A short history of the GP contract and its evolution in England

The first contract in 1948 – the General Medical Services (GMS) contract – was held between individual GPs and the Secretary of State for Health and drew GPs into the NHS. 

This built on the terms of engagement of doctors providing care to workers under the National Insurance Act 1911. National Insurance doctors received approximately 50% of their income from capitation, topped up with additional payments. The GMS contract cut and pasted many of these terms. It required GPs to take 24/7, 365 days a year responsibility for patients, with penalties if not fulfilled. GPs also acted as ‘gatekeepers’ to specialist care, meaning that, except in emergencies, patients could only see a specialist if referred to one by their GP (Kmietowicz, 2006).

By the mid-1950s, GPs were feeling overworked (seeing around 50 patients a day plus home visits) and received lower pay than hospital doctors, resulting in a recruitment crisis. The Collings report on general practice in England in 1950, described as an ‘Exocet’ fired through general practice, laid bare the grim working environment, described inner-city practices as ranging from unsatisfactory to dangerous, and argued for the need to define the functions of general practice in the NHS, to raise its status compared with specialties and to form group practices. Drawing on Collings’ excoriating work, the recently formed (1952) Royal College of General Practitioners (RCGP) campaigned for better terms of engagement with the NHS and the Family Doctor Charter, enacted in 1966, is seen as the starting point for re-professionalising general practice and for boosting its standing within the health system. It introduced a new contract – set out in the ‘Red Book’ – which addressed many of the grievances of GPs about their working arrangements, provided a loose definition of services to be provided and included much detail on terms for service delivery, including capitation arrangements, reimbursement terms for premises and staff, and payments for other items including preventive work and pensions.

The next iteration of the contract was imposed on GPs in 1990 by then Secretary of State, Kenneth Clarke, despite GPs voting to reject it. It sought to address policy-makers’ concerns about quality and productivity and limited preventive care, linking the new contract to wider market reform of the NHS through the introduction of GP fundholding (Bower, 2018). The contract:

• introduced performance measurements for the first time
• extended roles for nurses related to chronic disease management
• increased the focus on preventive care through the introduction of health checks
• included the option of holding a ‘fundholding’ budget for some hospital services.

The 1990s also saw a new variant of the GP contract – the Personal Medical Services (PMS) contract – described further in the next section.

A decade later, the Department of Health still saw problems with the GP contract:

• not enough GPs
• inadequate access to appointments
• variable quality of care
• limited choice for patients
• difficulty transferring hospital outpatient services that could be delivered in the community with GP support.

Politicians sought a new contract with local flexibility and performance-based rewards, but the British Medical Association (BMA) fought for a new national contract alongside the optional, local PMS contract as the best way to retain GPs in the workforce and improve access to care.

The revamped 2004 GMS contract fundamentally changed the terms on which services were provided:

• moving the contract from individual GPs to practices
• ending 24/7 responsibility for care
• introducing a national micro-incentive system for quality improvement – the Quality and Outcomes Framework (QOF) – and the option of additional local (Local Enhanced Service [LES]) and national (Designated Enhanced Service [DES]) incentive schemes for specified activities.

The addition of a Minimum Practice Income Guarantee (MPIG) was required for the contract to be accepted by the BMA’s General Practitioners Committee (GPC). And another variant of the GP contract – the Alternative Provider Medical Services (APMS) contract – was launched, allowing general practice service delivery by others than general practice partnerships (see the next section for more detail).

In 2015, the NHS launched its ‘Vanguard’ programme of new care models, including the ‘multispeciality community provider’ (MCP) model for general practice services in which groups of general practices were encouraged to collaborate or merge together and to work with local community services. Although never rolled out nationally, the MCP contraction is included here because it can be seen as a precursor of current policy to involve general practices in integrated neighbourhood teams (INTs) with other local health and care organisations (NHS England, 2022).

The prototype MCP contract included three options for general practices to link to other services:

• virtually through alliance agreements
• partially integrated, in which participating providers sign a shared contract to deliver specific services but remain as separate organisations
• fully integrated, with this version creating an integrated care organisation in which GPs would be salaried employees (NHS England, 2016).

Fourteen MCP vanguard sites were launched to test the model. They varied in scale, in the focus of their joint activities and in their choice of MCP contract, with only one site – Dudley in the West Midlands – working towards the fully integrated model. The MCP contract ended when the Vanguard programme closed in 2018, without ever being developed into a formal GP contract option for all practices. It had demonstrated potential to deliver new services through a contract-driven collaboration between practices, but also highlighted the complex political and organisational challenges of building links between general practices and other providers, described further in the evaluation section below.

The 2019–24 GP contract framework built on learning from the Vanguard programme. It reflected growing interest in the potential for larger general practice organisations to remain resilient to the financial and workload pressures faced by smaller practices and to have the infrastructure, skills and financial resilience to innovate, improve and transform care. Rather than introducing an alternative contract for general practice networks, the framework introduced a voluntary national incentive scheme to the core contract – the Network Contract DES – which aimed to increase ‘provision of proactive, personalised, coordinated and more integrated health and social care’ (NHS England, 2019).

The DES required practices to form primary care networks (PCNs) with a combined list size of at least 30,000 patients, through collaborations or mergers with other practices through which to deliver  extended-hours services, provide additional general practice services to care homes and employ staff through the Additional Roles Reimbursement Scheme (ARRS). Further financial incentives were introduced through the Investment and Impact Fund (IIF) to drive quality improvement (for example, improving cancer detection). Although the DES was voluntary, it was associated with significant additional funding. In the first year of the contract, 99% of practices had signed up for the significant PCN DES payment, resulting in 1,250 PCNs forming across England. The five-year duration of the framework recognised the time needed to develop shared services and shift some elements of core general practice services out of individual practices and into PCNs (NHS England, 2023).

Alternative contracts for general practice services

The first major GP contract variant was the 1997 Personal Medical Services (PMS) contract. This locally negotiated contract was entirely optional, and aimed to increase services in under-doctored areas, meet specific local population health needs – including those of deprived groups such as homeless people – and increase the number and range of staff providing care (a small number of nurse-led practices were commissioned through PMS contracts). Local contract negotiations enabled commissioners (then in primary care trusts) to be more flexible with local practices in return for meeting set quality standards and tailoring the contract to focus on specific local population health needs. In 2022, 29% of general practices held these locally agreed contracts (NHS England, 2023) and £1.65 billion of the total general practice budget of over £8 billion was spent on locally commissioned PMS services (NHS England, 2023).

The Alternative Provider Medical Services (APMS) variant of the contract, launched in 2004, was the first to allow general practices to be owned and delivered by non-GPs, with potential providers including private companies and third sector organisations. Contracts were locally negotiated, and could specify additional services, such as longer opening hours or services for specific patient groups, beyond those in GMS and PMS contracts. Importantly, for the first time, private companies could hold contracts for general practice services. Subsequent policy initiatives to increase provision of general practice services (for example, the ‘Fairness in Primary Care Procurement [FPCP] process) and introduce a ‘GP-led health centre’ in every area were contracted through the APMS contract.

Bolt-on incentive schemes – LES and DES incentives

The 2004 GMS contract introduced two types of ‘bolt-on’ financial incentives for practices to deliver additional services on top of the core contract. Local Enhanced Services (LESs) are agreed with local commissioners to address local health needs, while Designated Enhanced Services (DESs) are national incentive schemes offered to every practice in England.     

Local – LES – schemes have become a standard part of the local general practice commissioning landscape. Each scheme is locally devised and implemented to tackle specific local health issues and participation is optional. Examples include additional payments to increase diabetes detection, record 24-hour blood pressures and manage anticoagulation medicines.

National – DES – initiatives have to be offered to every practice and participation is also voluntary. They are a regular feature of annual GP national contract negotiations, and used to incentivise the formation of PCNs.

The combination of LES and DES incentive schemes and QOF incentive payments creates a significant task-focused workload for practices and PCN staff. A review of incentives in the GP contract is underway and due to report soon (Department of Health and Social Care, 2024). 

Evaluations of the impact of GP contract changes

Each contract iteration or variation has sought to tackle perceived problems with the previous contract and achieve improvements in different aspects of general practice care (summarised in Table 1). There have been many national and local contract evaluations shedding light on the extent to which each iteration has achieved its intended effects, with the main national evaluations summarised below.

The National Audit Office (NAO) evaluated the 2004 GMS contract in 2008, reporting against its multiple objectives. The evaluation concluded that its aim of attracting 1,950 new GPs was exceeded and the contract enabled patients to have more choice of GP and appointment length. There was some improvement in infrastructure and premises, a modest increase in the range of services offered in general practice and a slight simplification of administrative processes. However, the expected increase in productivity did not occur, with a fall in quality-adjusted productivity of over 2% between 2003 and 2005, and the observed increase in activity (number of patients seen) was at a lower rate than that of increasing costs.

The report also described initial improvements in GP morale, associated with greater job flexibility and being able to opt out of out-of-hours work, but these were not sustained, mainly because of subsequent public criticism of the reported increase in GP income arising from the new contract. Finally, the NAO noted that the Minimum Practice Income Guarantee (MPIG) had prevented the expected redistribution of resources to areas of higher deprivation and need. And the expectation that primary care trusts would commission LESs to reflect local health needs was thwarted by a significant overspend, mainly due to high achievement of quality payments and the high cost of new  out-of-hours services.

A full review of the impact of QOF is beyond the scope of this article, but various evaluations of it provide a mixed picture of its impact. For example, Campbell and others (2009) reported improvements in asthma and diabetes care and no improvement in heart disease care two years after the introduction of quality payments, and a decline in quality of care for non-incentivised activities. Ryan and others (2016) reported no significant impact on overall or condition-specific mortality associated with QOF, although more recent research reports that patients achieving QOF targets for type 2 diabetes have lower mortality rates and fewer complications (Majeed and Molokhia, 2023). And a 2017 systematic review of the role of QOF on long-term conditions by Forbes and others reported modest improvement in diabetes care and a modest slowing in emergency admissions for people with severe mental illness. However, Marshall and Harrison (2005) have highlighted some of the problems with large-scale incentive programmes. They argued that the use of targets and financial incentives may have unintended consequences on practitioner behaviour, such as goal displacement and rule following, leading to the ‘crowding out’ of and reduction in focus on non-incentivised tasks, as evidenced by Campbell and others. And researchers in Scotland found reduced performance in 12 of 16 quality measures studied one year after withdrawing QOF financial incentives and in 10 out of 16 measures three years after QOF was abolished (Morales and others, 2023).

A national evaluation of first-wave PMS pilots reported on several effects of the contract, including  an improvement in access for population groups often poorly served by general practice, a changing  skill mix with more care provided by nurses and modest improvements in quality of care. And the pilots introduced a cadre of salaried GPs at lower cost than general practice partners and more productive, with little or no impact on quality of care (Lewis and Gillam, 2002). Campbell and colleagues’ 2005 mixed-methods evaluation of the quality of care in a cluster of PMS practices compared to GMS practices reported only marginal differences in quality  across multiple quality measures, with only angina and elderly care statistically significantly better. A cluster of nurse-led practices launched through the PMS contract to improve access for vulnerable patient groups was evaluated by Lewis and others (2001), who described broad success in delivering care to vulnerable patients. The services were generally valued by users, although they identified some hostility towards PMS practices. In terms of the additional cost of these services, NHS England reviewed PMS contracts in 2014 (NHS England, 2014) when 22% of practices held the contract. There was no evidence that deprived areas received more PMS funding despite reducing inequalities being a core objective of the contract and the review identified £325 million of additional cost compared with GMS practices. The review linked £67 million of the excess to enhanced services and key performance indicators but could not define specific benefits associated with the remaining £248 million, suggesting the specification of PMS contracts was weak.

A 2015 evaluation of APMS contracts across 17 quality indicators found that, 10 years after the contracts were first launched, 4.1% of practices were held by alternative contract providers. They tended to serve younger, more diverse and more deprived communities than other contracted providers – which was consistent with part of the rationale for introducing the contract. The study also reported that APMS providers performed worse than other practices on 15 out of 17 indicators, albeit that many were serving more deprived communities where, more broadly, practices across contract types tend to perform less well on quality. And a qualitative study by Coleman and others (2013) highlighted the administrative burden of commissioning APMS services, noting how the transactional nature of this process contrasted with the more relational methods observed for contracting with traditional practices. But the authors also suggested that the perception of competition for GP contracts had ‘led existing practices to improve their services’ and offer longer opening hours.

Early and small-scale studies of the formation of PCNs described their formation and the work needed to establish the networks, and provided a mixed picture of their early achievements. Morciano and others (2020) reported that they had formed quickly, albeit with variable structures and processes, driven by the need to work together in response to the Covid-19 pandemic. A 2022 BRACE research collaborative evaluation described policy and financial incentives as key reasons why practices joined PCNs and reported that their early development was easier where prior collaborations had been in place, albeit that these were sometimes also a source of tension. The researchers described the need to:

• increase general practice involvement with PCNs through setting clear goals of local relevance, adapted if necessary for the needs of rural communities
• build leadership and management capacity, which was lacking in many PCNs
• clarify the role of PCNs in the wider health system
• strengthen monitoring and performance management.

An early qualitative study of seven PCNs by Checkland and others was consistent with the BRACE evaluation mentioned earlier, concluding that to continue developing and making an impact, PCNs will need considerable management support and that those in which practices had previous experience of working collaboratively were at a significant advantage. And drawing on interviews with 16 national policy stakeholders, the researchers also noted the multiple overlapping policy objectives associated with forming PCNs, each of which needed a different approach to delivery (Checkland and others, 2020). They argued for ‘temporal sequencing’ of objectives, with an initial focus on stabilising primary care before progressing to other developments and for crafting delivery requirements so all stakeholders see opportunities to meet favoured objectives as a way to avoid stakeholder dissatisfaction when implementation details do not match expectations.

Although never adopted as a national contract model for general practice, there is useful learning available from an evaluation of one of the most complex MCP vanguard initiatives in Dudley (Midlands and Lancashire Commissioning Support Group and Strategy Unit, 2016), which is included here because of its relevance to policy aspirations to draw PCNs into integrated neighbourhood teams. The evaluation described the potential of an integrated care provider contract to link general practice and community services with a whole-population budget, and shared population outcome measures to align incentives across all providers. The evaluation reported early progress in understanding population health but limited implementation of new service delivery models. It found that the procurement process for an independent care provider contract was so complex that it:

• was costly to run
• created divisions between participating organisations
• struggled to deal with the complexities of the supplier market.

The evaluation also highlighted the importance of continual communication between commissioners and providers to inform and engage organisations participating in contract delivery, and the need for continued work with providers after procurement to provide development support, given the level of programme and project management capacity needed for implementation and assurance.

Conclusion

Whatever contract type is developed to commission services through general practices, PCNs and integrated neighbourhood teams, there is valuable learning from past contracts. The PMS contract is already available as a local contract but did not, in its early stages, provide measurable value for money. The APMS contract (and the prototype MCP contract, which was never rolled out) highlight the administrative burden of developing alternative contracts, which must be funded and delivered as part of the contracting process. This will not be easy, with the centralisation of primary care commissioning in integrated care systems. And past experience with the micro-incentives of QOF and various national incentive programmes reminds us of the potential bear traps and unintended consequences of micro-incentive systems.