GP at Hand Evaluation: DRAFT Invitation to Tender

Transcription

1 GP at Hand Evaluation: DRAFT Invitation to Tender Introduction Hammersmith & Fulham CCG, together with their partners NHS England London Region and NHS England ( the clients ), invite bids for the evaluation of the GP at Hand service. GP at Hand is an innovative approach to offering digital-first primary care, based within a Hammersmith & Fulham (H&F) GP practice, but offering care to patients across London (and ultimately across England). It has the potential to transform the way primary care is delivered in England. The clients invite tenders from well-qualified bidders to deliver a rapid, policyfocussed, and robust piece of analysis. Successful bids are likely to incorporate qualitative, quantitative, and economic methods, as well as including considerable clinical, evaluative, digital and tech-sector specific expertise and input. We anticipate that, given the diverse range of skills required, successful bids are likely to come from consortia. Background to GP at Hand GP at Hand is the name of a GMS contract-holding general practice providing Primary Medical Services under the GMS Regulations 2015 in North West London, previously known as Dr Jeffries and Partners. They are now offering an innovative, digital-first model of primary care, primarily through use of a mobile app and video consultations provided by their subcontractor, Babylon Health. The app enables patients to access GP services 24/7, at short notice, via a virtual appointment using video conferencing and voice calls on a smartphone. Patients can also access symptom-checker services (driven by Babylon s Artificial Intelligence tool) and health monitoring software. If patients need a face-to-face appointment they can access in-person services in one of GP at Hand s six Londonbased sites. All patients accessing the service de-register from their previous NHS practice, and register with GP at Hand. The service takes advantage of the National GP Choice policy 1 to allow patients from across London to register with the service, with plans to expand the service nationwide. There are some limitations put on registrations from people with more complex care needs, who might need greater continuity of care, and more face-toface support with NHS England s clinical review recommending that those individuals seek advice before registering. GP at Hand has a current list size of over 16,000 patients (figure correct in January 2018), having increased from around 2,500 in July. Over 8,000 patients were added in December alone. These patients are predominantly young adults (over two-thirds aged 20-34, compared to less than a third for other H&F CCG practices), from outside H&F CCG (currently 6% of new registrations are from within the CCG area). 1 This allows patients to register with a GP away from their place of residence, and was introduced primarily to allow patients to register with a GP near their place of work. 1

2 There were initially more men than women registering, though this difference declined in December newly registered patients are on the severe and enduring mental illness register. Evaluation of GP at Hand GP at Hand is likely to have wide-ranging implications for: users of the service; nonusers of the service; Hammersmith & Fulham CCG; other GP practices; the wider health and social care system; and local and national policy. An independent evaluation is needed to provide a fact-based, robust, and fair assessment of what these implications are. A number of evaluative activities have already been undertaken, or are being undertaken at the moment. These include: A first clinical review, carried out by H&F CCG and the Medical Directorate of NHSE London. This recommended a more limited roll-out of the service, and the commissioning of an independent evaluation 2. A second clinical review (on-going), also carried out by H&F CCG and the Medical Directorate of NHSE London, focussing on the safety and effectiveness of GP at Hand, and whether it has addressed the concerns raised in the first review. Rapid analysis (by H&F CCG, supported by the Operational Research & Evaluation Unit (ORE) of NHSE) of the financial implications of GP at Hand for the CCG, and for other GP practices. On-going analysis of key data (particularly descriptive data on who is accessing the service, how, and what their utilisation of wider services is like), by NHSE to answer emerging NHSE policy questions. We would expect the independent evaluation to build on, and (where appropriate) extend these findings. We would expect any relevant analysis undertaken by the CCG or NHSE to be included in the evidence base that the study draws upon, and that the independent evaluator should work closely with all stakeholders, including representatives of the local medical committee (LMC) 3 to ensure that there is no repetition of work, and that all insights are shared and capitalised on. We would also expect the evaluator to provide some peer review of key analytical outputs, such as the second clinical review. Key research questions and objectives The evaluation should answer a set of research questions around the impact of the service on users, non-users, the wider health and social care system, and the workforce. Below we outline some of the key themes we would expect to be considered, with a fuller list of research questions annexed at the back. Areas where significant internal analytical work is likely to be undertaken are highlighted, but 2 The Clinical Review is available here: Clinical-Review.pdf 3 The body recognised as the representative body of local GPs under the NHS Act 2006 c97 and the GMS regulations 2015 part 13. 2

3 bidders should also dedicate some time within the scoping phase to ensure they fully understand and include all existing evaluative work. We would also expect the scoping phase to include some prioritisation from amongst the long-list of research questions and themes. We do not expect the evaluation to be able to answer everything. Bidders should note that the evaluation is relatively tightly focussed on the GP at Hand service itself, not (currently) other digital models (though these could be included as counterfactuals), or on the effectiveness of the new AI triage system. However, we might expect findings from this evaluation to be used by policy makers working in these areas. The GP at Hand practice s offer to patients is new, and our understanding of how it works, how people use it, and what implications and outcomes this will have is limited. For example, there may be reasons to expect GP at Hand to both decrease attendances at A&E/ UTCs (due to quicker access to primary care), and to increase attendances at the same place (due to longer distances to travel to a face-to-face appointment potentially past an A&E). The successful bidder will therefore need to develop a research methodology that is able to cope with this uncertainty. The evaluation should look at impacts on: Users of the service/registered patients Experience e.g. understanding, convenience, speed, satisfaction, continuity of care, relationship with the doctor; Safety e.g. accuracy of diagnosis, speed of and access to treatment, safeguarding. How do handoffs between the digital and face-to-face elements of GP at Hand work? Effectiveness e.g. earlier diagnosis and treatment, additional/more effective treatment, continuity of care, signposting/ referring to appropriate local services, access to and use of social prescribing, promotion of self-care, ability to reach currently unmet need, gaps in care. Cost and efficiency e.g. cost per patient per year, needs adjusted. Equity e.g. equity of access, and equity of experience for GP at Hand users. For all of these elements, the evaluation will need to understand who the service is working particularly well (and not well) for. This could feed into discussions around defining particular population segments for whom this approach might be beneficial. The wider system Analysis is already underway on the immediate cost implications of GP at Hand on the CCG and other GP practices; this is being produced by the clients to inform the CCG s and NHS England s response to the service. The evaluator will be expected to build on and extend this work. In particular they will be expected to understand the demand implications across the wider system, and to take an overall view of the value offered by the service. Where work is primarily undertaken internally, this is noted. 3

4 Key themes will include looking at the impact on: Other GP practices and their patients. Given the demographics of the GP at Hand population, what does this mean to the finances of practices they leave? What are the consequences to practice resources due to the Carr-Hill Formula? What does this mean for short and longer term quality, profitability, sustainability of the traditional GP, and the quality of the service they can afford to provide? Will there be equity in the service received by the cohorts of patients served by GP at Hand and patients of traditional practices? Will there be positive impacts for other GP practices? Might it encourage them to innovate, and offer similar digital services themselves? Some work to answer this question will begin internally; bidders will need to collaborate and build on this work, for example in developing the counterfactuals. CCG finances. What are the financial implications of increases in numbers of patients registered to a GP practice in Hammersmith & Fulham CCG? Are there any quality implications? Are there any overall changes in costs for the NHS? This research question will be primarily addressed by internal analysis, but evaluators may need to extend this work. Referral pathways. Is there any impact on clinical pathways into community, secondary and tertiary care as a result of non-local GPs making referrals? What is the impact where CCGs hold local block-contracts, for example for IAPT or community services, can patients access these? Are there changes in quality, speed and appropriateness? How does this interact with policies promoting models of care based around defined geographies (place-based care) both within Hammersmith & Fulham and nationally? Overall demand and costs for primary, community and acute care? How does this affect patients utilisation of primary, community, and acute care? Are there any issues around supply-induced demand? What effects does this have on primary care as a demand management gateway? What cost implications might this have, and who do the costs fall to in the short and longer term? Some work to answer this question will begin internally; bidders will need to collaborate and build on this work. Productivity, efficiency and value. Does the GP at Hand model deliver value 4 for patients; the CCG; NHSE and the wider system? Is it a more productive and efficient way of delivering primary care? Does it offer a good-value, cost-effective alternative to other services (including 111, integrated urgent care, and GP out of hours services)? The workforce GP at Hand workforce. What are the possible effects of GP at Hand on staff including, contract terms and conditions, job descriptions, pay, training, workload, appraisal and job wellbeing/ satisfaction? What impact does GP at Hand have on the nature of the doctor/ patient interaction and relationships? Wider GP workforce e.g. GPs leaving to join Babylon/GP at Hand. 4 By value we mean a nuanced view of a service that takes into account both cost and quality. 4

5 Wider policy questions Further research projects looking at some of the wider policy questions raised by GP at Hand are also being considered. At this time, options include a broader review of digital models, and a horizon-scanning and scenario planning project to understand how this service, and the market might develop. These may be undertaken internally or externally, but are not part of this specification. To understand why and how impacts are arising, the evaluation should also answer a fuller set of questions. These will include questions about who is accessing the service, in what ways, for how long, as well as the business model and services being offered by GP at Hand. This fuller set of research questions is included as an annex to this paper. Methodological approach Below is a brief outline of some of the approaches we would expect the bidder to consider when designing this evaluation. We would welcome additional, innovative suggestions from bidders on alternative methodologies, and would expect any approach to be flexible, and balance robustness with a need to rapidly produce findings. We would expect bids to include some elements of: Literature reviewing. To understand the existing evidence base around video consultations; digital-first approaches to primary care; and disruptive technologies. Qualitative research. To understand the experiences and outcomes of GP at Hand for users, non-users, and staff. This is likely to include some case study approaches. For example, to understand how the approach to care may differ between a typical GP practice and GP at Hand, potentially including a focus around specific consultation types both those that are particularly common (e.g. hayfever prescription), and those that might be particularly different or problematic in a digital first setting (e.g. identifying sepsis and antibiotic prescribing ; consultations with those with mental illness). Some direct consultation with service users, GPs, and primary care staff will be essential. Quantitative research. Some quantitative analysis will already have been undertaken by the CCG, and NHSE, and the evaluation would be expected to build upon this. In addition, we would expect evaluators to use primary, community, and secondary care datasets to answer the research questions. Bidders should include details and timelines of how they expect to gain access to this data.we would also expect appropriate and robust comparator groups to be established going beyond simple pre/post designs (particularly given individuals are likely to sign up shortly before they need access to primary care). We understand that establishing this counterfactual is likely to be very challenging; both methodologically, and in terms of securing access to data. NHS England s analysts would work closely with any successful bidder to develop an approach that is as robust as possible, whilst being deliverable within a reasonable timeframe. Economic research. Questions on the immediate financial implications of GP at Hand for CCG and other GP practice finances will be answered by internal 5

6 analysis. However, the external evaluator will be expected to build upon this research including, as appropriate, answering questions around overall resource use and cost-effectiveness. For all research methods, we would expect the evaluator to establish robust comparators and control groups. These counterfactuals should, as much as possible, take into account some of the nuance around the changes GP at Hand will be expected to implement. For example, comparing the GP at Hand service to the service that would otherwise be received, but also recognising that there will also be insight from comparing it to a benchmark for idealised primary care (e.g. as outlined in the General Practice forward view, one that makes use of a multidisciplinary team, uses digital technologies effectively, and works collaboratively with other local services). We would expect the evaluator to work closely with key stakeholders to develop a fair and objective definition of what this might be. Additionally, evaluators should be aware that the service GP at Hand offers, and the context it works in, is likely to change. For example, it may expand nationwide, or introduce greater triage through its Artificial Intelligence. Evaluators should therefore design a methodology and timeline which is flexible enough to respond to these changes. Governance and project management The evaluator will formally report to the evaluation steering group. This contains members from H&F CCG; NHS England London Region; and NHS England s Digital, Primary Care, and Operational Research and Evaluation Teams. The evaluator will provide regular updates to the group, and the group will review and sign off all formal publications. In turn, the steering group will report to the Primary Care Commissioning Committee of Hammersmith and Fulham CCG. The bidder will identify a project lead who will be the day-to-day contact for the work. They will be expected to work collaboratively with NHS England and CCG stakeholders to deliver the evaluation. Where bids are part of a consortium, lines of responsibility and approaches to quality assurance should be made clear. Evaluators should be prepared to offer flexibility in their approach; the GP at Hand service is likely to change, as are the associated policy priorities. Where research asks are more significant, and/or lie outside of the agreed evaluation questions, NHS England s Operational Research and Evaluation Unit will provide this analytical resource, but bidders should be available to support NHSE as appropriate with data, research findings, and critical appraisal. Bidders should demonstrate in their bids examples of where they have worked in this way in the past. Outputs and reporting We anticipate the evaluator producing three formal reports. This should include a scoping report (including detailed evaluation framework) at the end of the scoping phase; an interim report; and a final report. The exact timings of these reports are open to discussion; we would like them to be guided by when research findings are likely to be available. Some academic publications may be developed as a result of the research. In addition, there may be certain evaluative findings which are not 6

7 formally published, due to commercial sensitivities. We would expect to agree on approaches to this during the scoping phase. We would expect that the scoping report, which should include a full evaluation framework, is produced relatively rapidly. An interim report, containing some substantive findings on the GP at Hand service, should be available around the autumn time. The final report should include: a description of the GP at Hand service; detailed analysis of the outcomes and impacts of the service for users and non-users of the service, neighbouring practices, the CCG, the workforce, and the wider system. It should include a set of concrete recommendations for both the CCG and NHS England. Alongside the formal reports, evaluators should also provide fortnightly updates to the steering group. We would also expect the evaluator to budget a small resource for additional ad-hoc presentations, and short papers, to respond to reasonable requests from NHSE and the CCG. This will include providing some critical review of the second clinical review, due to report in July Timetable A final report is expected by March At the end of the scoping phase a scoping report, including full evaluation framework should be shared. An interim report will be expected in Autumn 2018, but exact timings are open to discussion. Budget Resources of approximately 200, ,000 (incl. VAT) will be available for this evaluation. The providers are asked to suggest the cost for this work, and provide a detailed breakdown of the budget by task and staff member. Evaluators are encouraged to present options for additional/ different analysis. Bids will be assessed on the value for money provided. Queries Please direct any questions to Holly Krelle, Senior Analytical Manager at NHS England, holly.krelle@nhs.net. In addition, we will hold a bidders meeting/ teleconference on 16 th March, where questions can be asked of the CCG and NHSE. 7

8 Activities and outputs Key research questions Below is a comprehensive list of research questions, developed through scoping consultations with key stakeholders. They are broken down into three overarching themes, related to: activities and outputs; outcomes and impacts; and policy implications and horizon scanning. This evaluation should focus on the first two themes, and to work with stakeholders to prioritise questions from amongst these. The questions on future policy direction are not the subject of this tender. Theme Who is accessing the service and what attracted them? To what extent to users understand the service, and its implications? How are patients accessing it? Questions To understand the cohort using it, and how this might differ from a usual practice. Demographics age, sex, SES/deprivation quintile of place of residence; Location; Conditions (e.g. LTCs, mental health) Past use of services primary, hospital, community, mental health service (will tell us about whether patients are high/low service users, and whether the service is meeting unmet need) Why did people join GP at hand? What attracted them, was it rapidity of access; video calls; having an immediate health need; or something else? For people that joined, did it live up to their expectations? What aspects did they particularly find positive and negative? Once registered, have any patients deregistered from GP at Hand and if so why? Who is not accessing the service? Why? Are there potential implications for health inequalities? How usable is the service? Are language, large text, etc. options available? How does this work for carers? To understand the extent to which users understand the GP at Hand service, and the implications it might have for their care: Do they fully understand what de-registering with a GP means? Do they fully understand they ll have to travel 45 mins or more for an apt.? Do they read, and fully understand the implications for how their data is used? To understand patterns of usage (proportions online v face to face), and turnover rates. No. and type of appts per person; Registration/ deregistration rates (ideally broken down by demographics/ location); 8

9 Outcomes and impacts OFFICIAL What are the average travel times to face to face appointments? What services are patients offered and using? What is the business model? What are the levels of satisfaction with the service? Are there any differences in clinical To understand how patients use the GP at Hand app and service: How does the AI triage system work? How confident are we in its outcomes? How do patients use it? What services are patients being offered? Who is staffing the service (e.g. are nurses, HCAs, AHPs used)? How is the service offer likely to change (e.g. introduction of nurse practitioners/ other members of an MDT)? Are patients offered any of Babylon s private services? When a patient is referred to another service or provider of care what is the method of handover and follow up, and how is continuity of care maintained? To understand the business and workforce model of GP at hand, including the extent to which it is sustainable: How is the GP at hand workforce structured? What is the breakdown of different professions? What is their level of experience? How do digital services synthesise with that of the host clinic? What are the business costs and overheads? What is the cost-per-acquisition of each patient? How might they generate profit? What scale do they need to operate at? Which elements of the service are likely to be profitable? To what extent could this model be replicated by other GP surgeries and/or other tech companies? What are the incentives for Babylon and GP at Hand, how might these differ from that of a usual GP practice? What are the cost structures within Babylon and GP at Hand, and how might these change as the business scales up? What are the levels of satisfaction and type of experience for: Users of GP at Hand; Comparable/matched non-users of GP at Hand; GP at Hand GPs (e.g. benefits of flexible working; issues around team working) Other GPs; Related health professionals. To understand/ assure whether GP at hand results in clinical outcomes that are as good as usual GP services: 9

10 outcomes for GP at hand patients? Are there types of patients is this most appropriate for? Do some patients benefit from, and value, this more than others? What happens when individuals develop illnesses that need longer-term care? What is the clinical: safety, effectiveness and triage accuracy of the service? How does this vary for a set of common consultations? To understand whether GP at hand results in increases, decreases or changes in the way patients use of health and social care services. What changes (if any) are there to patients use of health and social care services? Amount of use How does the number (and length) of contacts with GPs, nurses and clinicians in other practices change? How does the number of contacts with community and social services change? How does use of walk-in, urgent treatment, and A&E change? Are there any geographic patterns? How does use and prescription of medicines (particularly antibiotics and opiods) change? Is there an impact from supply induced demand? Has lowering the barriers to accessing GP advice increased use of services? Type of use How do relationships between: doctors & patients; doctors & wider practice staff; doctors & other services change? What effects does this have? How does the number and type (including to specialists, community and social services) of referrals change? Is there any difference in the quality or appropriateness of these referrals? What are characteristics of the GP at Hand and Babylon workforce? How does this differ to the wider primary care workforce? What are the workforce issues? What is the impact of the model on ways in which GPs work (eg team working, development, appraisal and mentoring)? Is there any effect of the GP at Hand on: Recruitment and retention of usual GPs; 10

11 Policy questions and horizon-scanning OFFICIAL Recruitment and retention of Babylon GPs; Training of new GPs? Primary care productivity (e.g. GP/other time per patient); Total primary care capacity? The incentives for GPs to work in a certain way? Clinical governance Effectiveness of MDT working? What are the financial implications? How does this fit with wider NHS policy, now and in the future? What options do NHSE/ CCGs have to effect change? How might this model change in the future? What other models are there for delivering a digital-first service? What is the impact of the model on indemnity, risk taking, and mistakes by GPs? Who is liable? Taking all the findings on changes in usage and workforce, does GP at hand offer value for money for: Babylon/ GP at Hand CCGs/ NHS England Patients To what extent might these VfM calculations change as the service scales up? What contractual implications does this have? What are the implications for the primary care and CCG funding formulas? How does this relate to place-based commissioning? Are there ways the two could be bought closer together (e.g. ACSs commissioning a GP at hand type service for their whole area)? What role (if any) could setting national prices (or methodologies) for digital primary care services play in driving the efficient delivery of these services, (and prevent sustainability issues arising for other primary services)? How might QOF change? What levers do NHS England, NHS Digital, and CCGs have to effect change and regulate this approach? What opportunities do NHS England, NHS Digital, and CCGs have to replicate, aid, or improve this approach? What happens/ what is the likelihood of (a) other individual practices piloting this and (b) other CCGs/ STPs develop the model? What changes is GP at Hand likely to make to its delivery model over the next year? Are there other models for delivering a digital-first service? (e.g. Hurley group, international models) What are the relative advantages and disadvantages of these approaches for patients, commissioners and GPs? How can other models be encouraged? How might other models adapt as a result of GP at hand? 11

12 What are the wider implications of GP at Hand What are the implications for the NHS of having independent providers delivering digital first primary care services? For productivity and innovation within the NHS? How is the GP at hand service, including the use of babylon technology, perceived internationally? 12

13 13