASE STUDIES: CASE STUDY OUTCOMES OF INTEREST 4 3 Mean ACB score 2 Care home site Site 1 Site 2 Site Time point (months) FIGURE 8 Mean ACB score across the three study sites over 12 months. Error bars: 95% CI. Staff satisfaction substudy Data coection for this substudy took pace towards the end of the case studies, by which time the research team were aware that the care homes were suffering from research fatigue foowing the case study process, and it was anticipated that engagement might be suboptima. A tota of 562 questionnaires were sent out by post, with two rounds of foow-up teephone reminders on a weeky basis to maximise response rates. The forms were misaid at two sites and further copies were sent. One care home subsequenty refused to return their staff questionnaires because of a change in management and a decision that supporting this fina stage of the study was no onger a priority. Consequenty, out of a potentia tota of 562 questionnaires, ony 94 were returned, a response rate of 16.7%. Sites 1, 2 and 3 returned 55, 14 and 25 questionnaires, respectivey. The buk of responses from sites 1 and 3 came from care assistants, whie in site 2, the number of responses was ow overa (Tabe 25). There were no differences between the sites in respondents age, duration in current profession or current post. Patterns of working differed, with participants significanty more ikey to undertake spit shift working on site 1 (Tabe 26). TABLE 25 Overview of responses by site Site, number of responses Profession Tota Care assistant Registered nurse Care home manager Activity co-ordinator Other No repy NIHR Journas Library

127 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 TABLE 26 How working hours were organised Site, number of participants How are working hours organised? Tota Schedued working hours with spit shifts Schedued working hours without spit shifts Part-time sick eave Fu-time sick eave Leave of absence Other No repy A tota of 83 out of 94 (88%) respondents thought that their workpace was staffed sufficienty, either sometimes or usuay, with ony 10 out of 94 (11%) respondents raising concern about staffing eves; 91 out of 94 (97%) respondents were somewhat or absoutey satisfied with their current working hours; and 79 out of 94 (84%) respondents fet abe to infuence their current patterns of working. There were no significant differences between sites for any of these variabes. In a sites, the buk of the working week was spent caring directy for residents (Tabe 27). Satisfaction with the care provided within the care homes was high across a sites (Tabe 28), but was significanty higher for information provided to residents about work routines and nurse in charge and activities provided to residents in site 1 (p < 0.05). Amost a (96%) of the sampe responded to the question about quaity of care provided to residents by the NHS on an anaogue scae from 0 to 10, whereby 10 represented extremey satisfied and 0 represented not at a satisfied. The mean score was 7.5 (range 3 10), with no significant difference between sites. The ow response rate means that the data may we be biased in favour of positive responses from those staff members who had remained enthusiastic about the project throughout. Despite this, the staff responses suggest a staff group that, on the whoe, reported high eves of job contro, job satisfaction and satisfaction with the services provided by their empoying care homes and the NHS services with which they interfaced. This is somewhat contrary to the narrative of a sector commony portrayed as beeaguered and on the edge of coapse. Few significant differences were observed between sites, but where they did occur they tended to favour site 1. TABLE 27 Breakdown of the average working week How working hours are distributed on an average week: a sites Activity 0% < 25% 25 50% 51 75% > 75% No repy Working directy with residents Administration Ceaning/service Trave Other Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 91

128 PHASE 2 CASE STUDIES: CASE STUDY OUTCOMES OF INTEREST TABLE 28 Staff satisfaction with care provided within the care home Satisfaction Very good Quite good Quite poor Very poor Cannot judge No repy Site 1 Information, work routines, nurse in charge Staff Activity Care Site 2 Information, work routines, nurse in charge Staff Activity Care Site 3 Information, work routines, nurse in charge Staff Activity Care Quaitative data on the outcomes of interest This section focuses on findings that reated to the five outcomes of interest: medication use; use of out-of-hours services; resident, carer and staff satisfaction; unpanned hospita admissions (incuding A&E); and ength of hospita stay. The findings were drawn from a participants accounts (residents, famiy, care home staff, HCPs, GPs and commissioners). This section considers what these accounts reveaed about the eements within the service deivery modes offered in the three sites that phase 1 findings had theorised were important to achieve the outcomes of interest. Specificay, how different contexts generated (or did not generate) staff responses that supported reationa working, which in turn supported residents care and integration of services, was considered. These data shoud be seen as compementary to the quantitative data about the outcomes of interest aready presented. Use of medications Across a three sites, HCP accounts identified a range of issues that were medication specific and coud not be identified as site specific. These incuded concerns about administration and recording errors, care home staff knowedge of pharmacoogy, difficuties of prescribing for wound management, the chaenges of mutipe prescribers visiting care homes and the importance of access to emergency end-of-ife medication out of hours. A ack of pharmacist invovement in medication reviews was highighted as a gap in service provision across the sites. Site 1 appeared to have the highest number of nurse prescribers. Speciaist nurses were invoved in reviewing specific medications in sites 1 and 3, incuding medications for dementia and osteoporosis. However, in a the sites nurse prescribing was not a substitute for GP invovement. There were situations when nurse prescribers deferred to the GP, for exampe when it was quicker to access the medication through the GP prescribing route, where the GP had initiated the medication or where there was a protoco in pace for GPs to review medications once the resident s condition was stabe. 92 NIHR Journas Library

129 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 In site 2, the tissue viabiity nurse speciaist highighted that nurses working as part of the care home staff did not identify products from the same formuary for wound care products as the tissue viabiity service and so requested different products from the GP. There were mutipe routes to obtaining a prescription for a resident and it was difficut to identify how care home staff decided whom to approach and invove. In site 3, there were no referra guideines for care home staff in reation to dietary suppements. This was seen as eading to dupication of consutations with the dietitian, nurse practitioner and GP, inconsistent prescribing practices and a ack of carity for the care home. The introduction of antipsychotic reviews by the dementia nurse speciaist was reported to have resuted in significant reductions in prescribing of these medications and there was evidence of ower ACB scores in residents at this site. These reviews were sustained over the ength of the study (Tabe 29). The care home staff interviews and focus groups provided an aternative perspective on medication management, with three main themes emerging: medication reviews, over-the-phone prescribing and chasing prescriptions and medications. Differences between accounts from the three sites were apparent. When GPs hed reguar cinics in the care homes, there were few or no references to medication-reated probems, and reviews were conducted more frequenty. In site 3, apart from a time-imited piot when GPs had offered care home-based cinics, GPs visited ony on request. Care home staff were either unaware that the GPs had competed reviews or reported having to remind GPs to compete them. Over-the-phone prescribing was a particuar issue for the care homes in site 3. In site 1 this ony occurred occasionay and care home staff were generay of the opinion that the practice was unacceptabe uness there were exceptiona circumstances, for exampe for the benefit of a resident with specific needs such as end-of-ife care: No, majority, they [GPs] wi come out. If I ve got someone on end of ife, and say ike they ve got thrush in the mouth, then they wi prescribe over the phone because this person s end of ife, especiay if it s a Friday, we re not going to wait unti the Monday because it s going to be uncomfortabe, they re not going to drink, they re not going to want mouth care. So I think certain aspects they wi, but nine out of ten I think the doctor wi come out. S1CH3, care home staff 01 interview For care home staff, the ogistics of obtaining prescriptions for residents were often described as being difficut. In sites 2 and 3, care staff described being continuousy on the teephone chasing prescriptions and acting as the go-betweens for GPs or district nurses and pharmacists to sort out medication issues. Some staff in site 2 residentia care homes found this particuary difficut to negotiate as a resut of their ack of medication-reated knowedge. Care home staff across the sites aso taked about district nurses TABLE 29 Medication review by GP approaches taken across the three sites Site Cross-cutting theme Mixed approach to medication review, but evidence of systematic approaches (taking patient records to care home) and working to reduce tota medication burden Mixed, mainy opportunistic review of medication when visiting or as a desk-based review exercise Panned care home-specific infuenza vaccination programme Exampe of targeting residents at risk, for exampe mutipe ca-outs, hospitaisations Piot cinic: reviewed medication of residents Confusion between GPs about what they have to do for care homes when doing medication reviews. Mainy opportunistic, when resident is seen (see most > 6-monthy) If a care home paid a retainer, GPs were wiing to do a monthy medication review No or minima pharmacist invovement; this was seen as a potentia area for deveopment Wiingness to review and pan future medication needs for peope recognised as dying Difficut prescribing for peope with behavioura symptoms that staff found chaenging Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 93

130 PHASE 2 CASE STUDIES: CASE STUDY OUTCOMES OF INTEREST prescribing dressings or other items for residents and expecting them to arrange their deivery, which they did not see as part of their roe. This was work for care home staff that arose from HCP visits and that was unknown and unacknowedged: Interviewer: So when things don t work so we with pharmacy, what s that about? S1CH1 care home staff focus group: What it is, isn t it, it s when the doctor s signed a prescription and it s our responsibiity to do medication, bah bah. When it s the district nurses right, it s happened a ot, hasn t it, where they expect us to get it across or whatever, dressings, but they order the, don t they, at the side of the... and they turn up and the item s not here. That s not our responsibiity to order that item, that s in their department, whereas ours is with the pharmacy medication. Care home staff who participated in the interviews did not appear to have received specific education or training in medication management from the visiting NHS services, incuding those in site 3 where there had been a focus on reducing antipsychotic prescribing. This may be a refection of the eve of staff turnover in care homes or that those interviewed had not been invoved in this programme of work. In a three sites residents identified access to medications as a concern. In Site 1, residents knew that the GP visited reguary and at predictabe times and wanted to have the opportunity to ask questions concerning their medication and treatment. To be seen, however, residents had to ensure that they or staff had written their name into a book. If this was not done, or the GP was fuy booked, this arrangement was perceived as deaying rather than faciitating access to care and, specificay, prescriptions. An awareness of GP invovement raised expectations about access that were not so apparent in the other sites as the foowing quotation shows: I have been waiting over a week for my prescriptions now. But I have marveous treatment here. The girs are marveous. S1CH3R01, resident Famiy members across the sites discussed medication. Severa acknowedged the compexity of the medication regimes and the importance of getting it right. For some, how medication was managed was a proxy measure for the quaity and personaised care provided by the care home: Yeah, because he s got a very compicated medication routine, and that s... so he s got, you know, he s given tabets every 2 hours, 2 and a haf hours, so somebody, somebody does that for him, and he aways, you know, if he s not feeing we, that s monitored fairy cosey. S1CH2F03 In a sites, the medication management was perceived as being mosty good, but there were some bad experiences. Exampes came from sites 2 and 3 and incuded not being informed about medication changes, over-the-phone prescribing, deayed prescriptions and errors in dispensing. Famiy members aso emphasised the importance of access to the GP to review, revise and access their reative s medication. The drawbacks of not invoving famiy members in decisions about starting and stopping medication were highighted by one reative, where a medication s positive effects were seen as having been ost: Now unbeknownst to us, the doctor put mum on a very mid antidepressant, without any consutation with us at a. But both us and the home noticed a significant change in mum s attitude, outook, she became much brighter... about a year they said that the doctor had just done a routine visit and decided that he was going to take her off of them... [this was seen as having resuted in a severe depression]. S3CH4F01 94 NIHR Journas Library

131 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Out-of-hours services A HCPs recognised that part of their invovement with care homes was to reduce unnecessary hospitaisations. In sites 2 and 3, some HCP posts had been introduced with a specific remit to reduce out-of-hours and emergency service use. This incuded the dementia nurse speciaist and care home nurse speciaists in site 3. In site 1, avoiding crisis admissions and GP ca-outs was the underying rationae for the case management work with new care home admissions. Despite this provision, additiona training for care home staff on sites and access to speciaist nurses were not aways seen as being abe to compensate for other service and system imitations. Apart from the GPs, a of the care home services worked office hours, and out-of-hours service provision did not aways fit around the needs of oder peope iving with dementia in care homes. An exampe was that residents had to be admitted to the emergency department at the weekend if they needed to be seen by the psychiatrist, running the risk that they might be admitted to a receiving medica unit even when a primary psychiatric diagnosis had been identified as the main issue by the care home staff or visiting HCPs. Staff at three care homes in site 1 had positive experiences of out-of-hours services responding quicky once they had been contacted. However, in sites 2 and 3, care home staff experiences of these services were poor. In particuar, no account was taken of care home staff knowedge of residents or experience, together with out-of-hours practitioners not differentiating between nursing homes and residentia care homes in terms of staff knowedge and skis. They found the out-of-hours process protracted, with ong deays between speaking to an advisor and residents receiving a visit. Care home staff coud wait for anything up to 12 hours for an out-of-hours GP to visit from the first contact with the service: S3CH2 care home staff focus group: But with the doctor ooh you can wait up to a night ong. Sometimes if you phone them at unchtime it can be five o cock in the morning sometimes. Interviewer: So what s the shortest ength of time before somebody wi visit? S3CH2 care home staff focus group: 4 hours. Interviewer: And what s the ongest? S3CH2 care home staff focus group: About 4 12 hours. They say 4 6 if someone is on end of ife. Some out-of-hours services were perceived as having negative attitudes towards care homes and staff. Exampes were given of do-not-attempt cardiopumonary resuscitation paperwork being ignored on occasion, and there was a perception that care home residents were a ow priority for out-of-hours and emergency services. It coud therefore mean that an out-of-hours service that was perceived as responsive was more ikey to be used by care home staff. Hospitaisation and ength of stay Prevention of hospitaisation was a priority for HCPs but the process of hospitaisation, ength of stay and faciitation of discharge did not feature in the HCP staff interviews and discussions, despite prompts in their interviews. The absence was striking. Care home staff, in contrast, were more preoccupied with how admissions and discharge were managed and the faiure of hospitas to ook after their residents or pan ahead. Hospita staff were perceived as having itte dementia expertise and no insight into how care homes might work or the difference between care homes with and without on-site nursing. The consequences of this ack of communication and common understanding were unpanned or unannounced discharges back to Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 95

132 PHASE 2 CASE STUDIES: CASE STUDY OUTCOMES OF INTEREST the care home, with no communication about the resident when they were in hospita or prior to discharge, and discharges at times when senior staff were not avaiabe. This posed particuar probems, as staff recognised that admission and discharge from hospita were often associated with deterioration in a resident s condition: Interviewer: OK, and what about when somebody has been in hospita and gets discharged back to you, what s that ike, how does that work? S1CH1 care home staff focus group: We we check them for bed sores (aughs). Resident, carer and staff satisfaction Genera practitioner services Where GPs hed reguar cinics in care homes, as in sites 1 and 2, the eve of satisfaction with the services provided was generay high. The highest eves of satisfaction were in site 1 where care homes 1 and 2 described their GP services as exceent and care home 4 reported that GP services were hugey improved as a consequence of the introduction of singe GP practices working with care homes. Reguar contact appeared to faciitate both working and interpersona reationships as we as GPs knowedge of residents heath-care needs. Care home staff highighted continuity of care and the opportunity to buid rapport as being key to satisfaction when working with GPs. The importance of a good reationship with the GP in the provision of residents care was highighted by one care home manager: We ve aways had a cose rapport with the GP surgery so we coudn t do what we do without them to be honest. S2CH2, manager In site 3, where GPs visited on request and speciaist nurses often substituted for them, there was a marked contrast, as opportunities to estabish reationships were imited and, correspondingy, eves of satisfaction were much ower. Care home staff focused on the difficuty of getting GPs to visit and the knock-on effects such as the high eves of over-the-phone prescribing. As the foowing quotation demonstrates, if the care home manager was not satisfied with the nurse practitioner empoyed by the GP practice, then she woud request a GP visit; substitution did not necessariy ead to reduction in demands on GP time: We if I was phoning doctor this morning for somebody, there s a good chance I get a nurse practitioner rather than a doctor but, at the end of the day, if that nurse practitioner doesn t fee happy with who she s seeing, then I get a doctor ater on. S3CH3, care home manager In S3CH1, access to GPs was further reduced over the course of the study as a resut of their reguar GPs being repaced by ocums who were reuctant to visit the care home. This increased the care home s reiance on speciaist nurses. Satisfaction with GPs was owest in S3CH4, where some GPs showed a ack of awareness of care home staff or their working patterns: Doctors just come stroing in ike they re god s gift sometimes sti and you know, they want to see the nurse and they want to know this and it doesn t matter if it s unchtime, tabet time, you have to accommodate them and it is just sometimes you want to say hang on a minute... S3CH4, care home manager 96 NIHR Journas Library

133 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 The benefits of reguar cinics on reationships and the quaity of care were evident to staff when comparing the different experiences of two care homes. In S1CH4, care home staff were dissatisfied with the quaity of service they received from GPs who visited ony on request, as such visits tended to be rushed:... because sometimes they [GPs visiting on a resident-by-resident basis] want to assess them [residents] so quicky... S1CH4FG Staff in S3CH2 had the contrasting experience of being abe to compare the impact of the introduction of a monthy GP cinic for residents as part of a piot scheme with their usua on-request GP service. Reguar cinics were perceived to be a good service because of their focus on residents we-being and the proactive approach to residents heath care: Monthy GP cinic is good because they [GPs] are taking an interest in their [residents ] we-being and their patients, and they have check-ups every month. You might get somebody go for ages and be fine on the inside but when they come round and do checks [they find a heath-care need]. S3CH4FG However, it did not resove the probem of not obtaining residents prescriptions prompty. Other services There appeared to be a simiar pattern in terms of overa satisfaction with other heath-care services, in that sites 1 and 2 were focused on what worked and site 3 showed the owest eve of satisfaction. Comparing types of care homes, the nursing home staff in sites 2 and 3 appeared to be the east satisfied with heath services. In S2CH4, nursing home staff perceptions were that they had itte access to heath-care services in genera and in site 3 the nursing home staff strugged to identify a service with which they were satisfied, with the exception of the chiropodist, optician and pharmacist. However, when care home staff deiberated their satisfaction with specific services there were wide variations, in some cases for the same services, within a site. In site 1, two contrasting accounts were given by two different care homes about the care home team; one was extremey positive, especiay in reation to the case management of new residents, and the other depicted staff as not knowing the residents and having itte knowedge about dementia. In most cases it was individua HCPs that they found difficut to work with rather than the service. For exampe, staff in S1CH4 were of the opinion that some district nurses coud earn from the dementia team or the GPs. In sites 2 and 3, two care homes reported that district nursing services were outstanding, but two care homes with on-site nursing did not receive district nursing services. It was recognised that it took time to deveop reationships; the manager in S3CH1 recounted that, as their reationship with the district nurses deveoped, so the performance of the care home team had aso improved: District nurses, when I first came here in 2012 there was a fractured reationship... and the reationship then with the district nurses improved, because they coud see the performance improving and the environment had improved and then they were coming on board with things to the point where we ve got an exceent reationship with them. S1CH3, care home manager In S1CH2 overa satisfaction with services was focused on how they worked together as a team to meet residents heath-care needs, a narrative of NHS staff heping out : Interviewer: What is it exacty about them you fee that works we? Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 97

134 PHASE 2 CASE STUDIES: CASE STUDY OUTCOMES OF INTEREST S1CH2 care home staff focus group: I think they a work reay we and I think depending on what services they are, I think they a do their job and hep us out and everything ese and they visit when required so I can t compain that way and obviousy every one of them services a ink together and provide the overa care for the residents here and not one of them don t do their job propery or hep you out. When evauating satisfaction with services in a sites, the speed of access to the service was important, but in sites 2 and 3 reocation of NHS staff and their heavy workoad were factors that affected satisfaction. In two of the care homes, the district nurses high caseoads imited the time they coud spend in care homes and compromised the deveopment of reationships with care home staff. However, the paiative nurse speciaist in site 2 and the dietitian in site 3, who both had an expicit and specified care home component to their roe, were abe to estabish a good rapport with the staff and give them reguar support with residents heath care. Residents and reatives perspectives on their heath and the services received The overarching theme for residents and their reatives across a three sites was the need to be ooked after, or checked over, irrespective of whether this invoved care home staff or HCPs, athough one resident attributed her satisfaction with her heath care as refecting the care received in the care home rather than that received from the visiting professionas: The heath service is marveous, it s not the heath service itsef, it s the peope that work there that cure the patients. S1CH1R04 When asked what was most important about their heath on a day-to-day basis, the majority mentioned having their physica conditions managed and the abiity to engage with ADL. These accounts echoed the phase 1 findings. Most residents did not distinguish between persona care and heath care when taking about good heath care and, with a few exceptions, resident satisfaction with care was high. The ony visiting service that was consistenty referred to by both residents and reatives was the GP, and satisfaction was inked to access, continuity of care and whether or not the GP was peasant:... the GP is aright... he has a sense of humour... no I think they are a reasonabe... S3CH2R06 There were areas identified for improvement: access to dentists, therapists and podiatry; advice and support; and monitoring for specific conditions such as pain secondary to arthritis and diabetes meitus. One reative wondered if a diagnosis of dementia had reduced her mother s access to speciaist care and commented that it woud be different if she had cancer. One need identified by two residents iving with diabetes meitus in different sites was that medication and advice around the management of their condition shoud be inked to how meas were prepared and offered. Apart from this there was a ack of specificity about services received. Some reatives were uncertain about service eigibiity and wondered in hindsight if they shoud have asked for services or chaenged or been incuded in care-reated decisions. For exampe, this reative wondered if physiotherapy services coud have maintained or improved the resident s function and mobiity post admission and regretted not asking:... Famiy [wife] fet that the resident [husband] needed physiotherapy to increase mobiity, sma movements such as eating, and bigger movements... but they have not spoken to anyone about it... and I fee guity about that now reay. S1F02 98 NIHR Journas Library

135 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Heath-care professionas satisfaction with the service deivered to care homes Heath-care professionas across the sites generay showed high eves of satisfaction when working with care homes but there were differences in how HCPs and GPs taked about care homes. Satisfaction was based on the quaity of the interpersona reationships and was experienced when care home staff did as they were instructed by HCPs. Frustration was expressed where handovers between care home staff did not ensure HCPs instructions were communicated and impemented: I think getting the information to start with for my assessment works quite we because there s aways someone there ready to give me the information. Sometimes it doesn t work as we when I ve put a particuar recommendation in pace that needs to be done reguary and it s not handed over we in the care home so the morning shift don t handover to the afternoon shift. So when I go in they haven t got a cue what I m on about so that s quite frustrating reay when that patient needs. For exampe, if I ve advised that the patient needs reguar standing and mobiity and it s not been handed over how far the patient can wak and they re ony doing two steps when in fact they can do 22 steps, they re not carrying out the correct recommendations. That s sometimes quite frustrating because there s such a big staff ratio in the home, if it s not handed over to every singe carer it doesn t get done for one particuar reason. S1CH4HCP04, physiotherapist dementia team The importance of the roe of the manager and the management structure were contributing factors to the overa experience of working with care homes identified in both sites 1 and 2. The paiative care nurse speciaist taked about a high eve of satisfaction when working with care homes, based on mutua trust and reationships that had taken 5 years to estabish. Some HCPs, incuding the tissue viabiity nurse speciaist in site 2 and the dietitian in site 3, noted that satisfaction was predicated on the extent to which care home staff were engaged in working with them. As one physiotherapist in site 3 stated: If care homes are not on board you are amost wasting your time. S3CH3HCP02 Resident behaviour secondary to dementia and its impact on service provision and working reationships Impicit in the phase 1 findings was that the majority of residents had a dementia diagnosis. However, we had not considered or recognised how this affected patterns of working and response between the care homes and the visiting heath-care services. An assumption that primary heath care and generaist community services had the reevant skis and, more importanty, referra networks was not supported in two of the three sites. Across the sites, but particuary in sites 2 and 3, dementia-reated behaviour was identified as the most chaenging condition because of its unpredictabiity and staff not knowing how to respond. In particuar, aggression and repetitive behaviours had an impact at a eves of care and coud adversey infuence the daiy ife of the care home: for the resident, it was the eve of distress experienced; for staff, it caused chaenges reated to knowing how to respond to the individua s needs, but aso to those of other residents affected by the behaviours; and for visiting HCPs and GPs, it affected their confidence in knowing what to do and who to invove by way of additiona expert support. Even when support was avaiabe, referras were either sow in achieving a response or perceived as ineffective; in these situations, it was an unsatisfactory outcome for a invoved. Box 6 provides one account of a situation in which mutipe services were utimatey invoved, but where there was itte evidence of the care being co-ordinated or the key issues resoved. In sites 2 and 3, without access to a speciaist team with care home responsibiities, there were case exampes of poice being invoved and residents being sectioned from the care home to speciaist Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 99

136 PHASE 2 CASE STUDIES: CASE STUDY OUTCOMES OF INTEREST BOX 6 Exampe of escaation of service use in reation to dementia care Resident in care home caing out continuousy Care home staff review with her reative possibe reasons and if she is in pain, in discomfort or distressed by something or someone in the care home. She cas out sighty ess when there is a member of staff avaiabe to hod her hand. Other interventions to distract, provide activities and reduce sources of possibe distress have not worked. Other residents, staff and visitors are increasingy upset by the woman s caing out. GP visits and suggests that the dementia speciaist team visit; a member of the team visits and suggests the community menta heath team, which visits and suggests a visit from the psychogeriatrician. The sequence of referras takes severa months and the woman s symptoms at the time of data coection were sti unresoved. psychiatric inpatient units. Even with access to a dedicated dementia team, the outcomes were mixed and, as the foowing quotation suggests, it needed more iaison with the care home staff about timing and frequency of visits, so that the HCPs coud be present to witness when and how a resident was aggressive: You try a ot of things but they just don t aways work do they? And then Dementia Outreach come back and they try a different strategy but it doesn t aways work. You know ike your [speciaist nurse] can t just come in and ike we say she s got aggression when we re changing her, a persona care, if she s [speciaist nurse] coming in on a morning and that, we reay it s absoutey fine, you know, that s a good day. But then they need to do a few days in seeing this aggression that these peope are giving... S1CHFG Simiary, in site 2, the care home staff gave an exampe of how often the assessment of the issues was inappropriate and did not consider the context of care as we as the presenting behaviours. One bad experience made them reuctant to re-engage with the service or trust the HCPs expertise: Intensive menta heath team: And was it hepfu when they came out? S2CH4FG: Not reay because they re a about, oh, ooking on paperwork, what s this resident been ike... rather than go and ook at the resident s heath, yeah, it s ony happened once, this has ony happened once, yeah, but once is enough, yeah. Site 2 provided another exampe of a situation not resoved. This ed to a safeguarding referra, which was subsequenty handed unsatisfactoriy and resuted in a forma compaint from the care home manager. Ony then was there a satisfactory response. This was the ony exampe of care home staff formay compaining about the quaity of the service provided by the NHS: Yeah. We ve had to put a forma compaint in before because they woudn t safeguard this resident and we tried desperatey to safeguard him. His behaviour was getting a ot worse and in the end we had to put a forma compaint in to the menta heath team. And it s funny how quick they came out, the psychiatrist, doctor and sectioned the guy because he was just getting even worse. S2CH2FG Not ony were behaviours probematic, but a ack of dementia knowedge and ski was shown by some HCPs, for exampe dentists expecting residents to visit the surgery when care coud have been deivered in the home. 100 NIHR Journas Library

137 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 The care home staff beieved that they had received adequate training in dementia, so they did not see that as the soution. In compex cases or situations when it was not cear why the person was distressed, staff wanted practica and expert support. Across the sites, care home staff taked about their frustration at being asked to compete behavioura charts for weeks, the purpose of which they did not understand, rather than working together with the speciaist teams and the resident. GPs were perceived as doing itte to manage a behavioura crisis other than referring to menta heath services: I ve had residents who have been, got themseves on the foor and have decided they re going to start kicking cupboard doors and everything ike that and the GP doesn t even ook at them. S2CHFG This observation was substantiated by GPs in site 2 who highighted deaing with aggression and vioence in peope iving with dementia as the most difficut aspect of their work in care homes. Cross-case comparison To deveop and refine the programme theory from phase 1, we wanted to know how HCPs interpreted their responsibiities to work with care homes. Specificay, what do the care homes require from them, who is the focus of their care and what infuences how they provide their services? In our interviews, we asked HCPs about activities that might support (or not support) reationa working. We ooked for evidence of the ways in which heath-care services were responsive to how care homes worked and what were their priorities and patterns of working. We considered the extent to which service deveopment refected co-design or a shared view of practice. Severa common narratives emerged from the HCP interview transcripts that suggested different mechanisms were at pay (Tabe 30). In site 1, there appeared to be an ethos of HCPs working together with care home staff and a view that they were a there for the same reason. This was expained as having deveoped because of infrastructure and resources that were designated for care homes and patterns of working that had evoved over severa years. These had evoved as HCPs had earnt to work with care homes within the context of a specificay commissioned care homes service. Individua practitioners appeared to share an understanding that to work with care homes they had to adjust how and when they visited, and that there was care home staff expertise that coud inform their discussions and practice when working with residents: Interviewer: What sort of things does the contact with the care home staff invove? S1CH2HCP03 occupationa therapy community rehabiitation team: So it woud be speaking to them to find out, you know, what, how they perceive, what, you know, the probem, or what the issue is, you know, find out from their point of view what s working and what isn t working, and then I woud, you know, I obviousy do my assessment and do kind of, normay do ike a demonstration with carers. Provide education and advice on, you know, why I m recommending something. I woud spend time ooking at the care pan and advising on, you know, maybe what ese needs to be incuded in the care pan, or updated foowing my visit. You know, education on different risk assessments that can be used to make sure that they ve got the right equipment in pace. So you know, for exampe, ike the use of ike the Braden Pressure Care Too, making sure that they ve got that, that they understand how to use that. Yeah, I think that s it. In site 1 there were numerous exampes of HCPs engaging with care home staff in an appreciative way. During interviews, HCPs frequenty highighted good practice from care homes without prompting. This contrasted markedy with more negative portrayas of care homes depicted within the transcripts from sites 2 and 3. Care homes we have issues with was a common phrase used by HCPs when discussing their working reationships with care homes in these sites. More HCPs in site 3 and, to a esser extent, Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 101

138 PHASE 2 CASE STUDIES: CASE STUDY OUTCOMES OF INTEREST TABLE 30 Main narratives across the sites arising from HCP interviews on working with care homes (n = 43) Site Language of working together for residents, a being there for the same reason HCPs recognise the importance of focusing on care home staff s perceptions of the issues HCPs pan their time to tak with the staff; they recognise the ora cuture for sharing information and care panning HCPs tak about achievabe goas for care home staff and working with residents Numerous exampes of working with care homes recognising care home staff experience, highighting good practice and interventions to de-escaate situations when residents were distressed AHP, aied heath professiona. Frequent references from HCPs to care homes they have issues with Narrative of having to fit visiting residents in with other work demands; few exampes of meeting with care home staff apart from resident specific visits HCPs threshod for support and care home staff s need are not agreed, frequent references to inappropriate requests and ca-outs Expectations differ between care homes and HCPs as to what care home staff shoud know and do HCPs think that care homes need to improve their quaity of care HCPs not seen as integra to this process Working with care homes presented as a chaenge (apart from one AHP) Represented residents and oder peope iving at home as eigibe for the same services. Did not differentiate between care home residents and other patients on the caseoad Exampes of community nurses taking a prescriptive approach to residents care rather than engaging with care home staff in the care panning process Some HCPs recognised care home staff expertise and knowedge of residents needs and their roe as mediators for residents Perception that working reationships with care home staff were faciitated if care home staff had previous NHS experience or knowedge in site 2, made derogatory comments about the standards and quaity of care provided in care homes. They described approaches to working with care homes that were prescriptive about the range of services and support that they woud and woud not provide, and this was presented as a strategy to manage what was seen as uncontroabe demand. There were two notabe exceptions in sites 2 and 3 where practitioners described a pattern of working that was not resident dependent, invoved care home staff in discussing care and offered training and support to staff, and where a HCP assumed responsibiity for iaising with the GP and referring residents to speciaist services on the care home s behaf. One was a paiative care nurse speciaist (site 2) and the other a dietitian (site 3). Both practitioners described their persona interest in the support of care home residents, had worked for severa years with care homes and knew (and were known within) the oca care economy and working with care homes was a designated eement of their work. Their choice of anguage and descriptions of how they worked with care homes, exampes of where care home staff had achieved good care and who they iaised with within the care home and within the oca care economy on behaf of the care home, were simiar to the HCP accounts in site 1. Regardess of service deivery mode or the infrastructure of care, there were common experiences that HCPs highighted around access to services, eectronic systems and sharing of information (Tabe 31). In site 1, the MDTs were meeting away from the care homes, but it was uncear how this affected residents care. Across a sites the opportunities for the different services to work together for the care homes were imited. 102 NIHR Journas Library

139 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 TABLE 31 Consutations and services Consutations and services Nurses and therapists consutations with residents were arranged by referras on request Cinics were not run by speciaists (reguar consutation sessions) in the home apart from one care home nurse speciaist in site 3 and the GPs in sites 1 and 2. Mutidiscipinary working imited to individua residents or meetings away from the care home that did not invove care home staff Optician and podiatrist worked directy with the care homes; visits 6-monthy or yeary to see a residents Domiciiary dentist in site 1 worked in the same way, but in site 3 the community dentist visited ony on request by care home staff, reatives or residents (it was not possibe to access a dentist for interview in site 2) Paperwork and eectronic systems There was no shared paperwork with care homes, but most services woud record information in the HCP section of residents care home notes or annotate care pans Eectronic systems, mainy SystmOne (The Phoenix Partnership, Leeds, UK), were used in a three sites by most HCPs for referras, recording consutations and assessments; however, not everyone was on the same system. Where nurses and therapists were using SystmOne, ony skeeton notes were eft in the care home so care home staff did not have a record of the HCP consutation Most HCPs coud access each other s information on residents if they or their reatives had given their permission Some GPs accessed information from other HCPs, but did not share their information Dedicated care home services in sites 1 and 3 worked office hours Concusion In summary, between sites, resident and reatives satisfaction with care and service use was not strikingy different. There were differences in the abiity of services to engage with the issues and care needs of peope iving and dying with dementia and questions about the different reasons and type of residents hospitaisations across the sites. Chapter 7 wi use these findings, and the detaied service descriptions outined in Chapter 5, to revisit the findings and emergent programme theory and possibe CMO configurations of phase 1 and refine these in ight of the phase 2 findings. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 103

140

141 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Chapter 7 Using the phase 2 case study findings to refine context mechanism outcome configurations and the expanatory framework Introduction Phase 1 (described in Chapters 3 and 4) mapped the range of NHS provision to care homes. Based on stakehoder views and inked evidence synthesis, we suggested different contexts and possibe mechanisms that supported care home residents access to heath care. This chapter revisits these arguments in ight of the findings from the case studies in phase 2, as described in Chapters 5 and 6. Based on a cross-case anaysis of recurring patterns within the phase 2 findings, it sets out a series of CMO configurations that chaenge and refine the expanatory framework of phase 1. Phase 1 posited that within the different types of service provision it is how and if a sense of common ground is achieved between HCPs and care home staff that is important. The activities within an intervention that we identified as key were those that aigned heath-care provision with the goas and priorities of care home staff, and aimed to buid reationships between care home staff and visiting HCPs. Activities, such as discussions, before setting up a service to identify shared chaenges and priorities, use of shared protocos and guidance and reguar meetings were the contexts necessary to generate mechanisms of co-working. Box 7 summarises the emergent programme theory that informed the case study recruitment and structure for data coection and anaysis. Mechanisms of successfu programmes are characterised by activities that provide visiting HCPs and care home staff aocated time together for discussion and refection and which aow reconfiguration of the intervention to match care home workfow and priorities in different care home settings. Contextua infuences, such as financia incentives or sanctions, continuity of contact and evidence-based approaches to assessment and care panning, are needed to enabe these mechanisms/staff responses to occur and to achieve improved resident and staff outcomes. BOX 7 Phase 1 emergent programme theory Interventions (whether or not they use sanctions and incentives, speciaist practitioners or care home-specific resources) are more ikey to achieve the outcomes of interest when the activities: trigger the engagement of care home staff from the outset create opportunities for heath-care and care home staff to work together structure the intervention to fit with the priorities and working practices of the care home as an institution. NHS interventions that use processes and approaches to working that reconcie competing priorities between care home staff and visiting heath-care staff wi engage the interest of the care home staff and generate opportunities for shared panning and working. NHS interventions that provide and fund visiting HCPs, and that recognise that engagement with care homes at an institutiona eve is important, increase the potentia for engagement across organisations and the association and engagement with care home residents and staff when there is carity and an agreed understanding of each service s roes and responsibiities. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 105

142 REFINING C M O CONFIGURATIONS AND THE EXPLANATORY FRAMEWORK Achieving common ground The cross-case anaysis broady supported achieving common ground as a mechanism that supported effective working between heath-care services and care homes. In site 1, with funding of care home teams and GPs to work with care homes, the narrative was one of co-operation, highighting best practice in care homes and finding shared heath-care soutions. This was repicated in specific instances in sites 2 and 3, where individua practitioners were empoyed to improve end-of-ife care and nutrition in care homes. As the foowing quotation demonstrates, there was an appreciation of each other s skis and evidence of coaboration:... So you know, kind of educating staff on how to, you know, appy the sing on the stand aid [referring to hoist] more effectivey and more efficienty. One of the, one of the carers in particuar was exceent, because the person that I went, the resident had quite advanced dementia, and the carer was particuary good at communicating and putting the resident at ease. In that, in that case I did, I wrote an e-mai to the manager of care home 2 and compimented the particuar carer, because she was very good. S1CH2HP03, occupationa therapy rehabiitation team... It worked and it worked reay we, the feedback was positive and then we roed it out... But because we, because we go into the homes, then it s not just a trainer deivering a training session, they coud come back and say, You know what you said on the course or coud you expain more. I didn t understand about this and can we go through it again? So, there s open communication with them. And they, at that one (care home), they a, I m aways greeted with a smie, I m aways greeted with, Oh reay nice to see you again, where ve you been? But it is a case of they fee confident that we have got a two-way communication and they fee that they can open up to me and I can open up to them as we and just to see that it is an open communication and that s what works we there and... And the staff are consistent, which aso works, where if you have got a pace where they have not got consistent staff, then you re sort of going over the same things a the time... S1CH4HP02, speciaist nurse care home team What was evident as a shared narrative across the care home teams and some of the practitioners in site 1, was echoed in individua practitioner accounts in sites 2 and 3 from the paiative nurse speciaist and the dietitian, respectivey: S2CH4HP01, paiative nurse speciaist: Yeah, so I woud say I aways have contact with the care home staff. Many patients do not have fu menta capacity so care home staff hep with assessing of situation, feedback and education is given to the care home staff on drug management, symptom contro, psychoogica and spiritua support. If a new member of staff is there, I try to encourage them to work aongside me so I m teaching them how to manage without me... the abiity of the staff to communicate we with famiies of patients and MDT. Researcher: Yes, so they can sort of earn from each other s experience. S2CH4HP01, paiative nurse speciaist: Yeah... if there s a care home that we know that the chef is reay invoved, we get them to come in and tak to the catering staff in other care homes... So some of our training is aimed at the care staff but some of it is aimed at the catering staff so that we can try to get things going that way. Researcher: OK, so it s sort of brought some... It s encouraging them to think about how to do that with their residents as we? S3CH2HP01 dietitian: Yeah, and with the newsetter that we send out... we woud put up areas of best practice as we so if we see something in a care home that we think is reay good, ike the midnight snack menu, we wi highight that on the thing so that other care homes can think about it as we. 106 NIHR Journas Library

143 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 When these activities did not happen (because of ack of funding, carity of roe, staff avaiabiity or time to meet), both NHS professionas and care home staff strugged to find an agreed way of working together. In particuar, under such circumstances, NHS professionas did not adapt how they worked with care homes to accommodate the different residents needs or care home and NHS staff patterns of working. This was most evident in site 3 where, in comparison with the other sites, response to residents needs and NHS services was observed to be reactive and unco-ordinated with changing eves of service invovement. Findings from phase 2 identified further contexts that informed the way in which NHS and care home staff worked together that generated mechanisms of mutua trust and wiingness to work together. These were: earning and working wrap-around care for frai oder peope iving and dying with dementia. Learning and working Commissioners and HCPs needed to deveop a anguage for an understanding of the compexities of supporting frai oder peope in non-medica, non-nhs settings. It took time for those NHS services funded to work with care homes to adapt and become embedded as a recognised part of the wider heath-care provision to oder peope. The three case study sites and the services within them represented a continuum of experience and intensity of association between the NHS and care homes. The case studies suggested that, in addition to activities that fostered opportunities for coaboration, if commissioners and services had pioted and triaed different ways of working together, this increased confidence and trust between the services. Site 1 had a ong history of earning how to work with care homes and had either retrained NHS staff to work with care homes or staff had moved between services, taking their care home expertise with them. Site 2 had begun to invest in staff, changing how often staff visited care homes and expanding their range of activities. Site 3 had recognised the need for transformation but had invested in singe practitioners and some GP piot cinics to achieve specific outcomes rather than system change. Staff from the site had, for exampe, met with care homes to discuss how to reduce the number of ca-outs to the ambuance service for residents who had faen. As the foowing commissioner quotation demonstrates, site 3 was focused on reducing unpanned admissions from care homes but was sti earning about the care homes in the surrounding area, who the staff were and how the care homes were run:... So our focus is predominanty on admission avoidance themes predominanty, and obviousy care homes is one that aways comes up as a significant issue for us in terms of the eve of demand that they create, we ve got a huge number of care homes in our ocaity, so we ve done some various things to try and improve our working reationship with care homes. We ran a piot with GPs ast year... we are just at the reay eary stages of trying to do that [work with care homes], so at the moment we do not have one definitive ist for a care homes in the area, but I m not sure that we have got sort of one ist with a of our contacts for our homes that we can send something out to because we have got so many. We have got over 200 care homes in the area and obviousy there s some sma independent companies that run homes as we as the bigger common ones, so it s difficut to make contact with them a. So what we are starting to do is we are trying to buid a ist of those that we sort of engage with. But aso we re working very cosey with our County Counci in terms of they re setting up some forums where they have invited a the care home managers to, and it s awayof sharing information and getting them engaged, so we re working with them to attend those and having a bit of joint approach reay... it s the fact that I do not think we have ever had care homes round the tabe before when we have come up with schemes you know. Every winter we tak about what schemes we need to put in pace to support demand but we ve never reay had care homes as part of those discussions. DS500042, commissioners interview site 3 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 107

144 REFINING C M O CONFIGURATIONS AND THE EXPLANATORY FRAMEWORK Where patterns of working had evoved over severa years, practitioners who were interested in working with care homes were supported by the commissioning organisation to deveop their work. They were abe to absorb or address what other practitioners had described as chaenges. Continuity of HCP and team input, being accessibe, responsive and wiing to provide education and training coud mitigate the effects of staff turnover in the care homes and support coaborative working. Such patterns of shared understanding and mutua professiona deveopment provided a patform through which more structured innovations coud be introduced (such as faciitating end-of-ife care training and nutritiona assessment, risk assessment or guideines on when to ca out a GP). Care home staff tended to be more confident and coud expect to receive ongoing support (as opposed to time-imited or singe issue input) as part of such initiatives. Consequenty, they were ess threatened by interventions designed to improve resident outcomes. Equay, NHS staff were abe to take the ong view, inking work that supported individua residents with wider initiatives to improve care for a residents. When NHS professionas did not modify how they worked to acknowedge differences between care home residents and oder peope iving at home or in hospita, this triggered a sense of frustration among both the NHS service provider and care home staff that residents did not fit with the service as offered. Such frustration was compounded when: NHS staff defined their work soey by individua resident encounters and were reuctant or had imited incentives to engage with care homes as an organisation visiting NHS staff fet that they acked particuar expertise in the care of those iving and dying with dementia the purpose or desired outcomes from working with care homes had not been identified or agreed by the commissioners in discussion with NHS practitioners invoved or care home staff. An observed consequence of some or a of the above was that a ack of constructive contact between the visiting NHS staff and care home organisations/staff coud become sef-perpetuating. NHS staff connecting primariy with individua residents, rather than with care homes as organisations, described a reuctance to engage with care home staff because of fears of being overwhemed by requests to visit. Where NHS staff either fet coerced to visit or perceived that the number of care homes they were working in was unmanageabe, there was an observed resistance to work with care homes to provide training and education to staff. Funding of posts and time to increase the contact NHS staff had with care homes was necessary, but generated greater impact when they created the opportunities to work together. When (because of experience, an interest in working with care homes or an interest in peer-to-peer earning) NHS staff had an expicit commitment to working with care homes, this triggered responses that emphasised the vaue of the work. Tabe 32 provides a revised CMO that expains how investment in additiona resources and services for care homes can be theorised in certain circumstances to achieve outcomes of care home staff satisfaction and residents access to heath care. In phase 1, the review of patterns of working with care homes and the reaist synthesis focused on particuar services, for exampe GP provision, care home teams and specific care home practitioners, incuding pharmacists and therapists. The reaist synthesis addressed the cross-cutting earning from different CMO configurations within particuar approaches to heath-care deivery. Phase 2 findings identified how the presence or absence of other services around the care homes aso affected residents access to heath care and specificay speciaist services. To understand what worked when and in what circumstances it was important to know to what extent NHS provision to care homes inked to other primary care-based services and the oca hospitas. Phase 2 found no measurabe difference in costs per resident across the three sites, but identified that the way in which resources were aocated and organised (support of singe practitioners or teams, care home 108 NIHR Journas Library

145 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 TABLE 32 Revised CMO of the impact of investment in NHS services on resident and service outcomes Mechanism Context + Resource Response = Outcome NHS services funded to visit care homes on a reguar basis NHS services staff working for care homes see this as a egitimate use of their time and skis Practitioners wiing to work with care homes, vaue their work and find ways to provide a package of care that supports residents and care home staff Services engage with care homes and residents have access to speciaist services Length of time the service staff have been working with care homes/have known particuar staff and care home routines Staff deveop ways of working and communicating with care home staff (both forma and informa) and are wiing to be accessibe and fexibe Services visit at times that fit with care home routines Number of care homes staff work with that are seen as manageabe Care home staff concerns about individua residents are described as being addressed before they become a crisis NHS services care home responsibiities understood and accepted. The roe has been through severa iterations Focus on residents access to heath care (not just prevention of admissions or monitoring) Aocation of time and resource and to work in care homes Wiing to engage in proactive care and discuss residents with vague or uncertain symptoms Vaue the work Note Care home working within a system of care: achieving wrap-around care for frai oder peope. Resident crises and GP ca-outs because of staff concerns are seen as either being reduced and deat with by care home staff or accepted by visiting NHS professionas as reasonabe use of the NHS services focus or not) triggered different responses from the NHS and patterns of referra. The structure of support either brought care homes into the economy of heath care with forma methods for referras and inking services and practitioners, or provided episodic outreach from heath care to care homes. This infuenced the abiity of the service to respond to residents with compex ongoing needs or who needed support from more than one professiona group. Prevention of hospitaisations and quaity monitoring were major preoccupations of a the heath and socia care commissioners and these were seen as important in a the sites. However, where this was the ony focus, it coud have an isoating effect on services and visiting HCPs. It aso risked short-termism in how NHS services were organised. They were under pressure to fix something or avert a crisis, both of which were difficut to sustain, especiay when singe practitioners were expected to achieve outcomes that were, often, consequences of patterns of practice across the wider heath and socia care economy. In two of the three sites, there were exampes of intensive interventions through which GPs were funded to improve medication reviews and reduce hospitaisations that had ended when the practitioner eft or the funding ceased. This approach miitated against forming working reationships both with care home staff and, just as importanty, other NHS services that had inks to care homes. Where differences between heath-care resource use and costs at sites were demonstrabe, it was for the number of GP consutations, which tended to suggest that these were both more frequent and more expensive in site 2, and use of acute hospita beds, where there was a trend towards increased use and increased ength of stay in site 3. Focusing service provision primariy around GPs did not reduce costs and did not reduce acute care utiisation. Site 1 had an infrastructure around the care homes that was characterised by a network of NHS teams and a reativey oose connection with GPs. Site 2 used shared data sets and deveoped services that coud ink together around the care homes. Being abe to refer residents to coeagues in a care home team or to Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 109

146 REFINING C M O CONFIGURATIONS AND THE EXPLANATORY FRAMEWORK other speciaist teams with a care home-specific responsibiity meant that residents had the potentia to be hed within a framework of referras and expertise. In these circumstances, care home staff appeared ess ikey to seek hep on an ad hoc basis from outside the supporting infrastructure. This was important, for exampe, when care home staff were unabe to manage or resove residents behaviours that they found chaenging. In such situations, the abiity to co-ordinate the care of the resident within the caseoad of interconnected speciaist practitioners who visited the home miitated against the tendency to ca for hep from outside services when things went wrong: Interviewer: OK. How do your visits work when you are at this care home, so this particuar care home? S1CH1HP02, dementia team speciaist nurse: Right, at this care home I arrange an appointment as usua and I either speak to the manager or a senior carer and we wi discuss whatever pan we ve put in pace, if it s being effective, I go and tak to the patient. And then of course if the pan s been effective and it s usefu and heping them we continue with that and if not we discuss how we can change and modify it. Then we come back, tak to other members of the team for advice for us or we can refer to, because we re a mutidiscipinary team we ve got physiotherapists, occupationa therapists, support workers and so we can refer on if there s any physiotherapy needs, occupationa therapy needs, whatever, get support from support workers and we can do that so, and we ve got consutant time as we, so that s broady how the visits work. The wrap-around effect of a range of oder peope-specific services being avaiabe to care homes, either through a forma infrastructure or informay through the connections of particuar practitioners working on the residents behaf, heped to co-ordinate residents care. This coud aso have other beneficia effects. Residents whose heath was deteriorating woud be referred earier for assessment, even if this was initiated by the NHS services. Care home staff coud enjoy better access to education and training from speciaist services. Investment in care home-specific services and GP time was an important underying contextua factor in securing care home residents access to heath care for probems besides urgent care needs. The quaity and consistency of access to the service that this investment prompted, however, was aso inked to how we the referra systems and the different practitioners contributions were known and understood by the care home staff. If this was not understood, diverse providers coud trigger mutipe referras or, when uncertain, a defaut response by care home staff of referring to the GP. In these situations, the outcomes and benefits of having the avaiabiity of a range of oder peope-specific services for care homes were reported more negativey or cautiousy by care home staff. Care home staff found it difficut to contro the number of NHS services visiting them. At times, they coud perceive this as a form of surveiance or covert quaity monitoring: Different bodies. Different, you know, sometimes we can have three different professionas come in to see one person, you know and it s a bit too much. I think care home ife is ike traffic, road traffic, everybody just coming and going, and demanding, and if you do not do this then the next thing you know you re in Safeguarding. S1CH1, care home manager The roe of the GP was important in a sites, even when services had been deveoped to suppement or substitute for care homes access to GP services. In site 1, it appeared that GPs were not overwhemed by the demands from the care homes. This was possiby for two reasons. First, other NHS-funded services predicated to support care homes absorbed those referras that GPs at other sites regarded as inappropriate or trivia and as resuting from staff inexperience or anxiety. Care home staff had permission and the opportunity to seek advice and hep from the services that visited about mutipe residents or probems that were not patient specific. Second, the GPs had responsibiity for fewer than three care homes. There was no 110 NIHR Journas Library

147 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 evidence that GPs withdrew from care in site 1. The number of GP contacts was not ower than in site 3, even after controing for the greater comorbidity and dependency at site 1. Sites 2 and 3 had GPs who received extra funding to visit care homes and, as part of these arrangements, in site 2, practices took over primary responsibiity for care homes whose residents had previousy been registered with other practices. They offered cinics in the care home, but the activities they described were very simiar to how they woud see patients in their own heath centre and at home. Ony in site 3 was greater attention (for the short time the service was funded) paid by GPs to medication reviews. This may have had an impact on prescribing as a consistenty ower ACB score was recorded for residents at this site. The issue for many of the GPs, particuary in site 2, was that working in care homes confated urgent and panned care. This required a shift in approach that they either coud not or were not prepared to make, ostensiby because of the commitments associated with their wider caseoad. As the foowing quotation aso shows, this moved decision-making about what is urgent care from the GP to the care home staff: Interviewer: Is it that you are saying there is dissonance between what you think you shoud be doing as a GP and what you are doing when you are in a care home? S2GP01: I am not sure that I woud ca it dissonance, I think that it is not good use of my time, care homes are probaby one of the ony paces where the GP is having to do urgent and panned work in the same pace and at the same time. When you see other patients it is a booked appointment or an urgent ca, there is some panning and anticipation it is harder to repeat that in care homes. I think that is an important difference. When I go in to see a patient then I can find mysef deaing with what the care home staff think are urgent cases. We were unabe to expore how NHS practitioner caseoad affected patterns of working, athough there were references in sites 2 and 3 to the difficuties of prioritising care home work over other responsibiities. The negative accounts about GPs carefuy managing their contact with care homes suggest that there may be a tipping point beyond which adding additiona responsibiities or roes to aready overburdened practitioners may become sef-defeating. Care home ownership and the eve of investment in training and education of care home staff may aso have shaped demand on NHS resources. In the interviews, NHS practitioners recognised that particuar staff were easier to work with than others, particuary if they were quaified nurses, but they did not attribute this to who managed or owned the care home. Tabe 33 provides a revised CMO that expains how commissioning mutipe NHS services to work with care homes on a reguar basis, incuding those with dementia expertise, where the referra networks are expicit may TABLE 33 Context mechanism outcome care home working within a system of care Mechanism Context + Resource Response = Outcome Mutipe NHS services staff commissioned to work with care homes on a reguar basis Known referra network, incuding services with staff who have dementia-specific expertise Referra systems for residents needs (e.g. dementia, fas prevention) Infrastructure supports review, feedback and opportunities to change patterns of service deivery Practitioners confident that they can provide or access services for residents and know the care home staff they work with Services engage with care home staff and residents have access to speciaist services for the support of peope with compex needs NHS services staff know other services because of the ength of association and stabiity of teams or structure of provision, or through staff moving between services Staff with other responsibiities to other patient groups have capacity to work with care homes Referra systems are cear to care home staff Reduce the need for hospitaisation Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 111

148 REFINING C M O CONFIGURATIONS AND THE EXPLANATORY FRAMEWORK ead to outcomes incuding service engagement with care home staff, access to speciaist services for residents and the reduced need for hospitaisation. Living and dying with dementia A cross-cutting theme in the case study sites was how the oca NHS responded to the heath needs of residents with dementia. Access to a inked dementia outreach team or speciaist expertise aowed both rapid responses to crises and the provision of staff training and support: Interviewer: You have described the chaenges of working with care homes. Are there any issues in terms of resident heath-care needs that you woud say, is compex, and stretches you as a GP? S2GP8: I think physica vioence is the biggest issue that I have to dea with and how to best manage that. Interviewer: With peope iving with dementia? S2GP8: Yes, it can be very difficut to manage and to know what is best for the person and sti think about the needs of the other residents. Symptoms of dementia that are not resoved. That has been an increasing area of my work. S2GP7: I agree, that and safeguarding issues and deprivation of iberties, you can get drawn into that and that is quite difficut, who to work with and how to resove it. The responsiveness of the oca NHS to residents with dementia was a key mechanism in securing residents timey access to services and the identification of dementia-sensitive soutions to minimise distress that coud be supported in the care homes. This was, however, a resource that was separate from other care home-focused services, and the reviews of medication did not suggest that any of the study care home staff were more skied than others in avoiding or reducing the use of antipsychotics. The probems that coud arise when there was separation of dementia expertise from other sources of NHS support were most evident in the accounts of residents whose behaviours and distress coud not be managed by care home staff, visiting primary care services or crisis-response menta heath teams. The narratives in these situations were those of visibe need, unco-ordinated responses, escaation of service invovement and demand on emergency services. The case study findings ed us to hypothesise that resident and organisationa outcomes woud improve where there is ongoing access to dementia expertise within the care home support services (Tabe 34). TABLE 34 Context mechanism outcome of iving and dying with dementia Mechanism Context + Resource Response = Outcome Care homes and visiting NHS practitioners have ongoing access to dementia expertise Abiity to provide training and support for the care of peope iving and dying with dementia Visiting practitioners and care home staff have a shared ski set to draw on to support peope iving with dementia Reduced use of antipsychotic prescribing Expertise in dementia care a prerequisite for working with care homes Range of resources and skis to anticipate and moderate the signs, and symptoms, of dementia that cause the resident distress, and address care home concerns around risk management and deprivation of iberty Visiting heath-care staff are confident when providing care to peope iving and dying with dementia Care of peope iving with dementia whose behaviours staff and residents find chaenging is managed within the care home 112 NIHR Journas Library

149 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Summary The phase 2 findings broady supported the expanatory framework proposed in phase 1 around mechanisms that promoted reationa working between visiting NHS professionas and care home staff. When NHS professionas, incuding GPs, did not receive dedicated funding for working with care homes, but were sti required to visit care home residents across the sites, this triggered responses of stress, resentment and frustration, and resuted in unco-ordinated care for residents. For GPs, in particuar, care home work was difficut to manage within existing caseoads. Estabished patterns of working with oder peope that were perceived to be successfu in other settings were difficut to repicate or sustain in care homes. As a consequence, HCPs and GPs were more ikey to emphasise care home staff shortcomings, the unacceptabiity of being asked to see urgent cases when making panned care home visits and the beief that care home staff shoud be better trained (though not by HCPs and GPs). The additiona resources needed to sustain NHS input to care homes shoud aim to buid over time a shared recognition and common narrative that care homes are integra to the heath-care economy. A ski set that is reevant to residents with dementia is especiay important here. Carity about how services were organised around and for care homes and their roes in managing different probems were important mechanisms for ensuring that care homes and NHS staff saw each other as vaued partners. It was important that they coud recognise how homes sit within wider care provision for oder peope care. Services that comprised isoated singe practitioners or services that offered episodic contact focused on a singe objective, such as medication review or avoidance of unpanned admissions, were ess ikey to be sustained or understood by care home staff and residents or other services working with care homes. Service staff were motivated when they recognised care homes at an organisationa eve and when engagement with care home managers, staff and structures was egitimised as part of the roe for NHS staff. This was more readiy fostered where service impementation and deivery was undertaken in the context of agreed or known goas. Shared priorities fostered by coaborative working over periods of severa years were aso important in this regard. It appeared difficut to buid, sustain or embed proactive modes of heath-care deivery when NHS services and NHS practitioners did not acknowedge care home staff as integra to how they worked, and when they treated residents as individuas who were simiar to those encountered in their own home. Continuity of care for care home residents was more ikey to be achieved when panned meetings with care home staff and ad hoc conversations were possibe and expected. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 113

150

151 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Chapter 8 Discussion and concusions Introduction This chapter brings together the findings to discuss what needs to be in pace to commission and provide heath care for oder peope iving in care homes. The starting point of this study was that it was unikey that there was one mode of service deivery that woud be effective in a situations. Oder peope resident in care homes are simiar in age and characteristics. However, ong-term care provision and NHS services provided to care home residents are quite heterogeneous. Consequenty, we required a theory-driven expanation of what works when, for whom and in what circumstances. To do this we took a ong view. We considered competing accounts of how to support care home residents and the experiences of past and current initiatives to expore the underying assumptions and supporting evidence. We then set out to test our candidate programme theories of what needs to be in pace to work with care homes by conducting detaied case studies prospectivey across three diverse heath and socia care economies. The work buit on what is aready known about what supports integrated working, 143,149,150 and specificay the findings of the APPROACH (Anaysis and Perspectives of integrated working in PRimary care Organisations And Care Homes) study. 2 This study had described differing and competing priorities and interests of care home and NHS staff that needed to be negotiated to support integrated working between the services. It aso drew on theories of co-production and co-design 151 and socia identity theory to deveop a common understanding of what needs to be in pace to reduce divisions in heath and socia care. 2,4,152 The goa of this study has been to deveop a mid-range theory 153 that, to quote Hedström and Yikoski: seeks to highight the heart of the story by isoating a few expanatory factors that expain important but deimited aspects of the outcomes to be expained. A theory of the midde range can be used for partiay expaining a range of different phenomena, but it makes no pretence of being abe to expain a socia phenomena, and it is not formed upon any form of extreme reductionism. Hedström and Yikoski (2010), 154 p. 61 Specificay, we focused on how the provision of extra resources, or the reconfiguration of existing resources, shaped care home residents access to heath care. We consider now the intended and unintended consequences of emphasising particuar approaches and outcomes and how these infuence the organisation and networks of co-operation with and around care homes. Patterns of service provision Our interviews with stakehoders, review of surveys and review of reviews 4,30,91 have provided a comprehensive account of how heath-care services can and do work with care homes, both in Engand and internationay. Phase 1 found that there was imited agreement in the intervention iterature on how to measure the effectiveness of heath-care provision to care homes or what quaity heath care might ook ike. This ack of consensus was evident in the review of surveys of heath-care provision. Athough there was evidence of the deveopment of care home-specific services, these were the minority and it was impossibe to estabish how many residents they supported. The absence of a nationa minimum data set on the heath-reated characteristics of residents in care homes (as is avaiabe in the USA) makes it difficut to judge the reationship between service provided and needs observed. The wide variabiity in the provision of services to care homes and widespread ack of denta services signaed that erratic and inadequate care was a persistent feature of heath-care provision to residents in care homes. The number of surveys identified Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 115

152 DISCUSSION AND CONCLUSIONS and the consistent nature of their findings, despite their methodoogica diversity, provides a strong argument for the need to move beyond surveying or auditing the status quo. Stakehoder interviews provided overapping accounts of what was necessary to achieve good heath care. These incuded education and training of care home staff, access to cinica expertise, the use of incentives and sanctions to achieve minimum standards of care, the vaue of champions and designated workers working in and with care homes and the importance of activities that buit robust working reationships between the two sectors. Combining this with the review evidence, and an initia scoping of the iterature, the reaist review theorised that it is activities that support and sustain reationa working between care home staff and visiting HCPs that expain the observed differences in how heath-care interventions are accepted and embedded into care home practice. Contextua factors such as financia incentives or sanctions, agreed protocos, cinica expertise and structured approaches to assessment and care panning coud support reationa working to occur. However, of themseves, these measures were ikey to be insufficient to achieve change if they did not ead to visiting HCPs and care home staff working together to identify, pan and impement care home-appropriate protocos for care. This was the expanatory theory that was tested and refined in phase 2. The case studies buit on this to provide a detaied account of how different primary care, community services and secondary care outreach services coud be resourced and structured and with what effect. We seected the three sites on the basis of contrasting approaches and incentives for providing NHS services to care homes. Our seection was not based on how they referred residents between services or how they inked to other NHS services, but our findings indicated that these patterns were an important context for cinica practice. For exampe, patterns of service provision and referra infuenced strongy what happened if more than one practitioner might need to be invoved with a resident s needs. We aso found that these patterns affected the reationships between secondary care, care home staff and visiting NHS services. Our findings corroborated a number of previousy reported observations. For instance, care home residents have variabe and inequitabe access to heath care and mutipe services (GPs, nurses, therapists, speciaists) are invoved in deivering care. Care is often poory co-ordinated and the paucity of strategic panning for care home residents is compounded by imited data about the costs and benefits of the services that are received by care homes. The new evidence presented here is ess about the number and type of services avaiabe and more about how they work together, particuary if they are inked as services around the care home and for how ong these patterns of association have existed. The reaist synthesis proposed that services for care homes had potentia to improve access to care where there were opportunities for meetings between visiting practitioners and care home staff beyond direct cinician patient contact. Reationa working Activities that fostered patterns of working, which in turn supported the deveopment of reationships based on trust and common interests, were more ikey to increase care home staff confidence and create a sense of coegiaity between the two groups. Possibe mechanisms to bring this about coud incude activities that invoved joint priority-setting and the shared use of assessments, protocos and documentation. The case studies broady confirmed this. We found itte evidence, however, apart from adjusting times of visiting and improving access, of NHS services organising provision to fit with the wishes and suggestions of care home staff or residents. Where there were different patterns of NHS provision these were defined and controed by the NHS. The care home staff and managers did not or coud not exercise much infuence over how the NHS worked with 116 NIHR Journas Library

153 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 them, apart from restricting access or registering a compaint when a service was not provided. Care home residents and their representatives coud not infuence when they woud be seen and by whom, especiay when there were questions and issues around medication. Access was one of the few issues that residents raised in their interviews. Where the patterns of working and visiting created opportunities to meet and discuss care, however, there was a greater mutua appreciation of the chaenges both NHS and care home staff faced each day. The case studies suggested that mechanisms that faciitated reationa working between NHS services around the care home as we as with the care home were important. Activities that fostered a sense of services working together for the care home and not just around individua residents were important. Using care homes as the hub of service provision was not more expensive than the aternative modes (incentives to GPs and no specific arrangements outside the basic GMS contract) and no ess efficient in the use of primary and secondary heath-care services. Indeed, there was some evidence that this approach fostered access to a wider array of services, freed up GPs to focus on GMS tasks and enabed an approximation of care/case management, even when such roes were not made expicit. At the macro or cross-organisationa eve of care, there were two approaches to reationa working, which we have described as outreach and encirced. In the outreach approach, individua practitioners or singe teams were funded to work systematicay with care homes aongside the episodic invovement of existing primary care and speciaist services. They were intended to be the ink with other services, but this aspect of their roe was not embedded into systems of care. In the encirced approach, care home teams had forma inks and referra mechanisms with other community-based and hospita outreach services, some team members aso had care home working as an expicit part of their job specification. We aso found that reiance on individua practitioners or a singe team risked isoation and service discontinuity. This approach was vunerabe to staff turnover, services being discontinued and residents being handed over from one service to another without consutation or review. The pubished iterature consistenty highights the probems of high turnover in care homes. The Nationa Care Forum s 2015 Personne Statistics Survey 155 found that more than haf of socia care staff (58.8%) eave within the first 3 years in post and amost one-third (30.7%) eave in their first year. This ceary was an issue for HCPs visiting care homes, as it prevents them from buiding working reationships with care staff who know the residents we. Conversey, NHS turbuence and NHS staff turnover were probematic for care homes, particuary in circumstances when the departure of an individua cinician ed to the oss of the service. An encirced approach to supporting care homes potentiay provided a network of support that was not specific to a singe person or team. It was therefore more ikey to be robust against changes in the system. It aso reinforced a view of care homes as one part of a system of care for frai oder peope. However, one unintended consequence of this network of support was evidence of dupication of effort, when mutipe MDTs were invoved, such as menta heath outreach, a community fas team and a rehabiitation team, each with overapping ski sets. It is possibe that reducing this dupication may bring costs down, which woud make this a more effective mode of care overa. However, our data provide no evidence of this at present. Importance of genera practitioners In a three sites, the invovement of the GP was important, even if other services had absorbed some of their activities, such as medication review, responsive care and case management. Taking specific tasks (e.g. reguar medication review or initia comprehensive assessment on resident entry to the home) from GPs aowed their contact with care homes to be narrower and more focused. This appeared to faciitate reationa working with care home staff, such that GPs were in effect aowed to practice in ways that they recognised ( to concentrate on being a doctor ). Services that provided intensive care home support, Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 117

154 DISCUSSION AND CONCLUSIONS through a mode of reationa working, sti needed inks to GPs for diagnosis, urgent care and discussions about unresoved issues of care with care home staff, residents and famiy. The importance of the GP roe is refected in the arger number of hospitaisations in site 3. It is possibe that this arose because the oca systems were more chaotic. Specificay, site 3 was acking in two contexts: (1) there was a ack of GPs and staff who were avaiabe and known to care home staff; and (2) referra systems or wrap-around care did not function between the NHS and care homes, nor between the different NHS services. We did not seect the three sites on the basis of how they referred residents between services or how they inked to other NHS services but it appeared to be an important context in how practitioners responded, particuary when faced with residents who needed other speciaist input. This was aso reated to the reationships between secondary care, care home staff and visiting NHS services. Investment in care home-specific work In a sites there was additiona investment in providing NHS services to care homes. The reaist synthesis had argued that financia incentives and targets were important resources (contexts) that coud trigger activities that supported reationa working. The case studies suggested that forma acknowedgement of working with care homes was important and vaued work and not something that coud be squeezed into a caseoad of aready overoaded primary care practitioners. Recognition of the importance of this area of work had a egitimising function that enabed NHS staff to engage more fuy with care homes and their residents. It potentiay freed them from anxiety that they were abrogating other competing responsibiities. This was most apparent when practitioners saw their roe as focusing on providing continuity of support and access to expertise. Where this was the case, the funding acknowedged that working with care homes takes time. Perhaps more importanty it aso recognised the need for the NHS to engage with care homes at an institutiona eve as we as with individua residents. A different response and pattern of invovement was triggered when the need for investment was expressed predominanty as concern about care homes being a drain on NHS resources. This negative mindset did not appear to aow for shared discussions about what kind of heath care or services residents needed. Instead, it ed to a focus on specific issues, such as fas prevention and reduction of emergency ca-outs, a commissioning approach we might ca hostie care (Dr Kenneth Rockwood, Dahousie University, 2016, persona communication). When activities were focused on reducing expenditure, this triggered short-term, negative responses. In these circumstances, commissioners assumed the worst, and measured outcomes in terms of what had not happened and how resources had not been used, rather than focusing on the benefits to residents and potentia job satisfaction for NHS and care home staff. This underying rationae coud affect outcomes, for exampe by an undue emphasis on safeguarding and addressing poor quaity care. Where practitioners or services had an ongoing commitment to the care homes, concerns about quaity of care were more ikey to be presented as probems to be worked through rather than decaimed and reported. Without opportunities to work with care home staff, NHS practitioners experienced frustration, and focused on care home staff shortcomings or what care home staff shoud (foowing extra education and training) be abe to do to support residents. The reaist synthesis argued that, for incentives to work, they needed to aign with the interests and priorities of the practitioners invoved. The cross-case study anaysis suggested that the provision of intensive support or additiona training to reduce demand on NHS resources and then a withdrawa of services satisfied no one. In such circumstances, NHS staff woud not be ikey to prioritise or vaue the deveopment of cose working reationships with care home staff. Direct financia incentives appeared to generate more GP activity. The GPs were incentivised to work as GPs, gatekeeping access to secondary care. We showed that if you invest in GPs they spend more time in care homes, but this did not (as commissioners often hope) automaticay ead to more proactive care, as it did not trigger a change in GP behaviour; as noted, behaviour change required the invovement of other services. 118 NIHR Journas Library

155 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Without an expicit, ongoing commitment to working with care homes and other services on behaf of residents, GPs sought to contro and imit the type and duration of interactions with care home staff, eaving care home staff strugging to find effective ways of working with NHS professionas. The commissioners shared common goas but they pursued different modes of care. There was evidence that the site 1 modes of commissioning had formaised aspects of care that were taking pace organicay esewhere. There was aso evidence that this responsive commissioning was informed by changes in patterns of working by front-ine staff over time. Other sites were earier in the evoution of service deivery modes. What the case study findings demonstrated was, to quote Manzano-Santaea: that programmes transform inexoraby but mosty quiety from their origina design. Manzano-Santaea (2011), 156 p. 21 In a three sites, there were exampes of services and individua practitioners who, over time, had changed how they worked with care homes. This was because previous interventions had not been successfu or, as they acknowedged, because the NHS had faied to understand how care homes worked. In the case of NHS professionas working with, and in, care homes, and drawing on the cross-case study anaysis, severa key mechanisms were identified; these incuded a shared focus on the oder person as the recipient of care (and not soey as a potentia drain on imited resources), fexibiity in the service to accommodate the fuctuations in care home workforce and expertise and a shared carity of purpose. The sense of shared purpose recognised that the care of frai oder peope iving with dementia requires expertise and time, working in a care home is difficut and compex, and work is often provided by the east quaified members of the socia care workforce. 84 These modifications and shifts in thinking were contextuay necessary to trigger a different response to the demands of working with care homes. When funding of NHS services to care homes was not sustained or intermittent, and/or when practitioners eft, there were fewer opportunities to adjust the focus and emphasis around what was important about working with care homes. Access to age-appropriate expertise: the case of dementia care In phase 1 we argued that an important resource (context) to improve heath-reated care is access to appropriate cinica assessment and care. We have suggested that this woud ead to improvements in assessment and in heath-reated outcomes. Phase 2 supported the ogic of this argument inasmuch as it found that pain, pressure ucer prevaence, medication use and comorbidities were predictors of increased heath service utiisation among care home residents. Our findings suggest that access to NHS expertise in dementia care is particuary important. We found that the greater the severity of cognitive impairment, the ess ikey it was that a resident woud see a primary care professiona. The presence of dementia compicated care provision, and not a services coud easiy dea with this compexity. In addition, quaitative accounts from NHS staff described how difficut they found visiting residents with dementia, notaby where there was no ready access to speciaist dementia services. The detaied and sometimes extreme accounts of distress, poice invovement and practitioners anxieties about heping care home staff to dea with vioent episodes underined the importance of access to, and integration of, dementia care expertise. Where a dedicated service of dementia speciaist expertise was provided, it needed to have a remit to work with care home staff concerning referras and working inks with other NHS visiting services. Care home staff coud then be more confident in caring for residents with behavioura and psychoogica symptoms of dementia. As others have found, this kind of speciaist support can have other benefits, such as reduced prescribing of antipsychotic drugs and giving staff new skis in dementia care. 157,158 In some instances, however, services worked in parae when they visited care homes or coud ony be mobiised at a time of crisis, so that residents were passed from one service to another without a cinician co-ordinating that process. This coud then resut in the resident being admitted to hospita. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 119

156 DISCUSSION AND CONCLUSIONS Based on the care home staff accounts of how hospita staff provided care and the ack of communication with care homes, the engths of stay observed for some residents were possiby more a refection of secondary care s inabiity to cope with oder patients with dementia than deficiencies in the care home per se. A person iving with dementia, on average, wi stay twice as ong as equivaent residents aged > 65 years and is three times more ikey to have a fa whie in hospita. 159 This may have been compounded when, as in site 3 (which aso had the highest rate of hospitaisation overa), a ack of structure in working with care homes and HCPs with few inks to hospitas meant that there were fewer opportunities to expedite residents discharge back to the care homes. Care home staff often fet that they knew more in this area than the HCPs with whom they interacted. Finding ways to harness and buid dementia-specific expertise within the care home sector itsef is a egitimate focus of ongoing enquiry. Programme theory In summary, our findings propose an expanatory theory that argues that specific commissioning arrangements are more ikey to work we where: there is a requirement and payment for dedicated care time as part of a job pan or service specification; this had a egitimising function and enabed staff to focus on working with rather than doing to care homes the arrangements acknowedged the need to engage with care homes at an institutiona eve as we as with individua residents the arrangements avoided a narrow description for heath service input to achieve specific outcomes in a very short time; excessive focus on certain activities coud trigger care home staff dependency upon the support received from the short-ived intervention; this coud deter NHS staff from working to estabish reationa inks with the care home team the arrangements avoided reiance on individua practitioners or singe staff groups with busy caseoads, such as GPs, without making space avaiabe for additiona commitments and responsibiities. Commissioning arrangements are ikey to be beneficia if they ead to services being organised around the care home and can adapt over time. Then individua teams can co-ordinate their activities and act in a mutidiscipinary fashion, without necessariy having expicit pathway navigation or case/care management. Buiding services around care homes as the hub of service provision does not necessariy require additiona resources, or the use of secondary or primary care ess productivey than in settings where care was ad hoc, or where care was focused primariy around the GP. This worked we when: the different services saw care homes as a egitimate and shared part of their workoad and respected the roe of care homes as organisations in care deivery, rather than simpy focusing on care deivery to individua residents the GP was part of the care deivery team it fostered access to a wider array of services, aowed GPs to focus on deivery of GMS and enabed an approximation of care/case management, even when such roes were not made expicit ongoing, proactive topic-specific expertise in dementia care was incuded in the range of services provided. Figure 9 sets out what needs to be in pace to achieve a service that is sustainabe and affordabe. This figure summarises what we found to be the constituent CMOs that informed what is required to achieve improved access to heath care. Figure 9 uses proxy outcomes, which are ikey to transfer into improvements in the outcomes we have foowed in this study, that is, those that are ikey to ead to specific improvements in medication management and reduction of urgent and secondary care and continuity of care. In practice, these features have often arisen through a combination of tria and error, 120 NIHR Journas Library

157 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Context Care homes have inequitabe access to heath care Resources Investment in care home-specific services Residents heath-care needs are variabe and compicated by dementia Range of inked services working together for care homes Care home residents use of urgent and emergency care has the potentia to be reduced Care home staff consuted on how and when to visit Proxy outcomes NHS staff assume case management roe for other services on behaf of residents Provision of dementia service for care homes Care home staff and NHS staff have opportunity to pan and work together FIGURE 9 Expanatory theory for NHS work with care homes. Response/mechanism Staff confidence working together Care home work is vaued Residents have network of support Episodes where residents symptoms reated to dementia are contained within the care home Staff deveop expertise in working together Outcomes Staff satisfaction Medication reviews Reduced ength of stay Service has capacity to address demand Care homes integrated into system of care Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 121

158 DISCUSSION AND CONCLUSIONS evoving systems of care and the voices of individua champions. The chaenge for commissioners is how to buid these mechanisms for good outcomes into their oca heath economy. Strengths and imitations A strength of this study is that it has been abe to capture the range of NHS support to care homes and provide an expanatory account of how particuar configurations of contexts and mechanisms are more or ess ikey to achieve certain outcomes. Integra to the design and anaysis is the recognition that programmes are aways affected by their contexts. In the rapidy changing care environment of the NHS, it provides a pausibe account of what shoud be considered in different modes of service deivery for care homes. This detaied examination of micro-eve actions, and the reationships between NHS practitioners, care home staff and residents over time, enabed us to provide an expanatory account, one that identifies the causa processes that underie what was observed. Our account describes the participants activities and their responses that characterise the deivery of heath care to care homes and the environment in which this was achieved (or not). The use of a modified version of interrai is both a strength and a imitation; it is a strength in that it demonstrated the potentia of minimum data sets for within- and cross-case comparisons of care home popuations, and examination of particuar residents service use; it is a imitation in that the findings are constrained by the number of residents we were abe to recruit (50%) and the quaity of the data over the 12-month period. This imited our abiity to test our programme theory against observed use of secondary care. It raises a methodoogica question about the feasibiity of prospective data coection for periods onger than 6 months. Despite financia incentives to participate and good working reationships with the research team, the care home staff s capacity to support resident-based data coection was difficut to maintain for a year. The absence of a nationa minimum data set meant that a disproportionate amount of researcher time was given to obtaining consent from individua residents, capturing detais of residents characteristics and service use, and chasing archived data (that were often not retrievabe). In other countries this kind of resident-eve information is readiy avaiabe as anonymised data. Our study design focused on care homes and the NHS services received from community providers and GPs. We recorded hospitaisations but were unabe to review how NHS professionas became invoved in co-ordinating residents discharge or confirm care home staff s suspicion that hospitas were i equipped to support peope iving with dementia. However, there is an increasing body of evidence that woud support this. 159 Athough we were abe to provide an account of residents resource use, we were unabe to estabish the costs of running the different services provided to the study care homes. An important imitation, with regard to the quantitative anaysis, is that numbers were sma and, therefore, the study was possiby underpowered to detect meaningfu intersite differences in the outcomes measured. This was a recognised consequence of the trade-off between detaied data coection that enabed in-depth description of the cohort and the abiity to recruit care home residents in arge numbers. We have, however, taken account of the numbers of zero-count outcomes and have cross-referenced numerica data with quaitative observations to aow us to more fuy understand any trends or statisticay significant differences identified. Our chosen method of reaist synthesis and evauation has certain strengths in terms of deaing with compicated situations where, for exampe, randomised controed trias wi never be possibe. Our approach, unike many reaist studies, was not to ook at a singe intervention (e.g. care home speciaist teams) impemented in different settings. Instead, we considered the contextua factors that are necessary to trigger the desired mechanisms. Distinguishing between contexts and mechanisms as part of the anaytic 122 NIHR Journas Library

159 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 process is chaenging. 160 Over the course of the project we refined our understanding of how processes within the approach to heath-care deivery coud be recognised in, and across, the sites. It is a strength of the reaist method and how it was appied that it enabed us to integrate and interrogate different forms of knowedge, using both primary and secondary sources. This aowed us to theorise and test how particuar contextua factors, such as how ong NHS staff had worked with care homes, triggered simiar responses. Thus, in one site methods of working together that supported integrated working grew from a series of commissioning decisions over time. Simiar reationa approaches to working were observed esewhere, deveoping organicay over time as individua practitioners earnt how to work with care homes. How anguage is used within reaist evauation is a recurring issue in the methodoogica iterature. The study questions began by suggesting that we woud focus on the features of the different service deivery modes, that is, what is done and how care is organised in order to understand the mechanisms (the interaction of these features with peope s reasoning). This is arguaby miseading and confates eements or features of different service modes with the observed mechanisms that emphasise the interaction of participants reasoning and resources. This study report offers a theory-based expanatory account that can, and shoud, be tested further for its pausibiity and reevance. The study was conducted in the UK, where the NHS creates a particuar context, ways of working and professiona (sub)cutures. Concusion The study has focused on the experience of (and responses to) providing and being the recipients of NHS services to care homes. From this, we have provided a theory-driven account of the underying causa processes that ead to some outcomes being achieved or not. The different contexts observed were not static. In a three sites studied, there were simiar services, but in different concentrations, with different referra systems and frequencies of contact with each other and with care homes. We found that an interest in, and endorsement of, what care homes achieve for oder peope, combined with provision of time to consoidate together, were mechanisms associated with our outcomes of interest. Financia incentives and investment in care home speciaist roes and teams made this achievabe. When both practitioners and the provider organisations were abe to meet frequenty and take a wider interest in care home residents heath, this provided an important context. From this position, care home staff became confident enough to accept, for exampe, criticay i patients because they knew they woud be supported. From the outset, a service deivery mode therefore needs to recognise, record and accommodate the diversity of service groups invoved in providing heath care to residents, and find ways of buiding mutua famiiarity among practitioners and the different systems of care represented. It is those activities that serve to break down the very rea (and documented in this study) sense of us and them between NHS and care home services staff. We found that it took years, as we as targeted funding, to deveop a recognition that working in care homes is difficut and compex work, which refects that time is required to achieve shifts in attitude and approach. Our study represents a microcosm of the ongoing battes to break down what Lewis 161 characterised as the hidden poicy confict that separates heath and socia care. There is penty of evidence that poicy-makers see this as an important area, with the Five Year Forward View from the NHS, 162 the introduction of vanguards 163 and the continuing poicy drive to achieve integrated care. However, our findings suggest that the socia care sector (in this case, care homes) has an important contribution that so far has been underutiised to inform the commissioning of optima heath care for oder citizens. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 123

160 DISCUSSION AND CONCLUSIONS Impications for practice There are mutipe ways in which the NHS works with care homes and it is unikey that there is one right way of working or mode of service deivery. This study has argued, however, that there are common features or aspects of how care home and NHS staff work together that are more (or ess) ikey to support residents access to heath care. This section sets out practice impications for commissioners of services for care homes, practitioners and care home staff and their organisations (a fim of the findings is in preparation; see aso Appendix 14 for the outine of the approach). When commissioning and panning NHS service provision to care homes it is important that residents in care homes have access to heath care that is equitabe and equivaent to those received by oder peope iving at home. We found that service provision to care homes is often ad hoc and reactive, and that some services (e.g. denta heath care, speech and anguage therapy) were either not offered to a care homes or were imited in scope. Heath-care professionas work with care homes shoud be formay recognised by NHS managers as key to the support of integrated working for oder peope. Recognised referra inks with other community and hospita services are more ikey to support continuity of care and management of acute episodes in the care home. Where care home services are a stand-aone service or an adjunct to an existing roe without protected time, practitioners can strugge to co-ordinate residents care and invove NHS services when needed. Investment and incentives to NHS services and practitioners working with care homes shoud be structured to support joint working and panning before services are changed or modified. Where funding and sanctions are designed to reduce inappropriate demand on secondary care and other NHS services this can have the unintended consequences of focusing on faiure. The study found that when funding supported care home teams and GPs to have more time to earn how to work with care homes and identify shared priorities and training needs, this was more ikey to faciitate co-operation, affirm best practice and motivate staff to find shared heath-care soutions. Care home providers referra guidance needs to fit with NHS referra protocos together with opportunities for diaogue where they are uncertain about how to identify different NHS services. The study found that care home staff were often unsure who to invove when they were concerned about a resident. Estabished reationships that had deveoped over time between care home staff and HCPs were aso observed to faciitate appropriate referras that in turn heped to reinforce best practice. Care home-based training needs to incude a care home staff working with residents, not just the nurses or senior carers, to support them to work with the NHS and communicate with famiy carers. New care home staff in particuar need to engage with NHS staff when working with residents and understanding their heath-care needs. The study findings suggested that when training incuded a members of the workforce (e.g. catering staff and junior staff), there was more ikey to be engagement at an organisationa eve and sustained impementation of service improvements. Genera practitioners need to pay a centra roe in residents heath care. How their work compements other care home-focused services shoud be specified and agreed between a those invoved in assessing, treating residents and making referras. Reguar GP cinics or patterns of visiting that were predictabe were associated with higher eves of care home staff satisfaction with heath care and fewer medication-reated probems and more frequent medication reviews. This was particuary true when there were opportunities to discuss care provision across the care home and not just individua residents heath care. Dementia expertise needs to be integra to reguar service provision, not part of a separate service. The study found that both care home and NHS staff coud benefit from ongoing access to training and resources to equip them to support residents iving with dementia. Care home staff pay a vita roe in managing and monitoring residents medication, but may need further training and support in this area. The study found that this was an aspect of care that was of particuar concern to both residents and their reatives. 124 NIHR Journas Library

161 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Recommendations for future research Our recommendations for future research reate both to aspects of research methods and to a number of research questions to further evauate and expicate our programme theory. We concude that there is imited vaue in further descriptive work on NHS heath-care service provision to care homes that is not inked to an understanding of how the services work with care home staff to improve care home residents heath-reated outcomes. There is an urgent need for research that can deveop and refine a minimum data set for residents that can ink with heath and socia care patient/cient data systems. The study findings suggest that when care home staff are confident in their decision-making and right to participate in panning care of their residents, this supported more equa patterns of working. Interventions that deveop care home staff abiity and confidence when working with visiting NHS staff need to test this further. This study found imited evidence of care home residents, staff or famiies infuencing or shaping how or what kind of heath-care support was provided. Further research is needed that can buid on the principes of reationa working and co-design to test different ways of supporting the meaningfu participation of residents, staff and famiies. We found very itte evidence of how famiy members contribute to, or monitor, the heath care that their reatives receive. There is a need for further research to understand how their knowedge of the resident and insights might inform care. Research is needed on how training and deveopment in dementia care across the NHS and socia care workforce (and not just for care home staff) can improve the quaity of care of peope iving and dying with dementia. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 125

162

163 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Acknowedgements We are extremey gratefu to a the study participants incuding the residents, reatives, care home staff, HCPs and stakehoders who took part in the study. In particuar, we are extremey gratefu to everyone invoved in the year-ong duration of phase 2 for their time, input and ongoing interest in the study. We woud aso ike to thank the members of the SSC for their invauabe contributions and the PIR group members. SSC members: Iain Carpenter, Sharon Backburn, Eieen Burns, Kate Grisaffi, Jan Lockyer, Russe Pitchford, Caroine McGraw, Gi Duncan, Jackie Morris, Judy Downey, Aan Rosenbach and Angie Siva. PIR group members: Michae Osborne, Karen Cooper, Katheen Sartain and John Wimott. We woud aso ike to acknowedge and thank Eaine Argye at Nottingham University for her work on the care home staff satisfaction survey and Lindsey Parker at the University of Hertfordshire for a her administrative support. Contributions of authors Caire Goodman (Professor of Heath Care Research, University of Hertfordshire) was the principa investigator, ed the study design, oversaw the whoe study, was invoved in a aspects of phases 1 and 2, and is ead author of the report. Sue L Davies (Research Feow, University of Hertfordshire) was a co-appicant, took day-to-day responsibiity for project management, was invoved in a aspects of phases 1 and 2, incuding recruitment and data coection for the case studies in sites 2 and 3, and was a co-author of the report. Adam L Gordon (Cinica Associate Professor in Medicine of Oder Peope, University of Nottingham) was a co-investigator, co-ed the study design, was invoved in a aspects of phases 1 and 2, oversaw the case studies in site 1, conducted the medication anaysis and care home satisfaction survey, and was a co-author of the report. Tom Dening (Professor of Dementia Research, University of Nottingham) was a member of the management group, invoved in a aspects of both phases of the study, oversaw the case studies in site 1 together with Adam L Gordon and was a co-author of the report. Heather Gage (Professor of Heath Economics, University of Surrey) was a co-appicant, was a member of the management group, ed the economic evauation of phase 2 of the study and was a co-author of the report. Juienne Meyer (Professor of Nursing Care for Oder Adut, City, University of London) was a co-appicant and a member of the management group, contributed to phases 1 and 2, in particuar faciitating recruitment of care homes in site 3 through the MyHomeLife programme, and design and deveopment of the DVD, and was a co-author of the report. Justine Schneider (Professor of Menta Heath and Socia Care, University of Nottingham) was a co-appicant, a member of the management group, contributed to phases 1 and 2 and was a co-author of the report. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 127

164 ACKNOWLEDGEMENTS Brian Be (Research Feow, University of Nottingham) was a member of the management team, provided the statistica support for phase 2 of the study and was a co-author of the report. Jake Jordan (Research Feow in Heath Economics, University of Surrey) was a member of the management team for phase 2, provided the economic evauation for the phase 2 case studies together with Heather Gage and was a co-author of the report. Finbarr Martin (Professor of Medica Gerontoogy, Guy s and St. Thomas NHS Foundation Trust) was a co-appicant and a member of the management team, contributed to a aspects of phases 1 and 2 of the study and was a co-author of the report. Steve Iiffe (Professor of Primary Care for Oder Peope, University Coege London) was a co-appicant and member of the management group, contributed to a aspects of phases 1 and 2, specificay the review of surveys in phase 1, and was a co-author of the report. Cive Bowman (Honorary Visiting Professor, City, University of London) was a co-appicant and member of the management group, contributed to phases 1 and 2 of the study and was a co-author of the report. John RF Gadman (Professor of the Medicine of Oder Peope, University of Nottingham) was a co-appicant and member of the management group, contributed to a aspects of phases 1 and 2 and was a co-author of the report. Christina Victor (Professor of Gerontoogy and Pubic Heath, Brune University) was a co-appicant and a member of the management team, and ed on the secondary data anaysis and write-up of residents interviews in phase 1 of the study. Andrea Mayrhofer (Research Feow, University of Hertfordshire) was invoved in data coection and quaitative anaysis for the phase 2 case studies in sites 2 and 3, was a member of the management team for that duration and was a co-author of the report. Meanie Handey (Research Feow, University of Hertfordshire) was invoved in the phase 1 reaist review, recruitment and data coection in sites 2 and 3 for the phase 2 case studies, was a member of the management team for that duration and was a co-author of the report. Maria Zubair (Research Feow, University of Nottingham), took day-to-day responsibiity for one of the study sites, was invoved in a aspects of phases 1 and 2, incuding recruitment and data anaysis, and was a co-author of the report. Pubications Goodman C, Gordon AL, Martin F, Davies SL, Iiffe S, Bowman C, et a. Effective heath care for oder peope resident in care homes: the optima study protoco for reaist review. Syst Rev 2014;3:49. Gordon AL, Goodman C, Dening T, Davies S, Gadman JR, Be BG, et a. The optima study: describing the key components of optima heathcare deivery to UK care home residents: a research protoco. JAmMed Dir Assoc 2014;15: Goodman C, Davies SL, Gordon AL, Meyer J, Dening T, Gadman JR. Reationships, expertise, incentives, and governance: supporting care home residents access to heath care. An interview study from Engand. JAMA 2015;16: NIHR Journas Library

165 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Goodman C, Dening T, Gordon AL, Davies SL, Meyer J, Martin FC, et a. Effective heath care for oder peope iving and dying in care homes: a reaist review. BMC Heath Serv Res 2016;16:269. Iiffe S, Davies SL, Gordon AL, Schneider J, Dening T, Bowman C, et a. Provision of NHS generaist and speciaist services to care homes in Engand: review of surveys. Primary Heath Care Res Dev 2016;17: Data sharing statement A avaiabe data can be obtained from the corresponding author. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 129

166

167 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 References 1. Gordon AL, Frankin M, Bradshaw L, Logan P, Eiott R, Gadman JR. Heath status of UK care home residents: a cohort study. Age Ageing 2013;43: Goodman C, Davies SL, Dickinson A, Gage H, Froggatt K, Morbey H, et a. A Study to Deveop Integrated Working between Primary Heath Care Services and Care Homes. London: NIHR Service Deivery and Organisation programme; p Goodman C, Robb N, Drennan V, Wooey R. Partnership working by defaut: district nurses and care home staff providing care for oder peope. Heath Soc Care Community 2005;13: Iiffe S, Davies SL, Gordon AL, Schneider J, Dening T, Bowman C, et a. Provision of NHS generaist and speciaist services to care homes in Engand: review of surveys. Prim Heath Care Res Dev 2016;17: Robbins I, Gordon A, Dyas J, Logan P, Gadman J. Expaining the barriers to and tensions in deivering effective heathcare in UK care homes: a quaitative study. BMJ Open 2013;3:e Whittingham KA, Odroyd LE. Using an SBAR - keeping it rea! Demonstrating how improving safe care deivery has been incorporated into a top-up degree programme. Nurse Educ Today 2014;34:e Prince M, Knapp M, Guerchet M, McCrone P, Prina P, Comas-Herrera A, et a. Dementia UK Update. 2nd edn. London: Azheimer s Society; Morey JE, Capan G, Cesari M, Dong B, Faherty JH, Grossberg GT, et a. Internationa survey of nursing home research priorities. J Am Med Dir Assoc 2014;15: j.jamda Mayrhofer A, Goodman C, Smeeton N, Handey M, Amador S, Davies S. The feasibiity of a train-the-trainer approach to end of ife care training in care homes: an evauation. BMC Paiat Care 2016;15: Gendinning C, Powe MA, Rummery K. Partnerships, New Labour and the Governance of Wefare. Bristo: Poicy Press, University of Bristo; Comas-Herrera A, Northey S, Wittenberg R, Knapp M, Bhattacharyya S, Burns A. Future costs of dementia-reated ong-term care: exporing future scenarios. Int Psychogeriatr 2011;23: Laing Buisson. Care of Oder Peope UK Market Report 2014/15. 27th edn. London: Laing Buisson; Hancock R, Maey J, Wittenberg R, Morciano M, Pickard L, King D, Comas-Herrera A. The roe of care home fees in the pubic costs and distributiona effects of potentia reforms to care home funding for oder peope in Engand. Heath Econ Poicy Law 2013;8: /S Demos. The Commission on Residentia Care chaired by Pau Burstow. London: Demos Magdaen House; Handey M, Goodman C, Froggatt K, Mathie E, Gage H, Manthorpe J, et a. Living and dying: responsibiity for end-of-ife care in care homes without on-site nursing provision a prospective study. Heath Soc Care Community 2014;22: Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 131

168 REFERENCES 16. McCauey K, Cross W, Moss C, Wash K, Schofied C, Handey C, et a. What does practice deveopment (PD) offer menta heath-care contexts? A comparative case study of PD methods and outcomes. J Psychiatr Ment Heath Nurs 2014;21: Parsons C, Johnston S, Mathie E, Baron N, Machen I, Amador S, et a. Potentiay inappropriate prescribing in oder peope with dementia in care homes: a retrospective anaysis. Drugs Aging 2012;29: Forder JE, Caies J. Measuring the outcomes of ong-term care. Soc Sci Med 2011;73: Witt E, Goodman C, Banerjee A, Scott T, Amador S. Patterns of attendance of care home residents to an Emergency Department. Eur J Emerg Med 2013;20: /MEJ.0b013e e1 20. Amador S, Goodman C, King D, Machen I, Emore N, Mathie E, Iiffe S. Emergency ambuance service invovement with residentia care homes in the support of oder peope with dementia: an observationa study. BMC Geriatr 2014;14: Steves CJ, Schiff R, Martin FC. Geriatricians and care homes: perspectives from geriatric medicine departments and primary care trusts. Cin Med 2009;9: cinmedicine Woods B, Aguirre E, Spector AE, Orre M. Cognitive stimuation to improve cognitive functioning in peope with dementia. Cochrane Database Syst Rev 2012;2:CD / CD pub2 23. British Geriatrics Society. Quest for Quaity: An Enquiry into the Quaity of Heathcare Support for Oder Peope in Care Homes: A Ca for Leadership, Partnership and Improvement. London:British Geriatrics Society; Potawski L, Goodman C, Iiffe S, Manthorpe J, Gage H, Shah D, Drennan V. Fraity scaes their potentia in interprofessiona working with oder peope: a discussion paper. J Interprof Care 2011;25: Spisbury K, Hewitt C, Stirk L, Bowman C. The reationship between nurse staffing and quaity of care in nursing homes: a systematic review. Int J Nurs Stud 2011;48: /j.ijnurstu Amador S, Goodman C, King D, Ng YT, Emore N, Mathie E, et a. Exporing resource use and associated costs in end-of-ife care for oder peope with dementia in residentia care homes. Int J Geriatr Psychiatry 2014;29: Szczepura A, Neson S, Wid D. In-reach speciaist nursing teams for residentia care homes: uptake of services, impact on care provision and cost-effectiveness. BMC Heath Serv Res 2008;8: Pawson R, Wong G, Owen L. Known knowns, known unknowns, unknown unknowns: the predicament of evidence-based poicy. Am J Eva 2011;32: Pawson R, Tiey N. Reaistic Evauation. London: Sage Pubications Ltd; Goodman C, Gordon AL, Martin F, Davies SL, Iiffe S, Bowman C, et a. Effective heath care for oder peope resident in care homes: the optima study protoco for reaist review. Syst Rev 2014;3: Wong G, Greenhagh T, Westhorp G, Buckingham J, Pawson R. RAMESES pubication standards: reaist syntheses. BMC Med 2013;11: NIHR Journas Library

169 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO Pawson R, Greenhagh T, Harvey G, Washe K. Reaist review a new method of systematic review designed for compex poicy interventions. J Heath Serv Res Poicy 2005;10: Wong G, Greenhagh T, Westhorp G, Pawson R. Reaist methods in medica education research: what are they and what can they contribute? Med Educ 2012;46: j x 34. Sheraw-Johnson C, Smith P, Bardsey M. Continuous monitoring of emergency admissions of oder care home residents to hospita. Age Ageing 2016;45: ageing/afv Connoy M, Broad J, Boyd M, Kerse N, Foster S, Lumey T, et a. Randomised controed tria of packaged evidenced interventions for reducing hospitaisations from residentia aged care (RAC): first resuts from the ARCHUS study. Eur Geriatr Med 2013;4:S j.eurger Mooy DW, Guyatt GH, Russo R, Goeree R, O Brien BJ, Bédard M, et a. Systematic impementation of an advance directive program in nursing homes: a randomized controed tria. JAMA 2000;283: Lewis R, Edwards N. Improving Length of Stay: What Can Hospitas Do? London: Nuffied Trust; Cowing TE, Harris MJ, Majeed A. Evidence and rhetoric about access to UK primary care. BMJ 2015;350:h Bunn F, Burn AM, Goodman C, Rait G, Norton S, Robinson L, et a. Comorbidity and dementia: a scoping review of the iterature. BMC Med 2014;12: s Bowers BJ, Fibich B, Jacobson N. Care-as-service, care-as-reating, care-as-comfort: understanding nursing home residents definitions of quaity. Gerontoogist 2001;41: /geront/ Rycroft-Maone J, Fontena M, Bick D, Seers K. A reaistic evauation: the case of protoco-based care. Impement Sci 2010;5: Rycroft-Maone J, McCormack B, Hutchinson AM, DeCorby K, Buckna TK, Kent B, et a. Reaist synthesis: iustrating the method for impementation research. Impement Sci 2012;7: Goodman C, Dening T, Gordon AL, Davies SL, Meyer J, Martin FC, et a. Effective heath care for oder peope iving and dying in care homes: a reaist review. BMC Heath Serv Res 2016;16: Brugha R, Varvasovszky Z. Stakehoder anaysis: a review. Heath Poicy Pan 2000;15: Davies SL, Goodman C, Manthorpe J, Smith A, Carrick N, Iiffe S. Enabing research in care homes: an evauation of a nationa network of research ready care homes. BMC Med Res Methodo 2014;14: Goodman C, Baron NL, Machen I, Stevenson E, Evans C, Davies SL, Iiffe S. Cuture, consent, costs and care homes: enabing oder peope with dementia to participate in research. Aging Ment Heath 2011;15: McMurdo ME, Roberts H, Parker S, Wyatt N, May H, Goodman C, et a. Improving recruitment of oder peope to research through good practice. Age Ageing 2011;40: /ageing/afr115 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 133

170 REFERENCES 48. Mor V. A comprehensive cinica assessment too to inform poicy and practice: appications of the minimum data set. Med Care 2004;42(Supp. 4):III mr e 49. Carpenter I, Hirdes JP. Using interrai Assessment Systems to Measure and Maintain Quaity of Long-Term Care. In OECD/European Commission, editors. A Good Life in Od Age. Paris: OECD/ European Commission; pp Carpenter I, Perry M, Chais D, Hope K. Identification of registered nursing care of residents in Engish nursing homes using the Minimum Data Set Resident Assessment Instrument (MDS/RAI) and Resource Utiisation Groups version III (RUG-III). Age Ageing 2003;32: /ageing/ Onder G, Carpenter I, Finne-Soveri H, Gindin J, Frijters D, Henrard JC, et a. Assessment of nursing home residents in Europe: the Services and Heath for Edery in Long TERm care (SHELTER) study. BMC Heath Serv Res 2012;12: Rosen R, Mountford J, Lewis G, Lewis R, Shand J, Shaw S. Integration in Action: Four Internationa Case Studies. London: Nuffied Trust; Shaw S, Rosen R, Rumbod B. An Overview of Integrated Care in the NHS. What is Integrated Care? London: Nuffied Trust; Parker G, Corden A, Heaton J. Experiences of and infuences on continuity of care for service users and carers: synthesis of evidence from a research programme. Heath Soc Care Community 2011;19: Hasson H, Arnetz JE. Nursing staff competence, work strain, stress and satisfaction in edery care: a comparison of home-based care and nursing homes. J Cin Nurs 2008;17: Gordon AL, Goodman C, Dening T, Davies S, Gadman JR, Be BG, et a. The optima study: describing the key components of optima heath care deivery to UK care home residents: a research protoco. J Am Med Dir Assoc 2014;15: j.jamda Bunn F, Goodman C, Brayne C, Norton S, Rait G, Robinson L. Comorbidity and Dementia: Improving Heathcare for Peope with Dementia (CoDem). Southampton: NIHR HSDR programme; Marcha B, Dedzo M, Keges G. A reaist evauation of the management of a we-performing regiona hospita in Ghana. BMC Heath Serv Res 2010;10: O Mahony D, O Suivan D, Byrne S, O Connor MN, Ryan C, Gaagher P. STOPP/START criteria for potentiay inappropriate prescribing in oder peope: version 2. Age Ageing 2015;44: Spinewine A, Schmader KE, Barber N, Hughes C, Lapane KL, Swine C, Hanon JT. Appropriate prescribing in edery peope: how we can it be measured and optimised? Lancet 2007;370: O Cathain A, Murphy E, Nicho J. Three techniques for integrating data in mixed methods studies. BMJ 2010;341:c Roya Coege of Physicians, Roya Coege of Nursing, British Geriatrics Society. The Heath and Care of Oder Peope in Care Homes A Comprehensive Interdiscipinary Approach. London: Roya Coege of Physicians; Gadman JRF, Chikura G. Medica Crises in Oder Peope. University of Nottingham Discussion Paper 8. Nottingham: University of Nottingham; NIHR Journas Library

171 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO Gage H, Dickinson A, Victor C, Wiiams P, Cheyne J, Davies SL, et a. Integrated working between residentia care homes and primary care: a survey of care homes in Engand. BMC Geriatr 2012;12: Quince C. Low Expectations: Attitudes on Choice, Care and Community for Peope with Dementia in Care Homes. London: Azheimer s Society; Carter C. Faiing the Frai: A Chaotic Approach to Commissioning Heathcare Services for Care Homes. London: British Geriatrics Society; URL: frai_fu_report.pdf (accessed 22 February 2017). 67. Monaghan N, Morgan M. Ora heath poicy and access to dentistry in care homes. J Disabi Ora Heath 2010;11: Darton R. Study of Care Home Residents and Reatives Expectations and Experiences. Canterbury: Persona Socia Services Research Unit, University of Kent; URL: images/pdfs/pssru_report_may2012rev.pdf (accessed 3 June 2017). 69. Seymour JE, Kumar A, Froggatt K. Do nursing homes for oder peope have the support they need to provide end-of-ife care? A mixed methods enquiry in Engand. Paiat Med 2011;25: Briggs R, Robinson S, Martin F, O Nei D. Standards of medica care for nursing home residents in Europe. Eur Geriatr Med 2012;3: British Denta Association. Dentistry in Care Homes Research - UK URL: dentists/poicy-campaigns/research/patient-care/documents/dentistry_in_care_homes.pdf (accessed 3 June 2017). 72. Commission CQ. Heath Care in Care Homes A Specia Review of the Provision of Heath Care to Those in Care Homes. London: The Stationery Office; Morgan MZ, Monaghan NP, Karki AJ. Waes Care Home Denta Survey URL: data/assets/pdf_fie/0010/47755/wales-care-home.pdf (accessed 3 June 2017). 74. White DA, Tsakos G, Pitts NB, Fuer E, Dougas GVA, Murray JJ, Steee JG. Adut Denta Heath Survey 2009: common ora heath conditions and their impact on the popuation. Br Dent J 2012;213: Leemrijse C, Veenhof C, Boer ME. The avaiabiity and use of aied heath care in care homes in the Midands, UK. Int J Ther Rehabi 2009;16: Guiford M, Morgan M. Access to Heath Care. London: Routedge; Jacobs S. A Nationa Survey of Access to NHS Services of Oder Peope in Residentia and Nursing Care Homes. Manchester: Nationa Centre for Primary Care and Research Deveopment; Guarado L, Cano FG, Damasceno GS, Rozenfed S. Inappropriate medication use among the edery: a systematic review of administrative databases. BMC Geriatr 2011;11: / Abbott RA, Whear R, Thompson-Coon J, Ukoumunne OC, Rogers M, Bethe A, et a. Effectiveness of meatime interventions on nutritiona outcomes for the edery iving in residentia care: a systematic review and meta-anaysis. Ageing Res Rev 2013;12: j.arr Rahman AN, Schnee JF, Appebaum R, Lindabury K, Simmons S. Distance coursework and coaching to improve nursing home incontinence care: essons earned. J Am Geriatr Soc 2012;60: Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 135

172 REFERENCES 81. Shier V, Khodyakov D, Cohen LW, Zimmerman S, Saiba D. What does the evidence reay say about cuture change in nursing homes? Gerontoogist 2014;54(Supp. 1): /geront/gnt Donad IP, Gadman J, Conroy S, Vernon M, Kendrick E, Burns E. Care home medicine in the UK in from the cod. Age Ageing 2008;37: British Geriatrics Society. A Quest for Quaity in Care Homes An Inquiry into the Quaity of Heathcare Support for Oder Peope in Care Homes. London: British Geriatrics Society; Dudman J, Meyer J. Understanding residentia home issues to meet heath-care needs. Br J Community Nurs 2012;17: Adred DP, Raynor DK, Hughes C, Barber N, Chen TF, Spoor P. Interventions to optimise prescribing for oder peope in care homes. Cochrane Database Syst Rev 2013;2:CD Ha S, Koiakou A, Petkova H, Froggatt K, Higginson IJ. Interventions for improving paiative care for oder peope iving in nursing care homes. Cochrane Database Syst Rev 2011;3:CD MacEntee MI, Kazanjian A, Kozak JF, Hornby K, Thorne S, Kettratad-Pruksapong M. A scoping review and research synthesis on financing and reguating ora care in ong-term care faciities. Gerodontoogy 2012;29:e Grabowski DC, Stewart KA, Broderick SM, Coots LA. Predictors of nursing home hospitaization: a review of the iterature. Med Care Res Rev 2008;65: Bradshaw SA, Payford ED, Riazi A. Living we in care homes: a systematic review of quaitative studies. Age Ageing 2012;41: Ha S, Goddard C, Stewart F, Higginson IJ. Impementing a quaity improvement programme in paiative care in care homes: a quaitative study. BMC Geriatr 2011;11: / Goodman C, Davies SL, Gordon AL, Meyer J, Dening T, Gadman JR, et a. Reationships, expertise, incentives, and governance: supporting care home residents access to heath care. An interview study from Engand. J Am Med Dir Assoc 2015;16: j.jamda Mannion R, Davies HT. Payment for performance in heath care. BMJ 2008;336: McDonad R, Harrison S, Checkand K, Campbe SM, Roand M. Impact of financia incentives on cinica autonomy and interna motivation in primary care: ethnographic study. BMJ 2007;334: Doran T, Fuwood C, Kontopanteis E, Reeves D. Effect of financia incentives on inequaities in the deivery of primary cinica care in Engand: anaysis of cinica activity indicators for the quaity and outcomes framework. Lancet 2008;372: X 95. Roand M, Paddison C. Better management of patients with mutimorbidity. BMJ 2013;346:f Charesworth A, Hawkins L, Marsha L. NHS Payment Reform: Lessons from the Past and Directions for the Future. London: Nuffied Trust; NIHR Journas Library

173 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO Evans G. Factors infuencing emergency hospita admissions from nursing and residentia homes: positive resuts from a practice-based audit. J Eva Cin Pract 2011;17: /j x 98. Barber ND, Adred DP, Raynor DK, Dickinson R, Garfied S, Jesson B, et a. Care homes use of medicines study: prevaence, causes and potentia harm of medication errors in care homes for oder peope. Qua Saf Heath Care 2009;18: Shah SM, Carey IM, Harris T, DeWide S, Cook DG. Quaity of prescribing in care homes and the community in Engand and Waes. Br J Gen Pract 2012;62:e bjgp12x Shah SM, Carey IM, Harris T, Dewide S, Cook DG. Quaity of chronic disease care for oder peope in care homes and the community in a primary care pay for performance system: retrospective study. BMJ 2011;342:d Witter S, Fretheim A, Kessy FL, Lindah AK. Paying for performance to improve the deivery of heath interventions in ow- and midde-income countries. Cochrane Database Syst Rev 2012;2:CD Russe G. Does paying for performance in primary care save ives? BMJ 2015;350:h Roand M. When Incentives Go Wrong 55 for Dementia Diagnosis is a Dead Cert URL: (accessed 2 June 2017) Fossey J, Baard C, Juszczak E, James I, Ader N, Jacoby R, Howard R. Effect of enhanced psychosocia care on antipsychotic use in nursing home residents with severe dementia: custer randomised tria. BMJ 2006;332: Loganathan M, Singh S, Frankin BD, Botte A, Majeed A. Interventions to optimise prescribing in care homes: systematic review. Age Ageing 2011;40: ageing/afq Cooke V, Aring G, Lewis T, Abrahamson KA, Mueer C, Edstrom L. Minnesota s nursing faciity performance-based incentive payment program: an innovative mode for promoting care quaity. Gerontoogist 2010;50: Stuck AE, Siu AL, Wieand GD, Adams J, Rubenstein LZ. Comprehensive geriatric assessment: a meta-anaysis of controed trias. Lancet 1993;342: (93)92884-V 108. Eis G, Whitehead MA, O Nei D, Langhorne P, Robinson D. Comprehensive geriatric assessment for oder aduts admitted to hospita. Cochrane Database Syst Rev 2011;7:CD org/ / cd pub Beswick AD, Rees K, Dieppe P, Ayis S, Gooberman-Hi R, Horwood J. Compex interventions to improve physica function and maintain independent iving in edery peope: a systematic review and meta-anaysis. Lancet 2008;371: Gadman J, Harwood R, Conroy S, Logan P, Eiott R, Jones R, et a. Medica crises in oder peope. Programme Grants App Res 2015;3(4) Lyne KJ, Moxon S, Sincair I, Young P, Kirk C, Eison S. Anaysis of a care panning intervention for reducing depression in oder peope in residentia care. Aging Ment Heath 2006;10: Morre CJ, Curran S, Topping A, Shaik K, Muthukrishnan V, Stephenson J. Identification of depressive disorder among oder peope in care homes a feasibiity study. Prim Heath Care Res Dev 2011;12: Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 137

174 REFERENCES 113. Eard DR, Thorogood M, Underwood M, Seae C, Tayor SJC. Whoe home exercise intervention for depression in oder care home residents (the OPERA study): a process evauation. BMC Med 2014;12: Hancock GA, Woods B, Chais D, Orre M. The needs of oder peope with dementia in residentia care. Int J Geriatr Psychiatry 2006;21: Cox H, Puffer S, Morton V, Cooper C, Hodson J, Masud T, et a. Educating nursing home staff on fracture prevention: a custer randomised tria. Age Ageing 2008;37: /ageing/afm Kenkmann A, Price GM, Boton J, Hooper L. Heath, webeing and nutritiona status of oder peope iving in UK care homes: an exporatory evauation of changes in food and drink provision. BMC Geriatr 2010;10: Merre J, Phipin S, Warring J, Hobby D, Gregory V. Addressing the nutritiona needs of oder peope in residentia care homes. Heath Soc Care Community 2012;20: /j x 118. Bamford C, Heaven B, May C, Moynihan P. Impementing nutrition guideines for oder peope in residentia care homes: a quaitative study using normaization process theory. Impement Sci 2012;7: Sackey C, Wade DT, Mant D, Atkinson JC, Yudkin P, Cardoso K, et a. Custer randomized piot controed tria of an occupationa therapy intervention for residents with stroke in UK care homes. Stroke 2006;37: Forsetund L, Eike MC, Gjerberg E, Vist GE. Effect of interventions to reduce potentiay inappropriate use of drugs in nursing homes: a systematic review of randomised controed trias. BMC Geriatr 2011;11: Livingston G, Lewis-Homes E, Pitfied C, Manea M, Chan D, Constant E, et a. Improving the end-of-ife for peope with dementia iving in a care home: an intervention study. Int Psychogeriatr 2013;25: Badger F, Pumridge G, Hewison A, Shaw KL, Thomas K, Cifford C. An evauation of the impact of the God Standards Framework on coaboration in end-of-ife care in nursing homes. A quaitative and quantitative evauation. Int J Nurs Stud 2012;49: /j.ijnurstu Hasson F, Kernohan WG, Wadron M, Whittaker E, McLaughin D. The paiative care ink nurse roe in nursing homes: barriers and faciitators. J Adv Nurs 2008;64: /j x 124. Sprakes K, Tyrer J. Improving wound and pressure area care in a nursing home. Nurs Stand 2010;25: De Visschere LM, van der Putten GJ, Vanobbergen JN, Schos JM, de Baat C, Dutch Association of Nursing Home Physicians. An ora heath care guideine for institutionaised oder peope. Gerodontoogy 2011;28: Mozey CG, Schneider J, Cordingey L, Moineux M, Duggan S, Hart C, et a. The care home activity project: does introducing an occupationa therapy programme reduce depression in care homes? Aging Ment Heath 2007;11: Schneider J, Duggan S, Cordingey L, Mozey CG, Hart C. Costs of occupationa therapy in residentia homes and its impact on service use. Aging Ment Heath 2007;11: NIHR Journas Library

175 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO Goodman C. Evauation of the End of Life Care Train the Trainer (TTT) Education Mode. Hatfied: University of Hertfordshire; Hayward AC, Haring R, Wetten S, Johnson AM, Munro S, Smedey J, et a. Effectiveness of an infuenza vaccine programme for care home staff to prevent death, morbidity, and heath service use among residents: custer randomised controed tria. BMJ 2006;333: /bmj Froggatt K, Hockey J, Parker D, Brazi K. A system ifeword perspective on dying in ong term care settings for oder peope: contested states in contested paces. Heath Pace 2011;17: Masterson-Agar P, Burton CR, Rycroft-Maone J, Sackey CM, Waker MF. Towards a programme theory for fideity in the evauation of compex interventions. J Eva Cin Pract 2014;20: Cooperrider DL, Whitney DK. Appreciative Inquiry: A Positive Revoution in Change. San Francisco, CA: Berrett-Koeher Pubishers; Gitte JH, Weinberg D, Pfeffere S, Bishop C. Impact of reationa coordination on job satisfaction and quaity outcomes: a study of nursing homes. Hum Resour Manag J 2008;18: Toes M, Anderson RA. State of the science: reationship-oriented management practices in nursing homes. Nurs Outook 2011;59: Goodman C, Amador S, Emore N, Machen I, Mathie E. Preferences and priorities for ongoing and end-of-ife care: a quaitative study of oder peope with dementia resident in care homes. Int J Nurs Stud 2013;50: Sackey CM, van den Berg ME, Lett K, Pate S, Hoands K, Wright CC, Hoppitt TJ. Effects of a physiotherapy and occupationa therapy intervention on mobiity and activity in care home residents: a custer randomised controed tria. BMJ 2009;339:b bmj.b Brooker DJ, Wooey RJ, Lee D. Enriching opportunities for peope iving with dementia in nursing homes: an evauation of a muti-eve activity-based mode of care. Aging Ment Heath 2007;11: Mayrhofer A, Goodman C, Homan C. Estabishing a community of practice for dementia champions (innovative practice). Dementia 2015;14: Goodman C, L Davies S, Norton C, Fader M, Morris J, Wes M, Gage H. Can district nurses and care home staff improve bowe care for oder peope using a cinica benchmarking too? Br J Community Nurs 2013;18: Kiney J, Stone L, Dewey M, Levy J, Stewart R, McCrone P, et a. The effect of using high faciitation when impementing the God Standards Framework in Care Homes programme: a custer randomised controed tria. Paiat Med 2014;28: Chenoweth L, King MT, Jeon YH, Brodaty H, Stein-Parbury J, Norman R, et a. Caring for Aged Dementia Care Resident Study (CADRES) of person-centred care, dementia-care mapping, and usua care in dementia: a custer-randomised tria. Lancet Neuro 2009;8: /S (09) Bakerjian D, Zisberg A. Appying the Advancing Exceence in America s Nursing Homes Circe of Success to improving and sustaining quaity. Geriatr Nurs 2013;34: /j.gerinurse Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 139

176 REFERENCES 143. Kodner DL, Spreeuwenberg C. Integrated care: meaning, ogic, appications, and impications a discussion paper. Int J Integr Care 2002;2:e Masso M, McCarthy G, Kitson A. Mechanisms which hep expain impementation of evidence-based practice in residentia aged care faciities: a grounded theory study. IntJNursStud2014;51: Bomsjö BB, Strandberg EL, Midöv P, Brorsson A. It is meaningfu; I fee that I can make a difference a quaitative study about GPs experiences of work at nursing homes in Sweden. BMC Fam Pract 2015;16: Lund S, Richardson A, May C. Barriers to advance care panning at the end of ife: an expanatory systematic review of impementation studies. PLOS ONE 2015;10:e /journa.pone Pubic Heath Engand. Heath Profies. London: Pubic Heath Engand; Office for Nationa Statistics. Changes in the Oder Resident Care Home Popuation between 2001 and London: Office for Nationa Statistics; URL: community/birthsdeathsandmarriages/ageing/artices/changesintheoderresidentcarehome popuationbetween2001and2011/ (accessed 3 June 2017) Gibert H. Supporting Integration Through New Roes and Working Across Boundaries. London: The King s Fund; Larkin M, Boden ZV, Newton E. On the brink of genuiney coaborative care: experience-based co-design in menta heath. Qua Heath Res 2015;25: Bate P, Robert G. Experience-based design: from redesigning the system around the patient to co-designing services with the patient. Qua Saf Heath Care 2006;15: /qshc Kreinder SA, Dowd DA, Dana Star N, Gottschak T. Sios and socia identity: the socia identity approach as a framework for understanding and overcoming divisions in heath care. Mibank Q 2012;90: Merton RK. Socia Theory and Socia Structure. New York, NY: Simon and Schuster; Hedström P, Yikoski P. Causa mechanisms in the socia sciences. Annu Rev Socio 2010;36: Nationa Care Forum. Personne Statistics Survey. Coventry: Nationa Care Forum; Manzano-Santaea A. A reaistic evauation of fines for hospita discharges: incorporating the history of programme evauations in the anaysis. Evauation 2011;17: / Baard C, O Brien J, James I, Mynt P, Lana M, Potkins D, et a. Quaity of ife for peope with dementia iving in residentia and nursing home care: the impact of performance on activities of daiy iving, behaviora and psychoogica symptoms, anguage skis and psychotropic drugs. Int Psychogeriatr 2001;13: Brooker DJ, Latham I, Evans SC, Jacobson N, Perry W, Bray J, et a. FITS into practice: transating research into practice in reducing the use of anti-psychotic medication for peope with dementia iving in care homes. Aging Ment Heath 2016;20: Azheimer s Society. Fix Dementia Care. London: Azheimer s Society; URL: org.uk/homepage/194/fix_dementia_care_homes (accessed 3 June 2017). 140 NIHR Journas Library

177 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO Marcha B, van Bee S, van Omen J, Hoerée T, Keges G. Is reaist evauation keeping its promise? A review of pubished empirica studies in the fied of heath systems research. Eva 2012;18: Lewis J. Oder peope and the heath socia care boundary in the UK: haf a century of hidden poicy confict. Soc Poicy Adm 2001;35: NHS Engand. Five Year Forward View URL: /10/5yfv-web.pdf (accessed 22 February 2017) NHS Engand. New Care Modes Vanguard Sites URL: futurenhs/new-care-modes/ (accessed 22 February 2017) Curtis L, Burns A. Unit Costs of Heath and Socia Care. Canterbury: Persona Socia Services Research Unit, University of Kent; NHS Engand. Nationa Careers Service NHS Agenda for Change Pay Scaes URL: www. nhsempoyers.org/case-studies-and-resources/2015/04/agenda-for-change-pay-bands-and-pointsapri-2015 (accessed December 2015) Curtis L. Unit Costs of Heath and Socia Care. Canterbury: Persona Socia Services Research Unit, University of Kent; Nationa Audit Office. Out of Hours GP Services in Engand. London: Department of Heath and NHS Engand; The Genera Optica Counci. Optica Business Reguation. Fina report for the Genera Optica Counci. London: Europe Economics; Department of Heath. NHS Reference Costs London: Department of Heath; Department of Heath. NHS Reference Costs London: Department of Heath; Ahearn DJ, Jackson TB, McImoye J, Weatherburn AJ. Improving end of ife care for nursing home residents: an anaysis of hospita mortaity and readmission rates. Postgrad Med J 2010;86: Adred DP, Standage C, Fetcher O, Savage I, Carpenter J, Barber N, Raynor DK. The infuence of formuation and medicine deivery system on medication administration errors in care homes for oder peope. Qua Saf Heath Care 2011;20: Badger F, Cifford C, Hewison A, Thomas K. An evauation of the impementation of a programme to improve end-of-ife care in nursing homes. Paiat Med 2009;23: / Barnes L, Cheek J, Nation RL, Gibert A, Paradiso L, Baantyne A. Making sure the residents get their tabets: medication administration in care homes for oder peope. J Adv Nurs 2006;56: Barnett K, McCowan C, Evans JMM, Giespie ND, Davey PG, Fahey T. Prevaence and outcomes of use of potentiay inappropriate medicines in oder peope: cohort study stratified by residence in nursing home or in the community. Qua Saf Heath Care 2011;20: /bmjqs Boumans NP, Berkhout AJ, Vijgen SM, Nijhuis FJ, Vasse RM. The effects of integrated care on quaity of work in nursing homes: a quasi-experiment. Int J Nurs Stud 2008;45: Cegg A, Cingo D, Hinchciffe S. Deveoping the community matron roe in care homes. Br J Community Nurs 2006;11: Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 141

178 REFERENCES 178. Cooney A, Murphy K, O Shea E. Resident perspectives of the determinants of quaity of ife in residentia care in Ireand. J Adv Nurs 2009;65: x 179. Davidson S, Koritsas S, O Connnor DW, Carke D. The feasibiity of a GP ed screening intervention for depression among nursing home residents. Int J Geriatr Psychiatry 2006;21: Dening T, Mine A. Menta Heath in Care Homes. Oxford: Oxford University Press; Doran T, Kontopanteis E, Vaderas JM, Campbe S, Roand M, Saisbury C, Reeves D. Effect of financia incentives on incentivised and non-incentivised cinica activities: ongitudina anaysis of data from the UK Quaity and Outcomes Framework. BMJ 2011;342:d /bmj.d Froggatt K, Davies S, Meyer J. Understanding Care Homes: A Research and Deveopment Perspective. London: Jessica Kingsey Pubishers; Gadsby R, Gaoway M, Barker P, Sincair A. Prescribed medicines for edery frai peope with diabetes resident in nursing homes issues of poypharmacy and medication costs. Diabet Med 2012;29: Gitte JH, Fairfied KM, Bierbaum B, Head W, Jackson R, Key M, et a. Impact of reationa coordination on quaity of care, postoperative pain and functioning, and ength of stay: a nine-hospita study of surgica patients. Med Care 2000;38: Orre M, Hancock G, Hoe J, Woods B, Livingston G, Chais D. A custer randomised controed tria to reduce the unmet needs of peope with dementia iving in residentia care. Int J Geriatr Psychiatry 2007;22: Hockey J, Watson J, Oxenham D, Murray SA. The integrated impementation of two end-of-ife care toos in nursing care homes in the UK: an in-depth evauation. Paiat Med 2010;24: Livingston G, Pitfied C, Morris J, Manea M, Lewis-Homes E, Jacobs H. Care at the end of ife for peope with dementia iving in a care home: a quaitative study of staff experience and attitudes. Int J Geriatr Psychiatry 2012;27: McCormack B, Dewing J, Bresin L, Coyne-Nevin A, Kennedy K, Manning M, et a. Deveoping person-centred practice: nursing outcomes arising from changes to the care environment in residentia settings for oder peope. Int J Oder Peope Nurs 2010;5: /j x 189. McDermott C, Coppin R, Litte P, Leydon G. Hospita admissions from nursing homes: a quaitative study of GP decision making. Br J Gen Pract 2012;62:e /bjgp12X Moyan T, Roberts M, Murray S. Medica needs and surviva of NHS continuing care residents. Scottish Med J 2008;53: Nyman SR, Victor CR. Oder peope s recruitment, sustained participation, and adherence to fas prevention interventions in institutiona settings: a suppement to the Cochrane systematic review. Age Ageing 2011;40: Ong AC, Sabanathan K, Potter JF, Myint PK. High mortaity of oder patients admitted to hospita from care homes and insight into potentia interventions to reduce hospita admissions from care homes: the Norfok experience. Arch Geronto Geriatr 2011;53: j.archger NIHR Journas Library

179 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO Parsons C, Adred D, Daieo L, Hughes C. Prescribing for oder peope in nursing homes: strategies to improve prescribing and medicines use in nursing homes. Int J Oder Peope Nurs 2011;6: Quinn T. Emergency hospita admissions from care-homes: who, why and what happens? A cross-sectiona study. Gerontoogy 2011;57: South A, Tandy C, Watt R, Corrado OJ. Comment on Care home medicine in the UK in from the cod. Age Ageing 2009;38: Szczepura A, Wid D, Neson S. Medication administration errors for oder peope in ong-term residentia care. BMC Geriatr 2011;11: Zermansky AG, Adred DP, Petty DR, Raynor DK, Freemante N, Eastaugh J, Bowie P. Cinica medication review by a pharmacist of edery peope iving in care homes randomised controed tria. Age Ageing 2006;35: Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 143

180

181 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Appendix 1 OPTIMAL Study Steering Committee members: 30 August 2013 Member Expertise 1. Professor Iain Carpenter (chairperson) Emeritus Professor (Human Ageing), Centre for Heath Services Study, University of Kent 2. Jan Lockyer Innovations ead, quaity improvement, Essex County Counci 3. Russe Pitchford Commissioning Manager for Oder Peope at Nottingham City CCG 4/5. Des Key/Sharon Backburn Nationa Care Forum 6. Caroine McGraw District nurse and ecturer in pubic heath, City, University of London 7. Dr Eieen Burns Community geriatrician 8. Angie Siva Care home manager 9. Karen Cooper Care home manager 10. John Wimott University of Hertfordshire PIR Group member 11. Dr Kate Grisaffi GP 12. Dr Jackie Morris Dignity champion/researcher in care homes 13. Aan Rosenbach Poicy ead, CQC 14. Gi Duncan Director of Adut Services at Hampshire County Counci, the Association of Directors of Adut Socia Services representative for nursing and care forum 15. Judy Downey Chairperson of the Reatives and Residents Association 16. Katheen Sartain PIR group member, Nottingham 17. Michae Osborne PIR group member, Nottingham Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 145

182

183 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Appendix 2 Screening form for OPTIMAL Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 147

184 APPENDIX NIHR Journas Library

185 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Appendix 3 Data extraction form used for theory area 1 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 149

186 APPENDIX NIHR Journas Library

187 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 151

188 APPENDIX NIHR Journas Library

189 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 153

190 APPENDIX NIHR Journas Library

191 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Appendix 4 Manager summary Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 155

192 APPENDIX NIHR Journas Library

193 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 157

194 APPENDIX NIHR Journas Library

195 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Appendix 5 Care home ink staff roe Roe of the OPTIMAL care home ink staff The OPTIMAL study aims to understand how the NHS works with care homes and to see whether or not specific ways of working improve the heath care that residents receive. Each care home that takes part wi be asked to identify two members of staff to act as OPTIMAL ink staff to work with and support the research team. As part of this roe they wi be asked to do the foowing: act as a first point of contact and iaison in the care home for the study researchers visiting the care home provide support to the researchers with recruitment of study participants and coection of some of the study information. The care home staff working with OPTIMAL wi hep with the foowing activities: introducing researchers to care home residents for recruitment purposes introducing researchers to interested reatives/famiy members of residents for recruitment purposes introducing researchers to other care home staff working with them identify and introduce researchers to the various NHS HCPs who visit the care home assist researchers with access to residents care home notes (where consent is in pace) assist researchers with the recording of information reating to the participating residents (1) heath and day-to-day function and (2) use of NHS heath-care services (e.g. GP visits, hospita admissions, outpatient appointments) this service information wi be coected every month. The data coection for the study wi take pace over 12 months. During this time, the OPTIMAL care home ink staff wi have contact with the researchers on a reguar basis, for exampe weeky or fortnighty. This wi be agreed between them on the days and times that are most convenient for everyone. We think that the amount of time put into the study for each OPTIMAL ink staff member woud be somewhere around 4 8 hours every month. We woud need more hep at the beginning when residents are being recruited and to coect information from their notes, but foowing on from this, coecting information on residents monthy NHS service use woud take much ess time. There wi be no change in care deivery as a resut of the staff s invovement in the study. The researchers wi work cosey with the staff and residents care wi aways take priority; no residents or staff wi be pressured or coerced into taking part in the study. Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 159

196

197 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Appendix 6 Staff satisfaction questionnaire Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 161

198 APPENDIX NIHR Journas Library

199 DOI: /hsdr05290 HEALTH SERVICES AND DELIVERY RESEARCH 2017 VOL. 5 NO. 29 Appendix 7 Resident service og Queen s Printer and Controer of HMSO This work was produced by Goodman et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 163