Reforming general practice in England: What can be learned from the 2018 Scottish GMS contract reforms?

The 2018 Scottish General Medical Services (GMS) contract was ambitiously designed to address deep structural challenges and reflected a consensus between those negotiating on behalf of GPs and government. Its main aims were to reduce GP workload and to narrow health inequalities. It intended to achieve these objectives through methods such as expanding multidisciplinary teams (MDTs), transferring certain services out of general practice, and removing the Quality and Outcomes Framework, instead incorporating GP clusters into the wider reform programme.

Eight years on, an independent audit, the academic evidence and frontline experience all point the same way: the contract’s central promises have brought some benefit but remain largely undelivered. In the main, the shortfalls were not failures of intent – instead, they appear to be failures of implementation, resourcing and measurement, with further unintended consequences. For policymakers in England, the value of Scotland’s experience therefore lies in how to do large-scale reform.

Specifically, there are lessons to be learned about the importance of evaluation and distributing resource according to need, as well as cautionary tales about transferring services out of general practice and the employment of practice-attached multidisciplinary staff. This briefing aims to share these lessons, so that England can learn them before embarking on reform of its own.

What we did

This briefing draws from three bodies of evidence:

1. The Audit Scotland Publication: General practice: Progress since the 2018 General Medical Services contract
2. Published academic work, cited throughout
3. Anonymous interviews conducted in late 2025 with eight individuals who had experience of designing, implementing or evaluating the 2018 contract reforms (including people who worked for or with the Scottish Government, the BMA Scottish GPs Committee, Audit Scotland, RCGP Scotland and the Scottish Deep End Project).

It is designed to give an overview of what has happened in Scotland over the past eight years, focusing on the lessons that England can learn from it.

Why this matters now

The UK government intends to embark on major reform of the GP contract in England within this parliamentary term. Scotland is the closest comparable UK system to have attempted major contract reform in recent years. It is also far enough through the process that both an independent national audit and peer-reviewed evaluations exist.

The 2018 GMS contract was Scotland's first distinctively Scottish GP contract, and it rested on a clear theory of change: freed from work other clinicians could do in primary care, GPs would have the capacity to operate as expert medical generalists. Workload would reduce, and health inequalities would begin to narrow.

The original commitments were due by April 2021; many were not delivered, a revised deadline of April 2023 was set, and several remain outstanding today. 

What the 2018 contract changed

The 2018 GMS contract was the first contract negotiated specifically for Scotland. It was agreed between the Scottish Government and the Scottish General Practitioners Committee (SGPC) of the BMA and was underpinned by a Memorandum of Understanding with Integration Authorities (which jointly plan health and social care services), and NHS boards. The contract proposed “a refocusing of the GP role as expert medical generalists”, concentrating GPs on undifferentiated and complex care and on clinical leadership of a wider team. The main proposed reforms were as follows.

• Multidisciplinary teams and service transfer. An expanded multidisciplinary team (MDT), employed almost exclusively by NHS boards but also with some staff within the practice, was to work across six priority areas to transfer care out of general practice: vaccinations, pharmacotherapy (providing pharmacy and prescribing support for general practices), community treatment and care services (CTACS – a range of services such as wound care, some vaccinations, phlebotomy and chronic disease monitoring), urgent care, expansion of professional roles, and support from community link workers. For additional professional roles and community link workers, this was substantial new capacity rather than the transfer of existing general practice work. For the remaining four, the intention was to transfer responsibility from GPs to Boards which already led the local design and delivery of a range of health and care services.
• Funding reform. A new practice funding formula (the Scottish Workload Formula), intended to better reflect practice workload, alongside a practice income guarantee to stabilise income, was introduced through a phased transition beginning in April 2018.
• Premises. A move away from GPs being expected to provide their own premises, as part of efforts to reduce risks for GP partners. GPs who own their premises would be eligible for an interest-free loan of up to 20% of property value, intended to reduce ownership risk, remove barriers to partnership and support eventual transition towards NHS-board ownership. GPs who lease their premises would be eligible to apply for a mix of schemes aiming to move responsibility for lease-holding from GP partners to NHS boards. 
• Quality. The Quality and Outcomes Framework (QOF) - a pay-for-performance scheme incentivising GPs to provide high-quality care by paying them for meeting specified quality indicators - was discontinued and replaced by GP clusters intended to drive peer-led quality improvement. Scotland abolished QOF and introduced GP clusters in 2016/17. The 2018 contract built on and embedded these earlier reforms.
• IT and information sharing. New GP clinical IT systems, and information-sharing agreements recognising GPs as joint data controllers with NHS boards.

Scottish government also committed to an additional 800 GPs (by total headcount) between 2017 and 2027.

A ‘Phase 2’ of reform – with a greater emphasis on assessing and reimbursing expenses – was planned, but has not progressed.

What happened?

GP workload

Reducing GP workload was one of the contract’s headline promises. The available evidence suggests workload did not reduce during this period. The national GP survey comparing 2018 and 2023 demonstrates how GP-reported job pressure rose by a statistically significant yet modest amount, while overall job satisfaction was unchanged. More GPs reported planning to reduce their hours or leave direct patient care. However, these changes coincided with the pandemic, and a similar fall in satisfaction in other countries. 

The reduction in workload was expected based on the combined effect of additional GPs, and moving tasks previously done by GPs to MDTs with staff employed within and outside practices. Instead of the pledged 800 GP expansion, between March 2017 and March 2025, total headcount, excluding trainees, rose by 197 (from 4,385 to 4,582) and estimated whole-time equivalent rose by 72 (from 3,520 to 3,592). The number of patients each full-time GP cares for rose accordingly, in part due to population growth. 

In the national GP survey in 2024 GPs were generally positive about the initiative aiming to expand use of MDTs to reduce GP workload, but fewer than half reported that it had succeeded. Some reported the opposite: training and supervising new team members, often early-career staff, many working in a new setting, added to their load rather than relieving it. The picture is complex, however, as there were rises in GP workload for other reasons too: the ageing demographic, more complex care with rising mental health care demand especially, and a transfer of some workload from secondary care (some inadvertent due to outpatient and treatment delays post-pandemic). 

Measuring GP workload over time could have been a way to evaluate whether a reduction in GP workload had been achieved. This did not happen. A single national IT system may have facilitated this, but the provider went into administration. Public Health Scotland has since begun collecting activity data, but robust, comparable national workload and activity data were unavailable at the point of evaluation. The result is that the contract’s flagship claim can be neither substantiated nor refuted. As one stakeholder put it, "if you want to introduce a policy, you have to be able to show how you will measure it".

Health inequalities

Evidence suggests that the 2018 contract did not tackle inequalities in patient experience. Patient satisfaction fell across the whole of the UK over this period – a decline widely linked to the pandemic and its aftermath, rather than to any single contract – and Scotland was no exception. Drawing on the national Health and Care Experience survey, the share of patients rating their practice's care as ‘excellent’ or ‘good’ fell by 14 percentage points overall between 2017/18 and 2023/24 – from 83% to 69%. Experience tended to be worse in poorer areas. In 2023/24, patients in the most deprived areas rated care positively less often than those in the least deprived (63% against 71%). A peer-reviewed analysis of the survey over six waves (from 2011/12 to 2021/22) found that this gap actually widened over the decade, and that the 2018 contract does not appear to have altered the downward trajectory.

One of the most striking findings comes from work comparing GP working life in 2018 and 2023. Across job satisfaction, job pressure and other measures, working life improved for GPs in the most affluent areas. But in the most deprived areas, job pressures rose. This may suggest that the contract “worked” better for GPs in affluent areas and “did not work” for those in deprived ones.

One possible reason for this divergence in both patient experience and GP experience is that new resource was not adequately distributed according to need. Notably, the principle of proportionate universalism – offering a universal service, but doing proportionately more for those with greater need – was largely absent from the contract’s design and delivery. Nationally, MDT roles were allocated without an explicit equity lens, and the Scottish Workload Formula has been criticised: it draws on older activity data that did not fully represent practices in deprived areas. Interviewees outlined that they do not believe it is sufficiently weighted for deprivation. The formula heavily weighted towards older people, meaning that some practices in the most deprived areas serving people with a low life expectancy did not benefit much - if at all - from the initial (and recurring) increases in funding.

Work funded by The Health Foundation reaches the same conclusion: tackling the inverse care law in general practice requires explicit, sustained, targeted investment in deprived-area practices – not the assumption that universal reform will reach them in proportion to need.

Some bright spots are worth highlighting in relation to health inequalities. Community link workers were seen by some interviewees as a genuine success in practices in Scotland’s most socioeconomically deprived areas (so called ‘Deep End’ practices), providing a route to addressing the social and economic dimensions of patients' presentations that GPs alone could not. However, the Community Link Worker programme was not one of the services chosen for prioritisation in 2021 (unlike vaccinations, CTACS and pharmacotherapy). This meant that they effectively had to compete for funding, alongside services earmarked to be prioritised, within a finite resource allocation for the MDT element of the 2018 contract. This had a disproportionate impact on areas of deprivation (where community link workers are recognised to be a more essential part of the core general practice team), and this, combined with short-term funding and contracts, have made the service vulnerable to budget cuts.

There has been some targeted funding, but this has been separate to the contract itself. The Inclusion Health Action in General Practice (IHAGP) provided funding for 66 general practices within NHS Greater Glasgow and Clyde, where most of Scotland’s “Deep End” practices are located. And in 2025, a new national Enhanced Service was developed for cardiovascular disease, which was designed with health equity in mind. Practices in areas of greater deprivation had more eligible patients, and therefore received more funding. Practices were encouraged to proactively consider how to reach patients who were traditionally missing from their service. However, these examples happened at the margins rather than being a core component of the contractual change with long-term funding commitment. 

Transferring services out of general practice

Of the services planned for transfer out of general practice, vaccinations is the only one to have been mostly moved, though a few areas still operate a mixed model.  The previous arrangement had resulted in variable outcomes: there was performance-related funding, but vaccination uptake was dependent on how proactive each practice chose to be. Many GPs welcome that they no longer deliver an increasingly onerous and complex programme, which regularly changes, and that has freed up nursing time. However, stakeholders reported that the programme now costs substantially more than when practices ran it; that uptake has fallen with a widening of inequalities (partly societal, partly delivery problems); and that delivery of seasonal vaccinations can be slow. Opportunities to carry out ‘opportunistic’ vaccination have been lost, particularly in deprived settings. The model has been particularly challenging in remote and rural areas. For example, the Highland region has moved to a mixed model, with some practices resuming vaccination delivery. The other unintended consequence of CTACS and transferring vaccines is that the complexity of practice nurse workload has increased with the reduction of similar procedures. 

There is a structural lesson here that policy makers in England should not miss: amending regulations transferred responsibility to boards, but without service specifications setting out what level of service patients are entitled to receive (for example, how many appointments should be offered) workload can still fall back to practices.  

Overall, it appears that transfer can be justifiable in principle, but it is operationally challenging and requires careful design, delivery and monitoring, and needs to specifically take account of local circumstances, including deprivation.

Clusters and quality after QOF

Replacing QOF with peer-led GP clusters appeared to be sound in principle, given that extensive evaluation of QOF has shown that it is not associated with improvements in mortality; is not cost-effective; and while the scheme initially improved health outcomes in some conditions, over time outcomes returned to pre-existing trends. In practice, clusters have not progressed as intended. Though there are pockets of excellent work leading to clear changes in outcomes, stakeholders described clusters as having neither adequate power nor money, and a lack of easily accessible data insight was a major limitation.

The loss of QOF also removed a quality-assurance mechanism, a data stream, and evidence-based insights without putting alternatives in their places. For the period when indicators were still being recorded, there is evidence of decline: a study comparing Scotland with England (which retained QOF) found that recorded quality of care fell significantly on 12 of 16 indicators one year after Scotland withdrew the financial incentives, and on 10 of 16 after three years. It is unclear whether these changes in documented performance reflect actual changes in clinical outcomes.

Stakeholders pointed to a real-world equity consequence of this: greater 'missingness' – the loss of the proactive recall of underserved patients that QOF used to incentivise – which is likely to widen health inequalities and variation in care. There is now a move in Scotland to a more considered ‘core’ quality data set. 

Funding and the employment model

Across much of the period since 2018, the share of funding going directly into core general practice fell while indirect funding for general practice rose. Stakeholders cited a fall in general practice’s share of the health budget to well below previous levels:  between 2021/22 and 2023/24, real-terms direct spending on general practice decreased by 6%. BMA Scotland also estimates a funding gap of around £290 million - roughly 1% of the Scottish NHS budget – which it attributes to year-on-year real-terms erosion of per-patient practice funding since 2008. Stakeholders also reported substantial GP capacity sitting unused in the locum market in 2025 due to a lack of core funding to employ it. Recognising GPs concerns about ongoing underinvestment, in October 2025 the Scottish government announced additional funding of £531m over three years to help boost practice-employed staff numbers and capacity, support day-to-day operations and improve access to GP services. 

The NHS board-employment model for MDT staff does have potential strengths: it removes financial and employment risk from individual practices, and government points to a substantial expansion of staff now working in these services. It also allows for flexibility in terms of allocating (and reallocating) staff to practices across localities depending on workforce and other factors, including hub-working. Audit Scotland reports more than 4,900 whole-time equivalent (WTE) staff working across the six priority services by March 2024, of whom just over 3,500 WTE were additional staff recruited specifically to expand MDTs. But it comes at a cost to integration. Around seven in ten GPs reported a lack of influence over the roles of MDT staff in their own practice; others cited problems with space and the time required for supervision. A recurring theme was a culture clash – for example, a CTAC nurse based in a practice but line-managed from outside the practice, working to protocols that are not integrated with practice routines, may take three times as long to take a blood sample. 

Premises

Scotland’s ambitious attempt to take GP premises into public ownership has been faltering. GPs faced multiple barriers to applying for loans, including the high cost of legal fees and securing agreements with mortgage providers. Scottish government was criticised for delaying the scheme, such that no loans were issued until 2020/21, and for over-promising – before restricting – the amount of funding made available. The scheme was paused in March 2024 because of oversubscription, with just £15.1m in loans having been issued over five years. Scottish government subsequently identified budget to re-enable the scheme in 2025/26, and intends to launch a second phase over two years from 2027/28, but there have been delays to this process too. 

Lessons for England

We now draw the above threads together into lessons England can learn from Scotland’s experience. Each lesson describes a case where Scotland's experience has been assessed as specific enough, and consistent enough across the audit, the academic evidence and frontline accounts to carry direct lessons forward into the choices facing English policymakers.

Lesson 1
Build the evaluation infrastructure before you reform, not after

This is a foundational lesson on which the others depend: you cannot tell whether any reform is working without first being able to measure its impact. Scotland changed the system without having sufficient data and evaluation capabilities that would have allowed it to judge whether the change was successful. The result is that we cannot say with confidence whether the contract's flagship promise – reducing GP workload – has been achieved. Most claims rest on qualitative evidence, assembled piecemeal after the fact. In June 2026, Healthcare Improvement Scotland published a review of a phased programme of investment into primary care. The findings highlight the lack of sufficient data for attributing short-term changes to policy change, and that existing data proved insufficiently robust for detailed economic analysis.

England should specify, fund and stand up a measurement framework as part of reform itself, not as an afterthought. That means pre-specified outcome and process measures should be created – these could cover elements such as workload, workforce, access, continuity, patient experience and clinical outcomes – with a clear baseline taken before reform begins. Measures should be stratified by deprivation from the outset, so that average improvements cannot mask widening inequality. Ownership of the framework should be clearly assigned, with regular publication built in. A simple principle should be borne in mind: if a goal is worth setting in the contract, the means of measuring it should be detailed in the contract too – ideally in a way that doesn’t increase administrative burden on frontline staff.

Lesson 2
Distribute new resource by need, explicitly, or expect the gap to widen

Distributing resource equally is different to distributing resource equitably. The latter means distributing it in accordance with the need of the population, with some populations therefore receiving more resource than others. Funding and workforce distribution in general practice and MDTs was done without an adequate, explicit equity lens in Scotland. This was associated with a widening gap between richer and poorer areas in terms of patient experience, as well as improving working life for GPs in the most affluent areas while those working in the most deprived areas experienced increased job pressures. This shows the inverse care law is still operating, despite the aim of the reforms to address health inequalities. 

If England is serious about tackling health inequalities, it should treat equitable allocation as a design principle, not a downstream adjustment. This means weighting any expansion of workforce and funding towards need, defined as the need to benefit across the life course, from the outset. The Scottish experience suggests that proportionate universalism applied retrospectively is not a substitute for proportionate universalism applied proactively by design.

Lesson 3
Specify the service, not just the responsibility

Moving a service "out" of a GP practice in legal and regulatory terms does not move the work, unless what will actually happen in practice is defined in detail. Scotland transferred legal responsibility for vaccinations, pharmacotherapy and community treatment and care to NHS boards, but in the latter two of these three services, no service specifications were issued. The work therefore continued to fall on practices, often without commensurate funding or recognition. 

Any planned transfer of responsibility within England's reforms should be paired with clear service specifications setting out what practices and patients are entitled to receive from each service. All parties should be clear from the outset, not just about what is expected but about how those expectations will be delivered, and clearly communicated with the public.

Lesson 4
Treat the removal of established services from general practice as high-risk

Vaccination is Scotland's clearest cautionary tale. Stakeholders claim that the service now costs more than when practices ran it, while vaccine uptake has reduced. For policymakers there may be a justifiable argument for transfer in principle (the previous arrangement produced inconsistent uptake and arguably unwarranted variation between practices), but the operational reality has been very challenging.

If England is to consider shifting activity out of practices, the lesson is not necessarily to avoid it, but to treat transfer as a high-risk option that requires careful consideration and planning. In practice, that could mean:

• piloting at meaningful scale before national rollout 
• measuring uptake, cost and patient experience against the status quo, with particular attention to uptake among communities that experience health inequalities
• building in the ability to reverse or hybridise the model where it is not working
• managing seasonal demands for services
• retaining flexibility for practices to continue delivering services where that produces better outcomes, particularly in rural and remote areas. 

Lesson 5
Resource the enabling work

Workforce policy focused on numbers and composition is not enough. In Scotland, the most cited barriers to making the MDT model work were not the funding for the staff themselves. Instead people pointed to the conditions that surrounded them: problems with having enough physical space to host staff, in addition to the time required for training and supervision of new staff (many having only received training or experience in specialist or other settings), and being unused to managing higher levels of risk or uncertainty in the community. 

The implicit assumption that MDT staff could simply be slotted into existing practices – or that they would arrive trained, supported and ready to integrate – was not borne out in practice. Some of the workload "reduction" promised never materialised because existing staff were absorbed into the supervision and integration task.

England has experienced similar problems with the Additional Roles Reimbursement Scheme, and any further reforms should look to resource the enabling work explicitly. This may mean upfront investment in practice estate to accommodate expanded teams; protecting and remunerating supervision time for existing GPs and nurses taking on new roles; and ring-fencing organisational development funding to help practices adapt their ways of working. 

Lesson 6
Decide where staff are employed and managed, and weigh integration against risk

Where MDT staff are employed shapes whether they genuinely integrate with general practice or merely coexist with it. Scotland chose health board-level employment, which reduces financial and employment risk to individual practices at the expense of integration. One downside is that GPs reported a lack of influence over the roles of MDT staff in their own practice, and a recurring theme was the cultural and operational friction of staff who sat in practices but reported elsewhere. 

Currently, England's PCN-based model gives practices closer involvement with the staff working alongside them. This may soon be up for debate again with the introduction of new neighbourhood contracts and neighbourhood health centres.

The lesson here is to be conscious of the trade-offs that exist within different employment models. Whatever the employment route, practices should have meaningful influence over the design and selection of roles working alongside them. Where strategic alignment with the wider system is needed – for instance to align with neighbourhood health structures – that could be best achieved through dual accountability rather than through removing practice-level operational control. And funding should not be conflated with employment: it is possible to channel funding through higher-level structures such as boards or commissioners, while leaving employment, line management or operational direction and discretion with practices. The English reform offers a chance to make this design choice consciously, with Scotland's experience as a guide rather than a template.

Conclusion

Despite the challenges faced, Scotland’s experience does not mean that England should avoid reform. Scotland's 2018 contract was a serious attempt at tackling the right problems, built on a genuine settlement between profession and government, and several of its aims were sound. The lesson is that reform on this scale is very difficult, and the gap between a well-intentioned design and a system that actually delivers is wider than most people expect. 

What the Scottish experience offers England is a chance to close that design/delivery gap deliberately.  Reforms should:

• Build the means of measurement before changing the system, so that what is promised can be evidenced
• Distribute new resource by need from the outset, or accept that the gap will widen by default
• Specify services to an appropriate degree, not just where responsibilities should sit
• Treat the removal of established work from practices as a high-risk option.
• Resource the enabling work that turns design into delivery
• Be conscious of where staff are employed and directed from, because that choice shapes whether new teams integrate effectively.

None of this is a counsel against ambition. Rather, it is a counsel for ambition, matched by proactive and detailed consideration of the delivery challenges and unintended consequences that may lie ahead. Scotland's experience is not a warning against reforming the GP contract; it is a guide to doing so with eyes wide open.