General practice across England: What should the government take away from these findings?

The large decline in patient satisfaction since 2012 – particularly since 2019 – should concern the government. The previous government quietly spent a lot of time and money reforming primary care, all while patient satisfaction declined. This section looks into what this government – not long off a year in office – should be looking at as it embarks on its programme of NHS reform, specifically to improve the picture in general practice. 

Most national trends are in the opposite direction to what patients value in general practice 

The three factors that we have found are most closely associated with patients’ experience of general practice are: 

1. More GPs – particularly partners – per patient
2. A higher proportion of appointments delivered face-to-face
3. Smaller patient list sizes. 

In addition, all three of these factors are positively related to the proportion of QOF indicators that a practice achieved in 2023/24. 

Yet all three of these trends have nationally been going in the opposite direction to the way that would improve patient experience. There were 13.8% fewer GPs per patient in December 2024 than in September 2015. Staff in general practice delivered proportionally fewer appointments face-to-face in 2024 (66.2%) than in 2019 (80.7%). The list size of the median practice grew, from 6,199 weighted patients in 2013/14 to 8,262 in 2022/23, an increase of 33.3%. 

Conversely, the enormous expansion of the DPC workforce, the last government’s signature primary care policy achievement, has coincided with the largest drop in patient satisfaction on record. Our regressions also showed a negative relationship between the change in DPC staff and the change in satisfaction between 2019 and 2023. In other words, the larger the increase in DPC staff, the more likely it was that patients’ satisfaction with a practice would fall. 

While patient satisfaction is by no means a perfect proxy for service performance, the findings of this report should give the government pause.

The current GP workforce plan will not increase the number of appointments or patient satisfaction as much as other approaches 

The current iteration of the NHS Long Term Workforce Plan (LTWP) forecasts that the number of GPs (both fully qualified and trainee GPs) working in primary care will increase from 37,841 in 2024/25 to 51,533 in 2036/37, a rise of 36.2%. This is clearly intended to increase access to general practice as demand is expected to rise. 

But that increase is driven almost entirely by rising numbers of GP trainees. Trainee GPs will more than double in that time, from 10,974 to 22,605. At the same time, the number of fully qualified GPs is only expected to increase by 7.7%, from 26,867 to 28,928.

That is a reasonable assumption. Since 2019, the number of GP trainees has almost doubled, while the fully qualified workforce has remained relatively stable. This modelling is therefore a continuation of existing trends. 

We can estimate the effect on the number of GP appointments by using the results of our analysis from regressions 3 and 4. Using those estimates, the current plans for the increase in GP trainees and fully qualified permanent GPs between 2023/24 and 2036/37 will result in 23.5 million more total appointments in general practice in that time, of which 20.5 million will be GP appointments with the remaining coming from other practice staff.

In contrast, if the government and the NHS reduced the implied leaving rate in the LTWP from just over 10% of fully qualified GPs per year to the average leaver rate of 7.4% since June 2023, then there would be 6,486 more fully qualified GPs in the workforce by 2036/37 compared to the baseline scenario (35,414 vs 28,928). In turn, that would result in 27.6 million more GP appointments in general practice, compared to the current LTWP assumptions, and 4.7 million more appointments from other practice staff. 

In total, therefore, there would be 33.1 million more appointments, including the increase in appointments delivered by ‘unknown’ staff members.*

The impact on access to general practice from these different scenarios is stark. It also raises the question of value for money. It is expensive for the NHS to train a GP. The government will achieve much better value for money if trainees spend years or decades in the NHS after qualifying than if they either never join the fully qualified workforce, or else leave soon after joining. 

The government should focus on addressing issues that discourage trainees from joining the fully qualified workforce and which push them out when they do eventually join. This will result in more appointments in the long run, higher patient satisfaction as a result, and better value for money.

_{* For more information on the assumptions underlying this calculation and the calculation below, please see the Methodology in the PDF version of this report.}

The government urgently needs to address the crisis in the GP partner workforce 

This work shows that GP partners are the members of staff associated with the largest increases in patient satisfaction, appointments and achievement against QOF targets. Larger patient lists – which are associated with lower patient satisfaction – are also at least partly due to declining partner numbers: fewer GP partners lead to more practice closures and mergers, resulting in larger patient lists in the remaining practices. 

There is a more subtle point about the importance of GP partners to general practice. GP partners deliver not only the most appointments themselves but are also associated with greater increases in appointments among other staff members. In other words, they likely play an effective management, co-ordination and resource allocation role. This is probably due to their incentives to ensure that the practice runs efficiently and within its budget. 

Finding a way to reverse the decline in GP partner numbers would be one of the most effective ways that the government could increase access to general practice without piling more pressure on existing GPs. 

But as we have shown, there has been a steady erosion of GP partners from the service since at least 2016: there are roughly a quarter fewer GP partners in 2024 than in 2016. The situation is worse among younger GPs where there are now half as many partners under the age of 40 compared to 2016. 

If the government intends to continue to use the partnership model to deliver general practice, it needs to urgently address some of the factors that make partnership less attractive. They include: the high cost of entering partnership, the unlimited financial liability that comes with most partnerships (particularly related to the ownership of premises), fears of being the ‘last partner standing’, and increasing workloads. ¹⁸ Department of Health and Social Care, ‘GP Partnership Review’, January 2019, p. 14, https://assets.publishing.service.gov.uk/media/5c3ca241ed915d50b4b47223/gp-partnership-review-final-report.pdf  

If the government thinks that the partnership model is no longer a feasible way of delivering general practice, then it should assess whether other models could be more appropriate and plan accordingly. If current trends continue, particularly given the tendency for younger GPs to reject partnership, there is a risk that it will become impossible for the government to deliver general practice with the existing partnership model. The government should have options in place if that does become more likely.

There are questions about the efficacy of the expanded DPC workforce 

Since 2019, the centrepiece of government policy in primary care has been the addition of DPC staff. That agenda has been incredibly successful, with almost 50,000 DPC staff now working in primary care. DPC staff have delivered most of the increase in appointments in general practice.

But as we have shown, patients do not report higher satisfaction in areas with more DPC staff, or when the number of non-GP appointments increases. Outside of our findings here, there are also concerns that DPC staff require substantially more management by GPs (particularly GP partners), which may reduce the time that they have available to deliver appointments. ²⁰ Baird B, Lamming L, Bhatt R and others, Integrating additional roles into primary care networks, The King’s Fund, February 2022, p. 26, https://assets.kingsfund.org.uk/f/256914/x/1404655eb2/integrating_additional_roles_general_practice_2022.pdf  

DPC staff contribute fewer appointments than other staff, though as has previously been discussed, this appointment data may not capture the full extent of the activity that DPC staff carry out in general practice. 

The rolling out of DPC staff has clearly had both positive and negative impacts on general practice. But regardless, the government should be concerned that the staff group into which the most additional resources have been poured over the last five years is not associated with higher patient satisfaction.

The government should improve the quality of DPC data 

It is very difficult to evaluate the efficacy of all new DPC staff with publicly available data. The government records PCN staff separately to GP practice staff while including PCN staff’s activity in general practice appointment data. The NHS’s guidance for how PCNs should record their staff’s activity is unhelpfully vague. It says that “PCNs can record their appointments in any way they feel allows them to provide the highest level of care for their patients”, which means they can record data in existing GP practice appointment systems, PCN hub appointment systems, in sub-contractor or GP Federation appointment systems or “a combination of the above”. ²² NHS Digital, ‘Appointments in general practice: supporting information’, (no date), retrieved 28 March 2025, https://digital.nhs.uk/data-and-information/publications/statistical/appointments-in-general-practice/appointments-in-general-practice-supporting-inf…  

That makes it close to impossible to unpick the effect of additional staff hired in PCNs on general practice activity, satisfaction, QOF outcomes or any other performance metric. 

The government also reports activity and staffing data at inconsistent levels of geography. PCN staff are – logically – reported at a PCN level. However, PCN activity is only reported at a sub-ICB level, limiting the ability to carry out analysis at a lower geography. 

This sounds like an academic complaint but it has real implications for evaluating policy. The expansion of the PCN DPC workforce since 2019 is one of the largest changes to primary care in recent history. Despite that, the government does not publish data that allows for effective external evaluation of the impact of those staff members. The government should take steps to improve the quality of publicly available data for primary care staffing and activity. The ultimate objective should be to allocate PCN DPC staff to a practice level depending on the amount of time that they work in each practice. The government should also require all PCNs to record activity data consistently so that it is possible to see exactly how many appointments are carried out in general practice and how many elsewhere.

There should also ideally be greater granularity of appointment data at a staff level. Currently, it is possible to see whether an appointment was carried out by a ‘GP’ or an ‘other practice staff’ member. That makes it hard to know how much an additional GP trainee or an additional pharmacist (for example) contributes to practice activity.

The government needs to understand the trade-offs in pushing for GPs to deliver more appointments 

The substantial variation in the number of appointments that individual GPs carry out might encourage the government to push more GPs at the lower end of the distribution to deliver similar amounts of appointments as their colleagues who deliver more. And there will probably be useful learnings to take from the way that some GPs organise their practices and their time. 

However, pushing GPs to deliver more appointments above all else may not be a wise policy. Patients care about more than just access to general practice; they also care about how those appointments are delivered. It is difficult to assess the quality of appointments, but this report has shown that practices that provide more appointments remotely are associated with lower patient satisfaction. There are other metrics of quality that are difficult to capture and might be inversely related to the number of appointments that a GP carries out – for example, the British Medical Association (BMA) has recommended that GPs move to an average appointment length of 15 minutes (where many practices currently average 10). ²⁸ British Medical Association, ‘Safe working in general practice in England guidance’, 3 December 2024, retrieved 28 March 2025, www.bma.org.uk/advice-and-support/gp-practices/managing-workload/safe-working-in-general-practice/appointments  

It will also increase GP workloads, raising the risk of burn-out. The most recent GP voice survey, commissioned by the Royal College of General Practitioners (RCGP), found that the two most cited reasons for why a GP might leave the service in the coming years was “to find a better work/life balance” and because they “find general practice too stressful”. ²⁹ Royal College of General Practitioners, ‘GP voice survey: chartbook of all questions’, October 2024, retrieved 28 March 2025, p. 20, www.rcgp.org.uk/getmedia/0d28acfe-532a-427d-a6b7-097ad5c53fbf/RCGP-GP-Voice-Survey-Chartbook-2024.pdf  Given the difficulties in retaining GPs (particularly younger GPs), and the impact that high workloads have on the attractiveness of partnership, the government should be wary of increasing access to general practice by requiring GPs to deliver ever more appointments. 

Finally, productivity is a slippery concept in general practice. We have used the number of appointments per GP as a proxy for productivity in this report because appointments are the only type of GP activity for which we have meaningful data. But appointments are only one part of a GP’s workload and raw numbers of appointments tell us little about the quality of those consultations. They also spend time carrying out activity such as indirect patient care – like writing referral letters – and managing other employees in the practice, among many other things. None of this activity is captured by the appointments data. One interviewee suggested to us that those GPs that deliver fewer appointments in a year are just substituting appointments with other work. ³⁰ Institute for Government interview.  Just because this work is not measured by NHS England does not mean that it is not valuable or that GPs are not being productive when they focus on it.

One of Wes Streeting’s three shifts is towards a more preventative health system. GPs are a crucial part of that ambition. But if GP care became more preventative, it might require longer appointments ³¹ Clet E, Leblanc P, Alla F, and Cohidon C, ‘Factors for the integration of prevention in primary care: an overview of reviews’, BJGP Open, 2023, vol. 8, no. 3, https://bjgpopen.org/content/8/3/BJGPO.2023.0141  and improved continuity of care. ³² Hill A and Freeman G, Promoting Continuity of Care in General Practice, Royal College of General Practitioners, March 2011, pp.12−13, www.rcgp.org.uk/getmedia/aeff056b-da14-40ad-8e6e-66ac9388a461/RCGPContinuity-of-Care.pdf  Delivering a more preventative model of general practice might therefore have a detrimental impact on the number of appointments that GPs can deliver per year.

The government must effectively communicate its reform plans 

General practice has changed substantially over recent years, moving away from the more ‘traditional’ model of delivery. If the government is successful with its three shifts (Streeting wants to shift care from hospitals to the community, from sickness to prevention, and from analogue to digital), it is likely that general practice will continue to change. 

But the analysis in this paper shows that patients tend to report higher satisfaction with factors that are usually associated with that more ‘traditional’ model of general practice. They like to see a GP, in a face-to-face appointment, in a practice with a smaller number of registered patients. This is despite other characteristics of general practice – for example, more appointments delivered by non-GP staff are associated with improved QOF achievement – being associated with better care outcomes. 

It seems very unlikely that the government will return to that ‘traditional’ model of general practice despite its stated ambition to return to the “family doctor”: it is unlikely that the NHS is going to reverse the influx of DPC staff; it is unlikely that there will be a rapid turnaround in the GP partner workforce; and it is unlikely that there will be a return to smaller patient lists. 

However, patient satisfaction is at least in part a function of patients’ expectations. When they fill in the GP patient survey, they are implicitly comparing their actual experience with their expected experience. Part of the decline in patient satisfaction between 2020 and 2023 could therefore be because general practice has changed substantially while patients’ expectations have remained relatively static. Some of that is likely because the government has not effectively communicated its reform plans or why they would benefit patients. 

This interpretation places an onus on the government to improve the way it communicates its policies: where the government believes that reforms to the ‘traditional’ model of general practice will generate better outcomes, it should ensure that the changes and their intended benefits are communicated more effectively to the public. 

As an example, the increase in DPC staff since 2019 has happened largely unnoticed. Aside from a narrow debate that focuses on the use of physician associates (PAs, which account for only 4.3% of the expanded DPC workforce) in general practice, there has been little attention given to the more radical shift in how patients access pharmacy, physiotherapy, dietary or other services. It may therefore be unsurprising then that a patient reports lower satisfaction following an appointment conducted by these staff members rather than the GP they were expecting.

It is not enough, however, to just communicate the change that is happening. The government should also ensure that it explains what it is hoping to achieve with the change. For example, the government might think that more DPC staff means that patients should be able to access an appointment more quickly. Or that larger patient list sizes represent better value for money for patients. Or that conducting more appointments remotely means that GPs can deliver more appointments overall and saves patients the difficulties of coming into the surgery. But patients, understandably, may not be aware of or agree with those rationales. It is up to the government to make the case for its reforms. 

If the government can communicate these changes more effectively, and then follow through by providing the promised benefits, it could be that patient satisfaction recovers without a return to a more ‘traditional’ model of general practice.