PROJECT REPORT DELIVERING AN INTEGRATED GP AND MEDICINES MANAGEMENT SERVICE FOR THE RESIDENTS OF CARE HOMES

PROJECT REPORT DELIVERING AN INTEGRATED GP AND MEDICINES MANAGEMENT SERVICE FOR THE RESIDENTS OF CARE HOMES 1

Contents Contents... 2 1 Executive Summary... 3 2 Introduction... 4 2.1 Background... 4 2.1.1 Project Rationale... 4 2.1.2 Previous Work... 4 2.1.3 Development of the project proposal... 4 2.2 Aim... 5 3 Methodology... 5 3.1 Project Approach... 5 3.2 Project Scope... 6 3.3 Project Outcomes... 6 3.4 Funding... 6 4 Project Delivery... 7 4.1 The Model of Care... 7 4.2 Project Implementation... 8 4.3 Challenges... 8 4.4 Costs and income... 8 4.5 Outcomes... 9 4.5.1 Place of Death... 9 4.5.2 Prescribing... 9 4.5.3 Emergency Admissions... 9 4.5.3 Patient Choice and Satisfaction... 10 5 Project Lessons... 10 5.1 Lessons Identified... 10 6 Conclusion... 11 Appendices... 11 2

1 Executive Summary From time to time, concerns are raised about the quality of care offered in care homes. There are a number of alternative models of care that have been utilised to tackle some of the issues that have been identified in the literature, but one integrated solution to address all these concerns together could not be identified. This project was designed to address that gap. The aim of this project was to fund the delivery of a high quality integrated GP and Medicines Management Service to the residents of care homes which incorporates the residents wishes and addresses inappropriate hospital admissions, both as a proof of concept and to gather data that would support the roll out of similar services on a wider scale. The key project deliverables were: Dedicated clinical and medicines management support to residents of care homes 100% advance care plans completed for care home residents by the end of the year and within one month for each new resident thereafter Evidence of allowing natural death status and evidence of end of life pathways for all residents Production of information on potential medication errors and production of comparative data across all care homes Evidence of a reduction in anti-psychotic prescribing Evidence of a reduction in the number of inappropriate admissions to secondary care Statistical information to support further roll out By December 2011, the project was supporting 450 residents in 7 care homes across Buckinghamshire and Milton Keynes. However, at this stage, the project plan had to take into account an unforeseen decision by the PCT. The business case was predicated upon some funding being provided by the PCT to enable the project to become established and to gain critical mass. In the event, the PCT made the limited funding available to all local GPs, meaning that the project could not recruit sufficient patients to generate the economies of scale to make the service viable in the medium term. As a consequence, the project was terminated after six months. The project demonstrated that providing a dedicated stand-alone service for a population of care home residents can become the focal point for delivering high quality care to some of the most vulnerable people in our society. It is cost effective and can deliver at least a 40% reduction in hospital admissions and a 17% reduction in prescribing. Moreover, during the life of the project, 89% of residents were able to die in the place of their own choosing. There are significant benefits to centralising such a service and asking GPs to work within this framework, rather than having a large number of autonomous practitioners working to variable standards. However, while this model has great potential for delivering change nationally, it has significant implications for the ways in which the boundaries between GP practices are currently managed. 3

2 Introduction 2.1 Background 2.1.1 Project Rationale From time to time, concerns are raised about the quality of care offered in care homes. For example, it has been suggested that as many as 40% of residents who died in hospital did not have medical needs that required admission; that some 70% of care home residents have experienced at least one medication error; that residents experience poly pharmacy (8 items on average per resident) and there is a lack of significant medication reviews; that the inappropriate prescription of antipsychotics may lead to 1,800 residents dying each year due to wrongly prescribed antipsychotics; that there can be a lack of dignity in services provided to care home residents, resulting in such outcomes as end of life episodes that are not in accordance with the individual s wishes; and, finally, that high level of admissions can be attributed to preventable dehydration. Indeed, the World Health Organisation report Better Palliative Care for Older People showed that England and Wales have the lowest proportion of nursing home deaths and the highest proportion of hospital deaths amongst all countries that they studied. This is likely both to lead to increased cost to the NHS and a less than optimum experience of End of Life care for some patients. 2.1.2 Previous Work There are a number of alternative models of care that have been utilised to tackle some of the issues outlined above, but we could not identify one integrated solution. This project was designed to address that gap. In December 2008, the Health and Re-enablement Team (HaRT) project was established in Buckinghamshire. Delivery of this project was achieved by commissioning The Practice plc to deliver a service that provided care homes with: expert GP support during Out of Hours (OOH); specialist GPs providing advance care plans and end of life support; expert GPs receiving calls directly from the care homes at times of crisis, and from the local ambulance service. However, the HaRT project did not include an integrated pharmacy service, overnight GP telephone cover or daytime GP care. The HaRT project ran until March 2009, and had a significant impact. It resulted in 95 avoided admissions in 3 months, saving an estimated 72,000. At this point, the project was completed, with admissions avoidance work being incorporated into a newly commissioned County-wide OOH service and responsibility for advanced care planning being given to local GPs with a financial payment for undertaking this work. However, to date, the on-going benefits of these enhanced services have not been formally evaluated. 2.1.3 Development of the project proposal In July 2010, The Practice plc submitted an application for funding to the South Central Regional Innovation Fund, proposing an enhanced Clinical and Medicines Management Service to the residents of a number of care homes within Buckingham shire. This proposal built upon the experience of HaRT and included provision of: An integrated Clinical and Medicines Management Service to Care Homes 4

Specialist GP support delivering routine visits, continuity of care and high quality chronic disease management A training programme for clinicians in elderly and end of life care An urgent care service, with visits, available 12 hours a day seven days a week in order to provide overnight support by telephone (including tele-conferencing) Support with access to care plans, all day every day Monitoring, managing, auditing and training to ensure consistency in medicine management Arrangements with the local Ambulance Service for TPCL GPs to be called each time an ambulance is requested, to ensure that, where possible, the resident is seen in the home rather than being taken to A&E Support for End of Life discussions, and training to provide care home staff with the necessary skills, to enable more residents to be treated according to their wishes in the familiar environment of their home, as opposed to going to hospital The initial application was unsuccessful. However, with the support of Thames Valley Health Innovation and Education Cluster (HIEC), a refined model and project proposal was submitted and was successful in gaining project funding. The successful bid resulted in a grant of 65,000 enabling the service to be extended to 1,000 care home residents, with any shortfall of funds to be covered by the Thames Valley HIEC up to 30,000. The Practice was also willing to sustain a loss during the proof of concept phase, believing that, once funds from reductions in prescribing and inappropriate admissions could be accessed, the service would be self-funding. The Practice plc met with many national and local stakeholders in the development of the model. These included the English Community Care Association (ECCA), the National Institute for Clinical Excellence (NICE), Four Seasons Health Care, Barchester Healthcare, NHS Diabetes and the End of Life Care Strategy. Locally, the model was evolved through discussions with South Central SHA, local commissioning bodies, South Central Ambulance Trust, community care providers, GPs, care homes and their residents, Florence Nightingale Hospice, Buckinghamshire Hospitals Trust and the Alzheimer s Society. 2.2 Aim The aim of this project was to fund the delivery of a quality integrated GP and Medicines Management Service to the residents of care homes which incorporates the residents wishes and addresses inappropriate hospital admissions, both as a proof of concept and to gather data that would support the roll out of similar services on a wider scale. 3 Methodology 3.1 Project Approach The project was planned to be a proof of concept lasting one year. Throughout the life of the project, The Practice plc was responsible for project management and delivery of the service. Thames Valley HIEC was responsible for project governance, reporting and facilitating engagement with stakeholders beyond those directly involved in delivering the project. The key project deliverables were: Dedicated clinical and medicines management support to residents of care homes 100% advance care plans completed for care home residents by the end of the year and within one month for each new resident thereafter Evidence of allowing natural death status and evidence of end of life pathways for all residents 5

Production of information on potential medication errors and production of comparative data across all care homes Evidence of a reduction in anti-psychotic prescribing Evidence of a reduction in the number of inappropriate admissions to secondary care Statistical information to support further roll out The project would provide a dedicated clinical team comprising specialist GPs and pharmacy staff who would conduct routine visits medicines management urgent care management of EOL pathways and advance care planning out of hours advice. Initial pharmacy input came from from Pharmacy Plus, including integrated medicines management, pharmacy support, and GP support within their call centre. Later, further pharmacy support was provided by the PCT care homes pharmacy. 3.2 Project Scope The project aimed to work with up to 1,000 residents of care homes run by The Freemantle Trust, Westgate Homes, Gold Hill and the Restful Homes Group, located predominately within Buckinghamshire and Milton Keynes. The project would fund routine cover GPs (one WTE GP per 500 residents) and the OOH GP for telephone support (1 x GP for up to 1000 residents @ c 40 per night). 3.3 Project Outcomes The desired outcomes of the project were: 10% reduction in prescribing in care homes 90% of EOL episodes in accordance with the patient s care plan Elimination of medication errors in the care homes All homes to be brought below the national average for anti-psychotic prescribing An average of a 20% reduction in the number of visits by the local ambulance service An average of 20% reduction in inappropriate admissions 3.4 Funding The Regional Innovation Fund grant of 65,000 enabled the service to be extended to 1,000 care home residents, with any shortfall of funds to be covered by the Thames Valley HIEC up to 30,000. It was envisaged that the funding would be allocated as follows with payment being tied to resident sign-up and performance: 1 May 2011 1 August 2011 1 November 2011 1 February 2012 Resident Target 350 900 1,000 Set-up Costs 13,000 6,250 OOH Cover 3,000 3,000 3,000 3,000 Routine Cover 8,750 22,500 25,000 TOTAL 16,000 18,000 25,500 28,000 6

The Routine Cover payment was set as 25 per resident signed up, up to maximum figure shown above. However, in the event that the target figure was not reached by the payment date, the payment would reduce by 25 for each resident below the target. 4 Project Delivery 4.1 The Model of Care The model of care was developed around the idea of creating a stand-alone service that was specific for care home residents. The residents would be notionally registered to one of the Practice plc s local surgeries in order to obtain the registration fee; however, the service had a separate office, telephone number and workforce. The basic concept was to ask residents and their families how they wanted to be treated, record their wishes, ensure they were carried out, and then audit an individual s care to confirm that correct action had been taken. A gap analysis would be performed when the outcome did not meet the original plan. A four hour visit by a regular experienced doctor was allocated for every 35 residents in a home (residential, nursing or dementia). The GP was able to go through letters and results with care home senior staff, see relatives as requested and work to a proactive agenda. This included a new resident review, medication reviews, completion of Do Not Attempt Cardiopulmonary Resuscitation (DNACPR) forms, reviewing preferred priorities of care forms and a variety of tasks identified in the service plan. When the size of a home dictated more than one visit per week, the units in the home had different doctors who shared the primary responsibility. Routine visits were also planned throughout bank holiday periods as these are a high risk period for inappropriate emergency admissions. Each home had a lap top and printer with access to clinical records and all were Information Governance Statement of Compliance (IG SoC) compliant. Monthly repeat prescriptions were printed in the home, where possible, with senior care home staff support. In addition to the routine visits, the service provided an urgent intervention team ready to visit at any time during normal GP hours in order to offer a rapid response. Moreover, at night and over weekends, there was a GP available on the telephone to offer advice to the homes with regard to supporting the care plans that have been made. Normal OOH GP services were still available should it be felt that a GP visit was required out of hours. The Practice plc also allocated a nurse responsible for chronic disease management and QOF issues within the homes. She could also carry out other duties, such as ear syringing, as needed by residents. In order to administer the system, a practice manager and two part time admin staff were provided for the service to answer the care homes phone line and to undertake all other functions normally done in a GP surgery. These staff also gathered the weekly data from the homes on admissions and deaths for the clinical lead. The practice manager and clinical lead visited each home monthly to discuss performance and any issues identified. Moreover, they supported the homes with issues such as training for verification of death. Nursing staff went to a Practice plc surgery for accreditation in phlebotomy. The Practice plc also arranged for pathology samples to be collected from individual care homes. In line with the proactive theme, plans were also put in place with the homes to establish the response to adverse weather conditions such as snow, including a list of relatives and staff who could help with 4 wheel drive vehicles where necessary. 7

4.2 Project Implementation The service went live in the first home, Milton Court, in May 2011. By July, the service was live in three homes and on-going discussions with further homes had identified potential participants to take the resident population up to the 1,000 target. By September 2011, the service was live in 6 homes and recruitment of residents was on track with the 350 resident target for August being exceeded. Moreover, initial savings were looking very promising. In June, 37,000 was saved from the Milton Court monthly prescribing costs following a medicines review. This equated to 400 per resident, and further savings of approximately 100 per resident were projected month on month. By December 2011, the project was supporting 450 residents in 7 care homes across Buckinghamshire and Milton Keynes. However, at this stage, the project plan had to take into account an unforeseen decision by the PCT. The Practice plc business case was predicated upon some funding being provided by NHS Buckinghamshire to enable the project to become established and to gain critical mass. In the event, NHS Buckinghamshire made the limited funding available to all GPs across Buckinghamshire, meaning that the project could not recruit sufficient patients to generate the economies of scale to make the service viable in the medium term. As a consequence, The Practice plc decided to terminate the service in early 2012. 4.3 Challenges A number of challenges were identified when implementing the service: Care homes were located in the catchment areas of two PCTs, resulting in nominal registrations in two surgeries with different IT systems. This doubled the IT challenges. The aim was to work with one pharmacy and align their incentives to rational prescribing. This was not possible as care homes have loyalties and financial connections with specific pharmacies, with deals sometimes agreed nationally if the homes and the pharmacies are national providers. Baseline data was extremely difficult to access, prior to service commencement. During the service, two homes moved site and therefore accessed different secondary care services. Two homes increased their patient base greatly. Such factors made comparisons of data more difficult. Moreover, given the relatively short life of the service, the data was not adjusted for seasonal differences, as would have happened by comparing the same months in different years. Linking this bespoke service to a surgery has a profound effect on a surgery s QOF performance and prescribing targets. Baselines and agreements need to be established in advance. Providing IT into every care home was complex as homes have different levels of internet capability. IG SoC compliance is mandatory and complex. The service averaged 4-5 deaths a week. It can be a complex process, getting a sessional doctor doing death certificates and cremations in a timely way. The local commissioning function and the provision of funding proved to be a challenge, as agreed funding was withdrawn part way through the project. The success of the model caused the commissioning groups to offer all GPs a payment to match the service developed through this project. 4.4 Costs and income The cost of providing doctors, nurses, admin, and IT support equated to 650 per resident. This would have reduced significantly if the Practice plc had been able to scale the service so that the urgent intervention team could cover more residents within the agreed staff base. 8

Income was generated from registration, QOF, cremation fees and flu vaccination. In addition, two homes paid retainers totalling 30,000. The intention was that these retainers would have ceased once the service was fully commissioned. Initially, the Practice plc was offered 400 a resident by NHS Buckinghamshire in return for delivering reductions in admission targets. Later, this figure was altered to a flat payment of 85 per resident. Overall, taking into account project funding, this left a shortfall of approximately 100 per resident. Such a shortfall seems insignificant in the light of a 200 saving per resident in reduced prescribing, and many thousands to be saved from inappropriate admissions, once such savings start to be realised. It is now the PCT s intention to offer 200 a year for each of the 3000 residents in care homes next year, if GPs comply with the service specification developed through this project. However, opening the service to all GP practices in Buckinghamshire reduces the benefits gained through centralising data and building an integrated service which is developing by working through problems as they arise. 4.5 Outcomes 4.5.1 Place of Death Throughout the life of the project, there were 67 residents who died. 89% of deaths occurred within the resident s care home, in accordance with their wishes, at an average rate of 4.2 a week. The week by week data supporting this, including admissions and deaths, is in Appendix A. 4.5.2 Prescribing The item per script cost fell from 13.10 to 10.90 an item, demonstrating a 17% reduction in prescribing costs. This does not include the significant sums saved by reducing the use of specials. The saving on item per script cost was projected to be 85,000 for the service over the course of a year (over 200 a resident). Antipsychotic prescriptions have not fallen. However, the dosages have reduced. 4.5.3 Emergency Admissions Hospital admissions have fallen significantly: In the twelve months previous to the start of the project, there were 237 admissions in total from the care homes to High Wycombe, Milton Keynes General and Stoke Mandeville hospitals, at an average rate of 4.56 admissions a week. This is the baseline that was used to compare the outcomes of the project with the situation before it started. Admissions in the first four weeks of the project totalled 22, an average of 5.5 a week. Overall, across the life of the project, there were 58 admissions, an average of 3.65 a week and a 20% reduction on the average for the previous year. In the last 4 weeks of the project, despite the very significant increase in the number of patients enrolled onto the project, admissions totalled 11, an average of 2.75 a week and a 40% reduction on the average for the previous year. A comparison of the first four weeks of the project, with the last four weeks, shows a reduction of 50% from the start of the project to the end. These figures are a significant underestimate of the true impact of the project for the following reasons: 9

The project has mainly been running across the busiest time of year for admissions (winter months and Christmas bank holidays), so it is expected that admissions would be lower still in summer months Two homes have increased significantly in size, increasing the likelihood of unscheduled admissions Two homes have changed status from residential to nursing care. Frailer patients are more likely to be admitted. Two homes previously admitted significant numbers to Wexham Park hospital. These admissions were not able to be quantified and added to our baseline data for the previous year s activity. One of the homes relocated within the catchment of High Wycombe and we arranged admissions from the other home also to High Wycombe. The service is based on gradually identifying gaps in service provision and engaging in negotiating changes in that provision from secondary care and community services. This is a process that will deliver improvements over years and not the first few months. Interpreting these figures, there is currently a reduction in hospital admissions of approximately 40% from the previous year. However, if the baseline figures from Wexham Park were included, the real figure is likely to be in excess of 50%. However, if the project had run for the full twelve months that were initially envisaged, there would be more data from which to derive trends. As it is, we can only speculate about the potential impact of a longer period in which to embed the service, and the impact of the summer months on unscheduled admissions. 4.5.3 Patient Choice and Satisfaction Reported resident satisfaction/care staff satisfaction is extremely high. Patient choice in selecting the new service was respected and indeed previously many residents were not given choice as they ended up being allocated to a GP by the PCT. 95% of residents have chosen to utilise the new service. 5 Project Lessons 5.1 Lessons Identified The following Lessons were identified: Not all GPs are suited to deliver a clinical pathway that is determined by a resident, which may not be in line with current clinician ideas of best practice. Sharing decision making requires confidence and good communication skills. Two GPs left shortly after starting. The perfect model for this kind of service is a central database and agreed KPIs, and a service that utilises the skills of interested local GPs. It is a challenge for external organisations to win the confidence of the local GP workforce. The Quality and Outcomes Framework (QOF) looks very different for care homes. The Practice plc had to accept undershooting some targets as that is often clinically appropriate for some individuals. However, the opportunity is there to compensate for this by scoring highly on Dementia and Palliative care registers. The impact on prevalence generally should not be ignored. Working with multiple pharmacies was very difficult logistically, but seemed unavoidable given the business relationships of the different care homes Further economies could be made by using nurse prescribers to help with repeat prescriptions. 10

GPs could be more productive looking after a larger number of residents and working in clusters. Little things cause big problems. Ear syringing is an issue as nursing homes rarely provide this. Different homes all had different arrangements for monitoring warfarin. Out of Hours support as a specific element in the service specification would be unnecessary with the correct assurance and communication with an engaged OOH provider. Care home providers are crying out to be involved, in order to deliver a higher quality of care to their residents. This model is an essential building block to building consistent high quality care in care homes that will support the increasing demands that will be placed on community healthcare in the coming years. We need to find a variety of providers to coordinate this that can engage local GPs, such as the care home organisations themselves. 6 Conclusion The project demonstrated that providing a dedicated stand-alone service for a population of care home residents can become the focal point for delivering high quality care to some of the most vulnerable people in our society. It is cost effective and delivers a 40% reduction in hospital admissions, 17% reduction in prescribing and allows 89% of residents that die to do so in the place of their choosing. There are significant benefits to centralising such a service and asking GPs to work within this framework, rather than having a large number of autonomous practitioners working to variable standards. However, while this model has great potential for delivering change nationally, it has significant implications for the ways in which the boundaries between GP practices are currently managed. Appendices A Admissions and Deaths Data 11

Appendix A Admissions and Deaths Data DATE No of admissions No of Deaths No died in place of choice? 14-21 Sep 2011 6 3 2 21-28 Sep 2011 7 5 5 28 Sep-5 Oct 2011 4 6 5 5-12 Oct 2011 5 5 5 13-20 Oct 2011 4 5 5 19-26 Oct 2011 5 2 2 26Oct-2 Nov 2011 2 1 1 2 Nov-9 Nov 2011 2 7 5 9-16 Nov 2011 2 3 3 16-23 Nov 2011 3 5 4 23-30 Nov 2 7 7 30Nov-7 Dec 5 4 3 7Dec-14th Dec 0 3 3 14Dec-21Dec 4 4 4 21dec-28Dec 4 3 2 28Dec-4Jan 3 4 3 TOTAL 58 67 59 89% died in care home A-1