You are here

  • Home
  • Archive
  • Volume 32, Issue 7
  • Predictors and population health outcomes of persistent high GP turnover in English general practices: a retrospective observational study

Article Text

Article menu
Original research
Predictors and population health outcomes of persistent high GP turnover in English general practices: a retrospective observational study
  1. Rosa Parisi1,
  2. Yiu-Shing Lau2,
  3. Peter Bower3,
  4. Katherine Checkland2,
  5. Jill Rubery4,
  6. Matt Sutton2,
  7. Sally J Giles5,
  8. Aneez Esmail3,
  9. Sharon Spooner2,
  10. Evangelos Kontopantelis1,3
  1. 1 Division of Informatics, Imaging & Data Sciences, School of Health Sciences, The University of Manchester, Manchester, UK
  2. 2 Health Organisation, Policy and Economics (HOPE) Group, Centre for Primary Care & Health Services Research, School of Health Sciences, University of Manchester, Manchester, UK
  3. 3 NIHR School for Primary Care Research, Division of Population Health, Health Services Research and Primary Care, School of Health Sciences, University of Manchester, Manchester, UK
  4. 4 Alliance Manchester Business School, University of Manchester, Manchester, UK
  5. 5 NIHR Greater Manchester Primary Care Patient Safety Translational Research Centre, University of Manchester, Manchester, UK
  1. Correspondence to Dr Rosa Parisi, Division of Informatics, Imaging & Data Sciences, School of Healh Sciences, University of Manchester, Manchester, M13 9PL, UK; rosa.parisi{at}manchester.ac.uk

Abstract

Objective English primary care faces significant challenges, including ‘persistent high turnover’ of general practitioners (GPs) in some partnerships. It is unknown whether there are specific predictors of persistent high turnover and whether it is associated with poorer population health outcomes.

Design A retrospective observational study.

Methods We linked workforce data on individual GPs to practice-level data from Hospital Episode Statistics and the GP Patient Survey (2007–2019). We classified practices as experiencing persistent high turnover if more than 10% of GPs changed in at least 3 consecutive years. We used multivariable logistic or linear regression models for panel data with random effects to identify practice characteristics that predicted persistent high turnover and associations of practice outcomes (higher emergency hospital use and patient experience of continuity of care, access to care and overall patient satisfaction) with persistent high turnover.

Results Each year, 6% of English practices experienced persistent high turnover, with a maximum of 9% (688/7619) in 2014. Larger practices, in more deprived areas and with a higher morbidity burden were more likely to experience persistent high turnover. Persistent high turnover was associated with 1.8 (95% CI 1.5 to 2.1) more emergency hospital attendances per 100 patients, 0.1 (95% CI 0.1 to 0.2) more admissions per 100 patients, 5.2% (95% CI −5.6% to −4.9%) fewer people seeing their preferred doctor, 10.6% (95% CI−11.4% to −9.8%) fewer people reporting obtaining an appointment on the same day and 1.3% (95% CI −1.6% to −1.1%) lower overall satisfaction with the practice.

Conclusions Persistent high turnover is independently linked to indicators of poorer service and health outcomes. Although causality needs to be further investigated, strategies and policies may be needed to both reduce high turnover and support practices facing challenges with high GP turnover when it occurs.

  • General practice
  • Health policy
  • Health services research

Data availability statement

Data may be obtained from a third party and are not publicly available. Data are either are available from NHSDigital upon request (enquiries@ nhsdigital. nhs. uk) or publicly available online from NHS Digital and NHS. Organisation Data Service (GP workforce, https://digital.nhs.uk/data-and-information/publications/statistical/general-and-personal-medical-services; GPs-by-general practices, https://isd.digital.nhs.uk/trud/users/guest/filters/0/categories/5/items/58/releases; GP patients’ survey, https://www.gp-patient.co.uk/surveysandreports-10-162022).

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/bmjqs-2022-015353

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • Increases in general practitioner (GP) turnover, especially ‘persistent high GP turnover’, are contributing to the crisis in English primary care. It is not known whether there are certain practice characteristics associated with ‘persistent high turnover’ and whether persistent high turnover is in turn associated with worse performance on commonly used indicators of service and health outcomes.

    WHAT THIS STUDY ADDS

    • Practices with persistent high turnover tend to be larger, located in more deprived areas and with a higher morbidity burden across serious chronic conditions. The distribution of these practices also varies across regions, with the highest levels of persistent high turnover being in NHS Cumbria and North East, South Central and West Midlands. Persistent high turnover was associated with higher emergency hospital attendance and admission rates, in addition to a lower proportion of patients obtaining an appointment on the same day or being satisfied overall with their practice.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • Persistent high turnover is an important parameter when considering quality of care and access provided by general practices. Although further research is needed to establish causality, persistent high turnover appears to be contributing directly or indirectly to avoidable health system costs. Strategies and policies are needed to support practices facing challenges with high GP turnover.

    Introduction

    General practitioners (GPs) have a key role in the healthcare system in England, providing comprehensive population healthcare and acting as gatekeepers to specialist care.1 Nevertheless, in the last few years, English primary care has faced major challenges with the number of GPs per patient reducing2 3 and an increasing number of GPs considering early retirement.3 4 Other GPs state intentions to reduce working hours, due to excessive workload, dissatisfaction with physical working conditions,3 low morale, or reduced job satisfaction5–7 and high burnout rate.8 9

    In 2015, the UK government promised 5000 new GPs by 202010 but failed to deliver on that promise.11 The contributing factors to the steady decrease in the number of GPs can be found in the insufficient number of newly trained GPs joining the workforce, lack of overseas recruitment and more GPs retiring early.3 The decreasing number of general practices is also partly due to the inability to recruit staff and lack of resources allocated to the service as the population grows, patient consultations increase and people are living longer with complex health needs.12

    GP turnover, which can be defined as the proportion of GPs leaving a practice,13 is a major issue for the healthcare system. GP turnover has increased in the last decade, and it is unevenly distributed across England,13 with practices in more deprived areas likely to experience higher levels of turnover.13 In these deprived areas, average GP consultation rates are higher, and GPs need to deal with increasingly complex problems in less time, with no additional resources.14–16

    There is limited information available on the reasons GPs leave their practices.17 Therefore, it is unknown whether a GP leaving a practice is retiring, moving to a different practice or leaving the profession. For this reason, it is difficult to discern whether GP turnover may be a temporary situation for the practice or when the practice is facing enduring problems with retention. Previous work has highlighted that some practices experience high GP turnover for long periods of time,13 defined as ‘persistent high turnover’. This phenomenon may indicate recurring problems with recruitment and retention18 and may have an impact on safety of patients, quality, access and continuity of care.

    Very little is known about high GP turnover13 and its association with practice-level outcomes. A 2015 US study found that primary care provider turnover was associated with worse patient experiences of care but did not have a major effect on ambulatory care quality.19 However, high turnover could inevitably lead to poorer continuity of care, when the latter has been associated with higher risks of hospitalisation20 21 and mortality,21 higher costs of care,22 higher emergency care use23 and lower patient satisfaction.24 High turnover may also affect the ability to deliver primary care services,25 so it may affect outcomes directly as well as indirectly (through poorer continuity of care).

    Continuity of care matters to patients, as shown by the fact that more than half of GP patients taking part in a national survey expressed willingness to wait longer to see their preferred GP.26 Continuity of care has also been associated with lower hospital admissions for conditions that are primarily managed in primary care.23

    It is currently unknown whether certain practice characteristics are associated with persistent high turnover and whether practices facing persistent high turnover are associated with worse performance on commonly used indicators of service and health outcomes. This study aims to address that gap by examining if practice list size, deprivation and other characteristics are associated with persistent high turnover in English primary care, and whether practices with persistent turnover are associated with higher hospital activity, lower continuity of care, poorer access and lower satisfaction.

    Methods

    The study follows the Strengthening the Reporting of Observational Studies in Epidemiology guidelines.

    Data sources and study design

    The study used a retrospective observational design (2007–2019 and 2009–2017), combining data from four resources13: GP workforce,27 GPs-by-general practice,28 Hospital Episode Statistics29 30 and the GP Patient Survey (GPPS).31 Details are provided in the online supplemental material text 1.

    Supplemental material

    Variables

    We defined GP turnover as the proportion of GPs who leave a practice during a year. We defined ‘persistent high GP turnover’ as more than 10% of GPs leaving a practice in each of at least 3 consecutive years.13

    For outcomes, we included emergency attendances at all type 1 and 2 emergency departments in England, where there is a consultant-led 24-hour service or a consultant-led single specialty with full resuscitation facilities and designated accommodation for the reception of accident and emergency (A&E) patients.30 Emergency admission included all non-elective admissions to hospitals. Both types of measures are reported as the number of A&E attendances or admissions per 100 registered patients.

    Three outcomes were derived from the GPPS because they were considered the most relevant: the proportion of patients who were able to see their preferred GPs ‘always or almost always’ or ‘a lot of times’ among those patients who answered they had a preferred GP, the proportion of patients who were able to obtain an appointment with a GP or a nurse on the same day, and the proportion of patients who reported being ‘very satisfied or fairly satisfied’ or gave a ‘very good’ or ‘fairly good’ score with the overall experience with the practice.

    Statistical analyses

    Practices with fewer than 750 registered patients, or with no GPs left at the end of the year in question, or active for less than 3 years were excluded from the analyses (1455 in total or 1.5%). Two sets of analyses were conducted. The first aimed to identify risk factors associated with persistent high turnover. The second aimed to investigate the association between persistent high turnover and practice-level outcomes.

    Multivariable logistic regression panel data model with random effects was used to quantify the association between risk factors and persistent high turnover. The outcome persistent high turnover was modelled as a binary outcome, according to whether a practice experienced or not persistent high turnover (classified as 1 and 0, respectively). Multivariable linear regression panel data models with random effects were used to investigate the association between persistent high turnover and the various outcomes. Additionally, the Hausman test and the Breusch-Pagan Lagrange multiplier were used to assess whether random effects were justified for these models.

    All models were adjusted for practice area deprivation captured by the 2015 Index of Multiple Deprivation (IMD) in quintiles32; size of the practice population in quintiles; NHS regions; rurality, defined according to the 2011 rural–urban classification33; and the aggregate prevalence of seven serious chronic conditions extracted from the Quality and Outcomes Framework (QOF).34 35 These conditions were coronary heart disease, stroke/transient ischaemic attack, hypertension, diabetes, heart failure and chronic kidney disease. When examining associations between turnover and outcomes, we included interactions between persistent high turnover and either IMD, practice population, NHS regions or rurality. Postestimation average marginal effects were calculated to obtain the predictive probability for the interactions of interest. Due to the small proportion of missing values in the analyses (which varied between 0.2% and 4.5%), complete case analyses were conducted.

    When exploring the association between persistent high turnover and outcomes, the following sensitivity analyses were performed: analysing turnover, modelled as continuous or binary (GP turnover above 10%); restricting the analyses to 2015–2017 and including full time equivalent(FTE) per patient ratio in the analyses (also when persistent high turnover was the outcome); restricting the analyses of the GPPS to 2011–2017 since the question relative to the ability to get an appointment on the same day or overall experience of the patients with the practice changed slightly after 2010.

    All analyses were performed in Stata V.16.

    Stakeholder involvement

    The main stakeholders in this study were GPs. In view of this, GPs were involved in two ways, as part of the research team and during separate stakeholder involvement meetings. The purpose of the stakeholder involvement meetings with GPs was to gain their feedback on the study findings. This involved two discussion groups with a total of four GPs. They welcomed the findings of the study and recognised the need for research in this area. They highlighted a number of issues they faced, including issues with occupational health, which can be poor for GPs, workload pressure, limited opportunities and contribution to decision making and the management of their practice, particularly salaried GPs, with lack of funding and investment from the government.

    Results

    Descriptive statistics

    The average number of practices included was 7526 per year, and on average, 373.5 (6%) of them had experienced persistent high turnover. In particular, the proportion of practices with persistent high turnover increased during the study window from 2.7% (211/7946) in 2009 to 6.3% (417/6585) in 2019 and with a minimum in 2009 and a peak in 2014 equal to 9% (688/7619). The distribution of the number of practices included and those with persistent high turnover is provided in table 1 and online supplemental figure 1. There was regional variability in the proportion of affected practices (figure 1).

    Figure 1
    Figure 1

    Distribution of number of practices and proportion of practices with ‘persistent high turnover’ by NHS regions.

    View this table:
    Table 1

    Number of practices and number of practices with ‘persistent high turnover’ by year

    Risk factors associated with persistent high turnover

    Characteristics associated with persistent high turnover were: higher practice location deprivation (OR of highest deprivation quintile vs lowest 1.21, 95% CI 1.01 to 1.46), larger list size (OR of highest list size quintile vs third quintile 2.69, 95% CI 2.35 to 3.08), urbanity (rurality OR 0.79, 95% CI 0.66 to 0.93) and larger QOF morbidity burden across serious conditions (OR per one unit increase of 1.03, 95% CI 1.02 to 1.04). A change in QOF morbidity burden from the 25th percentile to the 75th percentile would correspond to increased odds of persistent high turnover of 1.27 or an increase in the probability of persistent high turnover from 3.8% to 6.1%. We also observed regional variation, after controlling for other covariates, with the highest adjusted rates observed in NHS Cumbria and North East (OR 1.73, 95% CI 1.32 to 2.27) and the lowest in NHS Cheshire and Merseyside (OR 0.54, 95% CI 0.39 to 0.75), compared with London (see table 2 and online supplemental table 1).

    View this table:
    Table 2

    Multivariable analysis of risk factors associated with ‘persistent high turnover’

    Persistent high turnover and emergency attendances or admissions

    During the study window, the mean number of hospital emergency attendances or admissions was 24.62 (SD 10.34) and 9.28 (SD 2.58) per 100 registered patients, respectively. Emergency visits increased steadily from 17.44 in 2009 to 26.57 in 2017, as did admissions from 9.02 to 9.85 over the same time period (online supplemental table 2).

    Persistent high turnover was associated with an increase of 1.80 (95% CI 1.55 to 2.05) hospital attendances per 100 registered patients, controlling for other covariates. Variables associated with increased emergency attendances included deprivation and the QOF morbidity burden across serious conditions (table 3). There was regional variation in the impact of persistent high turnover on the outcome, as expected. This implied that in some regions (NHS Cumbria and North East, NHS East Midlands, NHS South East and NHS Wessex), the difference in emergency attendances between practice with and without persistent high turnover was greater than that in other regions (online supplemental figure 2).

    View this table:
    Table 3

    Multivariable analysis of the association between ‘persistent high turnover’ and outcomes (emergency attendances, emergency admissions, proportion of patients seeing their preferred doctor, proportion of patients getting an appointment on the same day or proportion of patients overall satisfied with the practice)

    Persistent high turnover was significantly associated with 0.1 more emergency admissions per 100 patients (0.11, 95% CI 0.06 to 0.16), after controlling for other covariates. Other variables associated with an increase of emergency admissions were higher practice area deprivation, urbanity, larger practice list size and QOF burden of serious conditions (table 3). Again, there was regional variability, with the majority of NHS regions significantly associated with higher emergency admission compared with London, except for NHS Central Midlands and East Midlands, and NHS South Central. Examining the persistent high turnover with practice location deprivation interaction, we found that there were small differences in the outcome for practices located in the most deprived areas, which increased for the more affluent areas (online supplemental figure 3).

    Persistent high turnover and continuity of care, patient satisfaction and access

    The mean proportion of patients who were able to see their preferred doctor was 65.1% (SD 17.2%) during the study window, decreasing from 72.5% in 2009 to 56.1% in 2017; whereas the mean proportion of patients who were able to obtain an appointment on the same day was 45.9% (SD 23%) during the study window, decreasing from 80.3% in 2009 to 32.3% 2017, and the mean proportion of patients who were satisfied with the practice overall was 86.3% (SD 9%) and remained stable during the study window (online supplemental tables 3–5).

    Persistent high turnover was significantly associated with a 5.25 decrease in the proportion of patients seeing their preferred doctor (−5.25, 95% CI −5.64 to −4.86). Higher practice area deprivation, large practice size, urbanity and greater QOF morbidity burden of serious conditions were also associated with a decrease in the proportion of patients seeing their preferred doctor. There was also regional variability in the distribution of the proportion of patients seeing their preferred GPs (table 3), but also in the association between persistent high turnover and the proportion of patients seeing their preferred doctor. The size of the association of interest also varied across practice location deprivation strata (online supplemental figures 4 and 5).

    Persistent high turnover was significantly associated with a 10.61 decrease in the proportion of patients getting an appointment on the same day (−10.61, 95% CI −11.40 to −9.81). Large practices and those with a greater QOF morbidity burden of serious conditions were also associated with a decrease in the proportion patients getting an appointment on the same day (table 3). Interaction terms indicated that the size of the association was smaller for larger practices and those located in deprived areas (online supplemental figures 6 and 7).

    Persistent high turnover was also associated with a small decrease in patient satisfaction (−1.34, 95% CI −1.56 to −1.12). Other variables associated with a decrease in patients’ overall satisfaction were practice area deprivation list size, QOF morbidity burden of conditions, large practice size, and practices in London or in urban areas (table 3). Interaction terms indicated variation in the association of interest, by practice location deprivation and list size (online supplemental figures 8 and 9).

    All the sensitivity analyses were broadly consistent with the main results (see table 4 and online supplemental tables 6–13).

    View this table:
    Table 4

    Sensitivity analysis of the multivariable analyses of the association between ‘persistent high turnover’ or high turnover and different outcomes

    Discussion

    Main findings

    A small but significant number of general practices in England experience high GP turnover for at least 3 consecutive years, a phenomenon here defined as persistent high turnover. Findings revealed that practice area deprivation, larger practices, a greater burden of serious health conditions and urban area are associated with persistent high turnover. Regional variability in persistent high turnover was also observed. Furthermore, our results showed that practices having issues with GP turnover for a longer period of time were associated with higher emergency attendances and admissions. Patients from those practices were less likely to see their preferred doctor, to obtain an appointment with a healthcare professional on the same day of contacting the practice and were less satisfied with the overall service of the practice. Available data provide limited information about the reasons why GPs leave their practice; therefore, it is unknown whether a GP is retiring, moving to a different workplace or leaving the profession. To have a more accurate understanding of the impact of GP turnover, it is important to focus on those practices that have recurrent problems with retention of staff.

    Strengths and limitations

    The study has several strengths. It uses national data from England, which has allowed GPs’ actual turnover to be quantified rather than ‘intention to leave’ the practice. Second, it covers a long time window, which has enabled examination of whether a practice experiences turnover issues during consecutive years. Third, the method used to calculate GP turnover uses the exact dates when a GP joins and leaves a practice, and is more accurate than using aggregate data.13 18

    Limitations of the study need to be acknowledged as well. The datasets used do not include information on the reasons the GPs leave the practice; therefore, we were unable to know whether the GPs moved to another practice, left the profession or retired.28 We were not able to distinguish between salaried and partner GPs; therefore the results are combined for these two categories. It is worth noting that salaried GPs do not have the same legal responsibilities or financial investments as GP partners, giving them more flexibility.12 36 It was not possible to adjust for GP workload in the analyses because NHS Digital revised the methodology to calculate GPs’ FTE per patient ratio after 2015. This variable was included in the sensitivity analysis restricted to 2015–2019 or 2015–2017. We had access to only limited demographic information from the datasets, and therefore, we could not take these into account in the analyses. The time window analysed ends in 2020 and does not take into account all the stress and pressure GPs have faced during the COVID-19 pandemic. It is likely that GP turnover during and following the pandemic response was modified by multiple factors including temporary return to work by recently retired GPs,37 increased opportunities for GPs to work remotely38 and continuing increased need for healthcare from GPs.7 Further analyses will be needed to understand how these and other changes may affect GP turnover. Finally, this was an observational study, and despite the fact that we could test for an association between potential risk factors and persistent high GP turnover or between persistent high GP turnover and poor outcomes, we cannot exclude these associations are due to unknown confounders.

    Comparison with existing literature

    Our study has explored the prevalence of persistent high GP turnover, the practice characteristics associated with it, or its association with patterns of healthcare delivery or patient experience health outcomes. The existing literature has explored changes in healthcare used linked to GP turnover,39 the unequal distribution of the GP workforce according to area deprivation or continuity of care in community setting and its association with hospital activity23 and GP patients’ satisfaction.24 Our findings are in line with Sabety et al, who found that GP turnover was associated with lower use of primary care and increased used of specialty, urgent and emergency care.39 Our findings are also similar to those of Asaria et al, who found that the GP workforce is smaller in more deprived areas40 and similar to those studies which found that continuity of care is associated with higher hospital admissions and lower patients satisfaction.23 24 An analysis by the Health Foundation has found that patients living in more deprived areas report a poorer overall experience with their GP practice and that practices in these areas receive the lowest overall satisfaction scores compared with more affluent areas.26 These findings are broadly confirmed in our analyses where practices with persistent high turnover and located in more deprived areas were associated with lower ability to access to practice and lower overall satisfaction with the practice. Our results, however, are in contrast to Anderson et al, who systematically reviewed GPs’ intentions to quit their patient care and found that neither small or large practices were associated with GPs’ intentions to quit their job.41 However, these studies explored GPs’ intentions to leave the job rather than actual movement of GPs.

    Interpretation of findings

    Our findings have revealed that characteristics associated with practices experiencing high GP turnover for sustained periods of time include high practice area deprivation, large practice size, high morbidity burden of serious conditions and urban area. The association between persistent high turnover and deprivation may be explained by the burden and challenges GPs face in deprived areas in managing patients with more complex health needs with no additional resources, and the unequal distribution of the GP workforce in deprived areas. A recent study exploring how socioeconomic deprivation impacts GP work found that in highly deprived areas, GP work typically extends beyond the management of the illness but that they are not resourced to perform those additional tasks.42 For example, they need to manage the increased burden of multimorbidity at an earlier age and balancing increased medical complexity with social complexity compared with more affluent areas.42 The Health Foundation has shown that, after adjustments to take account of increased workload in more deprived areas, these practices receive around 7% less funding per need-adjusted registered patient than those serving less deprived populations.26

    The association between practices with persistent high GP turnover and higher emergency attendances and admissions might be explained by the impact of GP turnover on continuity of care, as avoidable emergency attendances have been linked to poor quality of primary care.43 Regional variations were found when we explored risk factors for persistent high turnover and for the relationship between these practices and hospital activity or GP patients’ satisfaction. These differences might be due to different levels of social deprivation, unequal distribution of the GP workforce and different pressures on the healthcare system.

    High GP turnover, especially when it persists for consecutive years, is a concern for the healthcare system. While we cannot estimate the financial cost attributable to each GP practice with persistent high turnover, we can estimate the associated cost of persistent high turnover for emergency hospital care (A&E and non-elective admissions). If the associations found in these studies are causal, they would suggest considerable costs to the healthcare system from the effects of turnover. For example, the costs generated by an average practice with problems with persistent high turnover and high emergency admissions/attendances is an additional £73 200 per annum (£2.9 million per annum to the healthcare system).44 We would expect persistent high turnover to be associated with levels of use of other health and social care services as well as impact of unmet health and social care needs which would likely increase future financial costs to the healthcare system.

    Practices with persistent high GP turnover need to be better supported by local and national authorities; policies and strategies to maximise retention of GPs should facilitate sustainable GP workload and contractual requirements, as well as the need for personal and professional support, targeting areas which influence job satisfaction and work–life balance.45 This may require attention to the funding formulae, which determine the distribution of funding for practices, as these currently do not fully take account of the demands associated with practising in a deprived area.46

    Conclusions

    Practices with persistent high turnover of GPs are independently linked with poorer outcomes such as higher numbers of emergency hospital attendances and admissions compared with practices without persistent high turnover. Persistent high turnover has an impact on quality of healthcare and contributes to avoidable health system costs. One of the factors associated with a practice experiencing high turnover over a number of years is deprivation of the area where the practice is located, highlighting the need for more support for these practices. There is a need for more in-depth studies to explore the contexts and reasons of GPs leaving their practices. Strategies and policies are needed to support practices facing challenges with GP turnover.

    Data availability statement

    Data may be obtained from a third party and are not publicly available. Data are either are available from NHSDigital upon request (enquiries@ nhsdigital. nhs. uk) or publicly available online from NHS Digital and NHS. Organisation Data Service (GP workforce, https://digital.nhs.uk/data-and-information/publications/statistical/general-and-personal-medical-services; GPs-by-general practices, https://isd.digital.nhs.uk/trud/users/guest/filters/0/categories/5/items/58/releases; GP patients’ survey, https://www.gp-patient.co.uk/surveysandreports-10-162022).

    Ethics statements

    Patient consent for publication

    Ethics approval

    Not applicable.

    References

    1. ↵
      2. Marshall MN
      . How well do general practitioners and hospital consultants work together? A qualitative study of cooperation and conflict within the medical profession. Br J Gen Pract 1998;48:1379–82.pmid:http://www.ncbi.nlm.nih.gov/pubmed/9800393
      OpenUrlAbstract/FREE Full Text
    2. ↵
      2. Buchan J ,
      3. Charlesworth A ,
      4. Gershlick B , et al
      . A critical moment: NHS staffing trends retention and attrition. The Health Foundation, 2019.
    3. ↵
      2. Palmer W
      . Is the number of GPs falling across the UK? In: Nuffield Trust blog, 2019.
    4. ↵
      2. Gibson J ,
      3. Sutton M ,
      4. Spooner S , et al
      . Ninth national GP Worklife survey: University of Manchester: policy research unit in commissioning and the healthcare system Manchester centre for health economics, 2018.
    5. ↵
      2. Gibson J ,
      3. Checkland K ,
      4. Coleman A , et al
      . Eighth national GP Worklife survey: policy research unit in commissioning and the. Manchester: Healthcare System Manchester Centre for Health Economics, 2015.
    6. ↵
      2. Dale J ,
      3. Potter R ,
      4. Owen K , et al
      . The general practitioner workforce crisis in England: a qualitative study of how appraisal and revalidation are contributing to intentions to leave practice. BMC Fam Pract 2016;17:84.doi:10.1186/s12875-016-0489-9 pmid:http://www.ncbi.nlm.nih.gov/pubmed/27439982
      OpenUrlCrossRefPubMed
    7. ↵
      2. Odebiyi B ,
      3. Walker B ,
      4. Gibson J , et al
      . Eleventh national GP Worklife survey. In: University of Manchester: Policy research unit in commissioning and the healthcare system Manchester centre for health economics, 2021.
    8. ↵
      2. Iacobucci G
      . Burnout is harming GPs' health and patient care, doctors warn. BMJ 2021;374:n1823.doi:10.1136/bmj.n1823 pmid:http://www.ncbi.nlm.nih.gov/pubmed/34281907
      OpenUrlFREE Full Text
    9. ↵
      2. Cheshire A ,
      3. Ridge D ,
      4. Hughes J , et al
      . Influences on GP coping and resilience: a qualitative study in primary care. Br J Gen Pract 2017;67:e428.doi:10.3399/bjgp17X690893 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28483822
      OpenUrlAbstract/FREE Full Text
    10. ↵
      2. Hunt J
      . New deal for general practice. Jeremy Hunt sets out the first steps in a new deal for GPs; 2015.
    11. ↵
      1. The Guardian
      . GP number in England down every year since 2015, 2022. Available: https://www.theguardian.com/society/2022/apr/11/gp-numbers-in-england-down-every-year-since-2015-pledge-to-raise-them
    12. ↵
      1. Centre for Workforce Intelligence
      . In-depth review; 2014.
    13. ↵
      2. Parisi R ,
      3. Lau Y-S ,
      4. Bower P , et al
      . Rates of turnover among general practitioners: a retrospective study of all English general practices between 2007 and 2019. BMJ Open 2021;11:e049827.doi:10.1136/bmjopen-2021-049827 pmid:http://www.ncbi.nlm.nih.gov/pubmed/34420932
      OpenUrlAbstract/FREE Full Text
    14. ↵
      2. McLean G ,
      3. Gunn J ,
      4. Wyke S , et al
      . The influence of socioeconomic deprivation on multimorbidity at different ages: a cross-sectional study. Br J Gen Pract 2014;64:e440.doi:10.3399/bjgp14X680545 pmid:http://www.ncbi.nlm.nih.gov/pubmed/24982497
      OpenUrlAbstract/FREE Full Text
    15. ↵
      2. Mercer SW ,
      3. Jani BD ,
      4. Maxwell M , et al
      . Patient enablement requires physician empathy: a cross-sectional study of general practice consultations in areas of high and low socioeconomic deprivation in Scotland. BMC Fam Pract 2012;13:6.doi:10.1186/1471-2296-13-6 pmid:http://www.ncbi.nlm.nih.gov/pubmed/22316293
      OpenUrlCrossRefPubMed
    16. ↵
      2. Kontopantelis E ,
      3. Mamas MA ,
      4. van Marwijk H , et al
      . Chronic morbidity, deprivation and primary medical care spending in England in 2015-16: a cross-sectional spatial analysis. BMC Med 2018;16:19.doi:10.1186/s12916-017-0996-0 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29439705
      OpenUrlCrossRefPubMed
    17. ↵
      2. Sansom A ,
      3. Terry R ,
      4. Fletcher E , et al
      . Why do GPs leave direct patient care and what might help to retain them? A qualitative study of GPs in South West England. BMJ Open 2018;8:e019849.doi:10.1136/bmjopen-2017-019849 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29326195
      OpenUrlAbstract/FREE Full Text
    18. ↵
      2. Taylor DH ,
      3. Leese B
      . General practitioner turnover and migration in England 1990-94. Br J Gen Pract 1998;48:1070–2.pmid:http://www.ncbi.nlm.nih.gov/pubmed/9624750
      OpenUrlAbstract/FREE Full Text
    19. ↵
      2. Reddy A ,
      3. Pollack CE ,
      4. Asch DA , et al
      . The effect of primary care provider turnover on patient experience of care and ambulatory quality of care. JAMA Intern Med 2015;175:1157–62.doi:10.1001/jamainternmed.2015.1853 pmid:http://www.ncbi.nlm.nih.gov/pubmed/25985320
      OpenUrlPubMed
    20. ↵
      2. Nyweide DJ ,
      3. Anthony DL ,
      4. Bynum JPW , et al
      . Continuity of care and the risk of preventable hospitalization in older adults. JAMA Intern Med 2013;173:1879–85.doi:10.1001/jamainternmed.2013.10059 pmid:http://www.ncbi.nlm.nih.gov/pubmed/24043127
      OpenUrlPubMed
    21. ↵
      2. Sandvik H ,
      3. Hetlevik Øystein ,
      4. Blinkenberg J , et al
      . Continuity in general practice as predictor of mortality, acute hospitalisation, and use of out-of-hours care: a registry-based observational study in Norway. Br J Gen Pract 2022;72:e84–90.doi:10.3399/BJGP.2021.0340 pmid:http://www.ncbi.nlm.nih.gov/pubmed/34607797
      OpenUrlAbstract/FREE Full Text
    22. ↵
      2. Hussey PS ,
      3. Schneider EC ,
      4. Rudin RS , et al
      . Continuity and the costs of care for chronic disease. JAMA Intern Med 2014;174:742–8.doi:10.1001/jamainternmed.2014.245 pmid:http://www.ncbi.nlm.nih.gov/pubmed/24638880
      OpenUrlPubMed
    23. ↵
      2. Barker I ,
      3. Steventon A ,
      4. Deeny SR
      . Association between continuity of care in general practice and hospital admissions for ambulatory care sensitive conditions: cross sectional study of routinely collected, person level data. BMJ 2017;356:j84.doi:10.1136/bmj.j84 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28148478
      OpenUrlAbstract/FREE Full Text
    24. ↵
      2. Fan VS ,
      3. Burman M ,
      4. McDonell MB , et al
      . Continuity of care and other determinants of patient satisfaction with primary care. J Gen Intern Med 2005;20:226–33.doi:10.1111/j.1525-1497.2005.40135.x pmid:http://www.ncbi.nlm.nih.gov/pubmed/15836525
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      2. Shen X ,
      3. Jiang H ,
      4. Xu H , et al
      . The global prevalence of turnover intention among general practitioners: a systematic review and meta-analysis. BMC Fam Pract 2020;21:246.doi:10.1186/s12875-020-01309-4 pmid:http://www.ncbi.nlm.nih.gov/pubmed/33250045
      OpenUrlCrossRefPubMed
    26. ↵
      1. The Health Foundation
      . The Health Foundation’s response to the Health and Social Care Committee’s inquiry – The future of General Practice; 2021.
    27. ↵
    28. ↵
      1. NHS Digital TRUD
      . NHS ODS weekly prescribing-related data.. Available: https://isd.digital.nhs.uk/trud3/user/guest/group/0/pack/5
    29. ↵
    30. ↵
    31. ↵
      1. Ipsos MORI
      . GP patient survey. practice data (weighted), 2009-2017. Available: https://www.gp-patient.co.uk/surveysandreports-10-16
    32. ↵
      1. Department for Communities and Local Government
      . The English index of multiple deprivation (IMD) 2015 – guidance, 2015. Available: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/464430/English_Index_of_Multiple_Deprivation_2015_-_Guidance.pdf
    33. ↵
      1. Office for National Statistics
      . Rural urban classification (2011) of lower layer super output areas in England and Wales; 2011.
    34. ↵
      2. Kontopantelis E ,
      3. Mamas MA ,
      4. van Marwijk H , et al
      . Geographical epidemiology of health and overall deprivation in England, its changes and persistence from 2004 to 2015: a longitudinal spatial population study. J Epidemiol Community Health 2018;72:140.doi:10.1136/jech-2017-209999 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29263178
      OpenUrlAbstract/FREE Full Text
    35. ↵
      2. Doran T ,
      3. Fullwood C ,
      4. Gravelle H , et al
      . Pay-for-performance programs in family practices in the United Kingdom. N Engl J Med 2006;355:375–84.doi:10.1056/NEJMsa055505 pmid:http://www.ncbi.nlm.nih.gov/pubmed/16870916
      OpenUrlCrossRefPubMedWeb of Science
    36. ↵
      1. The King’s Fund
      . Improving the quality of care in general practice: report of an independent enquiry commissioned by the King’s Fund, 2011. Available: www.kingsfund.org.uk/publications/ improving-quality-care-general-practice [Accessed May 2022].
    37. ↵
      2. Haynes L ,
      3. Haynes L
      . How many retired GPs could return to work to fight COVID-19? GP Online; 2020.
    38. ↵
      2. Green MA ,
      3. McKee M ,
      4. Katikireddi SV
      . Remote general practitioner consultations during COVID-19. Lancet Digit Health 2022;4:e7.doi:10.1016/S2589-7500(21)00279-X pmid:http://www.ncbi.nlm.nih.gov/pubmed/34952678
      OpenUrlCrossRefPubMed
    39. ↵
      2. Sabety AH ,
      3. Jena AB ,
      4. Barnett ML
      . Changes in health care use and outcomes after turnover in primary care. JAMA Intern Med 2021;181:186–94.doi:10.1001/jamainternmed.2020.6288 pmid:http://www.ncbi.nlm.nih.gov/pubmed/33196767
      OpenUrlPubMed
    40. ↵
      2. Asaria M ,
      3. Cookson R ,
      4. Fleetcroft R , et al
      . Unequal socioeconomic distribution of the primary care workforce: whole-population small area longitudinal study. BMJ Open 2016;6:e008783.doi:10.1136/bmjopen-2015-008783 pmid:http://www.ncbi.nlm.nih.gov/pubmed/26787245
      OpenUrlAbstract/FREE Full Text
    41. ↵
      2. Anderson R ,
      3. Long L ,
      4. Robinson S , et al
      . Why do primary care doctors quit direct patient care? A systematic review of empirical research. Final report, 2016.
    42. ↵
      2. McCallum M ,
      3. MacDonald S
      . Exploring GP work in areas of high socioeconomic deprivation: a secondary analysis. BJGP Open 2021;5:BJGPO.2021.0117.doi:10.3399/BJGPO.2021.0117 pmid:http://www.ncbi.nlm.nih.gov/pubmed/34465578
      OpenUrlCrossRefPubMed
    43. ↵
      2. Parkinson B ,
      3. Meacock R ,
      4. Checkland K , et al
      . How sensitive are avoidable emergency department attendances to primary care quality? Retrospective observational study. BMJ Qual Saf 2021;30:884–92.doi:10.1136/bmjqs-2020-011651 pmid:http://www.ncbi.nlm.nih.gov/pubmed/33144351
      OpenUrlAbstract/FREE Full Text
    44. ↵
    45. ↵
      2. Chilvers R ,
      3. Richards SH ,
      4. Fletcher E , et al
      . Identifying policies and strategies for general practitioner retention in direct patient care in the United Kingdom: a RAND/UCLA appropriateness method panel study. BMC Fam Pract 2019;20:130.doi:10.1186/s12875-019-1020-x pmid:http://www.ncbi.nlm.nih.gov/pubmed/31514728
      OpenUrlCrossRefPubMed
    46. ↵
      1. The Health Foundation
      . 'Levelling up' general practice in England. What should government prioritise? 2021.

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Twitter @RosaParisi09, @MattXSutton, @SMS_Spooner

    • Contributors RP, EK and YSL led the study design and planned and performed the statistical analyses. RP wrote the first draft of the manuscript. SJG, RP and SS organised the discussion group with general practitioners. RP, YSL, PB, KC, JR, MS, SJG, AE, SS and EK contributed to the interpretation of the findings and the final draft of the paper. RP is the guarantor. The corresponding author attests that all listed authors meet the authorship criteria and no others meeting the criteria have been omitted.

    • Funding This project has been funded by the Health Foundation as part of the Efficiency Research Programme. All decisions concerning analysis, interpretation and publication are made independently from the funder.

    • Competing interests None declared.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

    Linked Articles

    • Editorial
      High cost of broken relationships
      Christine A Sinsky James Morrow
      BMJ Quality & Safety 2023; 32 373-375 Published Online First: 17 Feb 2023. doi: 10.1136/bmjqs-2023-015930