General Practice Workforce 31 March 2021

This publication presents statistics relating to the general practice workforce in England as at 31 March 2021.

Accompanying this publication are the following additional files and tools to enable further analysis:

• September 2015 to March 2021: High-level figures in the Excel Bulletin tables and the Power BI dashboard
• March 2021 only: High-level figures in the Excel Bulletin tables including:
  • Full-time equivalent counts by high-level staff group per 100,000 registered patients
  • Headcounts of general practice staff by high-level staff group and ethnicity  
• Regional tables (March 2021 only) - additional breakdowns with interactive pre-set pivoted tables which provide for region, STP and CCG:
  • Full-time equivalent work commitment by high-level staff group
  • Interactive tables and CCG-level records
• Experimental analysis showing annual counts of joiners to and leavers from the GP workforce (excluding registrars and locums)

• Practice information (contract type and dispensing status) in a Comma Separated Values (CSV) file

• March 2021 only: Individual and Practice-level data in the Comma Separated Values (CSV) files.
  • The Individual-level CSV is comprised of a row for each role in general practice, with a pseudonymised ID number - labelled UNIQUE_IDENTIFIER - assigned to each individual. If an individual is working multiple roles in general practice, the same UNIQUE_IDENTIFIER will appear for each role they are working. This means that, by counting distinct instances of UNIQUE_IDENTIFIER within various groups, accurate headcounts within those groups can be calculated. The UNIQUE_IDENTIFIER numbers are reassigned for each publication and so cannot be used to track individuals over time across different publications. The Individual-level CSV includes information about the CCG each individual works within and includes estimated records, which are calculated at CCG level.
  • The Practice-level CSV provides aggregated headcounts and FTE counts for each job role at each practice. Because we do not publish full estimates at practice level, the figures in this file do not include fully estimated records and so headcount and FTE totals will not equal those that can be calculated from the Individual-level CSV, or the figures in the Bulletin and Regional Excel files and Key Facts section, all of which include estimates. Partial estimates, where FTE has been estimated for a known individual, are included in these figures.

Figures in all regional tables are published against the regional hierarchy as at 1st April 2021. Historical Practice and Individual-level CSVs will be re-mapped to the new structure and republished in due course.

We are keen to ensure that our reports are as useful and relevant as possible for our users. We are therefore continually working to improve the quality of the data and analysis. Changes can include improving the coverage, completeness or accuracy of the data, or revising elements of the methodology and are made in consultation with colleagues, including within the Department of Health and Social Care (DHSC), NHS England and NHS Improvement (NHSEI) and Health Education England (HEE). We welcome feedback from all our users and you can contact us at [email protected], 

These General Practice Workforce statistics are relatively new, following the introduction of a new primary data source in September 2015 and any figures published within this series are not comparable with figures from earlier years.

The majority of the figures are a snapshot of the workforce at the date of data extract  which is the last day of the month. However, figures include an additional measure which reflects the amount of 'ad-hoc' or 'infrequent' locum usage during the period.

We now use two main data sources to produce these statistics:

• The workforce Minimum Data Set (wMDS) collected directly from general practices via the National Workforce Reporting System (NWRS) data entry module. This collects information on all staff except GP Registrars including information about ad-hoc or 'infrequent' locums which we collect in a separate module in the tool.
  Until January 2019, some figures were supplied to us directly by Health Education England regions, but the NWRS has included this HEE data since March 2019 making it the main data source for the general practice workforce.

• Health Education England's (HEE) Trainee Information System (TIS) provides all information about GP registrars and has been our primary source for this data since June 2018. We adopted use of TIS for GP registrar data as it provides more complete and timely data than the Electronic Staff Record (ESR) which we had used previously.
  A comparison of the TIS and ESR data sets enabled us to calculate estimates for GP registrar counts back to September 2015 to produce a more consistent national time series although regional comparisons cannot be made prior to June 2018.
  Details on GP registrar data source and estimations methodology are in the Data Quality Statement accompanying the publication of December 2018 data.

In addition to calculating estimates for GP registrars as already outlined, in November 2019 we also produced estimated figures for GP locums and revised our entire full-time equivalent time series, updating figures from September 2015 onwards. This means that the figures released in November 2019 supersede earlier releases and because the figures are not comparable, we have archived these publications to avoid confusion. More information about these estimates, the methodologies and the Data Quality are included n the General Practice Workforce 30 September 2019 publication, released 28 November 2019. 

More information is also available in What this publication can tell you about.

Not all GP practices provide valid data which means some figures are estimated. For the March 2021 data collection, 98.9% of eligible general practices provided valid GP data for non-registrar GPs, with GP data for the remaining 1.1% of practices estimated.  Full-time equivalent (FTE) figures were also estimated for otherwise valid records with no submitted hours, which affected GP records in 5.5% of practices in March 2021. More information about the estimation methodology can be found in the Background Data Quality Statement; the estimation rates for all four staff groups can be found under Estimation Rates below.

From July 2019 GP practices have been joining with other health and care providers to form Primary Care Networks (PCNs) and will work collaboratively with network members to strengthen primary care. Since that time, Clinical Directors for the PCNs have been recruited and information about these individuals was included in the September and December 2019 publications. However, a PCN specific data collection has been operative since February 2020 and information on Clinical Directors and other PCN staff is now published in the Primary Care Network Workforce series of Official Statistics. 

More information on these changes can be found in the accompanying Data Quality Statement and on the March 2021 Data Quality tab.

Estimates are made for both headcount and full-time equivalent (FTE) for those practices which did not provide complete and/or valid data for one of the four staff groups; this could be due to poor data quality or no submitted data.

Full Estimation - practices who provided no valid data for one or more staff groups. For these practices, CCG-level estimations were made.

Partial Estimation - practices whose data contained one or more otherwise valid records with no submitted hours. These records were retained and the full-time equivalent (FTE) estimated.

Table 1: Percentage of practices with estimated records, March 2020 and 2021

Due to the effect of seasonality on general practice workforce trends, comparisons between different releases should only ever be made on a year-on-year basis. 

It is not always possible or appropriate to compare figures over time, depending on how they were originally calculated. Where this is the case, clear indications have been included in the publication(s) and associated supporting material. Sometimes we may advise that comparisons might be made with caution, knowing that the impact of these changes is small. We will always aim to give guidance on how to interpret any changes in the series, to ensure time series comparability is as clear as possible.

These statistics are currently produced on a quarterly basis, with the time series going back to September 2015 when this data collection was first made. The collection was initially biannual, until December 2016 when GP data began to be collected quarterly. Data for the other staff groups began to be collected quarterly from December 2017.  Although  we collected and published monthly information for October and November 2020, data quality concerns mean that we will not be including figures from those releases in any time series of future analysis although the publications are still available for transparency purposes.  We have reverted temporarily to a quarterly collection and publication cycle although we expect that monthly extracts and publications will resume during 2021.