General Practice Workforce 31 December 2020

This release of the General Practice Workforce Official Statistics has been affected by several significant events which have had – or will have – an effect upon the figures.

We collect information about the general practice workforce directly from practices using the National Workforce Reporting System (NWRS).  We ask practices to ensure that the information about their staff is kept up-to-date and NWRS automatically extracts the data. There is evidence to suggest that the completeness and coverage of data extracted from the NWRS for March and June 2020 were adversely affected by a range of issues, including the COVID-19 pandemic, and that the exceptional pressures on the general practice workforce meant not all practices were able to update their NWRS data in time for the extracts on 31 March and 30 June.

For September 2020, we were confident the practices had returned to normal levels of NWRS data completion in time for the extract on 30 September.

'Infrequent ' or 'ad-hoc' locums and use of combined monthly extracts

This series is primarily a snapshot of the general practice workforce at the time of data extract. The only group for which this is not the case is an element of the GP locum workforce which one might describe as 'infrequent' or 'ad-hoc' locums.

Some practices employ locum GPs on a long-term or regular basis, for example to provide cover for maternity leave or for one or more recurring weekly or monthly sessions. In these cases, the locum’s details are recorded in the main part of the NWRS together with records for permanent practice staff and, like other staff roles, information about them is a snapshot.

However, some locum provision is needed on a more ad-hoc basis, for example to cover one or more sessions at short notice. In such circumstances, practices enter summary details of the locum GP along with the total number of hours worked during the entire reporting period. This means that this 'infrequent' locum element of the workforce is captured and measured differently to all other staff roles. The headcount and FTE of 'infrequent' locums is not a snapshot, but a reflection of their usage - total headcount and average weekly FTE - throughout the reporting period. The 'infrequent' locum FTE is therefore a calculated measure that is indicative of usage and activity of the cohort.

'Long term' and 'infrequent' locum figures are combined into a single GP Locum job role in this publication.

In October 2020, we collected and published information on the general practice on a monthly basis for the first time. At that time, we noticed significantly fewer than expected records for 'infrequent' locum GP records in the October data extract, with the total hours worked also much reduced. Clearly, we would expect the total hours worked by the 'infrequent' locum cohort to be lower when collecting data monthly than when we extracted on a quarterly basis, perhaps by around two thirds. However, the total 'infrequent' locum worked hours in October 2020’s data extract were far less than a third of September’s quarterly total. Although the recorded 'infrequent' locum worked hours increased in the November extract, they were still significantly lower than might be expected when considering usage in previous quarters. 

In addition, we are aware that a far lower percentage of GP practices logged onto the NWRS during October and November than we had expected. This may be because practices were not accustomed to the monthly process and had overlooked the need to ensure that all data additions and changes – particularly for their 'infrequent' locum staff – had been made by 31 October and 30 November. It is also possible that increasing workload pressures due to the COVID-19 pandemic meant that some NWRS users did not have time to enter details of 'infrequent' locums working at their practices during October and November.

We are aware that fewer practices employed locum GPs in the six months from March 2020 because of the effect of the COVID-19 pandemic and a decrease in the number of face-to-face, telephone and video appointments that practices were able to offer during that time. However, we know that the numbers of appointments that took place during October and November were somewhat closer to expected levels for the time of year.

With the December extract, the number of practices submitting information about locums, and consequently the counts of and total hours worked by locums during the month increased to a level which we have no evidence to suggest were underreported. Moreover, when combining the three months of 'infrequent' locum data from October to December, despite the apparent underreporting in October and November, the headcount of 'infrequent' locums is in line with historical quarterly figures.

This is partly because of the greater completeness of the December extract and means that, for headcount, any locums who worked in October/November AND December, but were not reported in the first two monthly extracts, are still being captured in headcount because they are in December’s extract.

Because combining the three months of 'infrequent' locum information provides what we believe to be comparable quarterly headcount figures, and to align with the temporary move back to quarterly extracts and therefore avoid the need for a time series break, the 'infrequent' locum element of this release's figures relate to a quarter, rather than a month.

'Infrequent' locum FTE estimation methodology

However, though we believe the headcount of 'infrequent' locums to be reliable across the three month period, what will not be captured as a result of low data completion rates in the first two months of the quarter are any hours which were not reported. For this reason we have adopted the following estimation methodology for the FTE of 'infrequent' locums, based on the assumption that the 'infrequent' locum usage reported for December was:

a) complete and reliable

b) indicative of the levels of usage that would likely have been occurred in October and November:

• sum hours from the three monthly extracts and calculate a weekly average FTE for each reported 'infrequent' locum. This is the known FTE and is the FTE figure for each record presented in the Individual Level CSV, and aggerated in the Practice Level CSV.
• sum hours for December for each reported 'infrequent' locum, and multiply this by 64/21 (number of working days October to December/number of working days in December alone). This adjustment, rather than simply multiplying the hours by three (number of months), accounts for an extra working day in October in 2020 compared to November and December.
• calculate a weekly FTE from these estimated hours and subtract the known FTE to give the difference between known FTE and estimated FTE
• this resulting difference in FTE - 71 - is included in the England-level FTE figures in the bulletin tables but assigned to the 'unknown' categories for gender or regional breakdowns.
• in the Individual Level CSV this estimated FTE is reported in a separate line.
• the practice level CSV does not include estimates and so the estimated FTE element for 'infrequent' locums is not included.

The releases for October and November 2020 presented FTE figures only. The figures for those months are not included in the timeseries in this publication. We are preserving those statistical publications for transparency and to enable analysts to do specific analysis understanding the caveats, but it is evident that the transition to monthly collection during the pandemic led to a decrease in quality and completeness of the data submitted by GP practices and we will not use figures  for October and November 2020 in any future time series or analysis.

Since July 2019, all GP practices in England have had the opportunity to join one of around 1,250 Primary Care Network and the vast majority have taken the opportunity to do so. PCN member organisations, which include GP practices, and other health, social care, mental health, and voluntary sector providers, will deliver accessible and integrated care to local communities.

These PCNs have their own distinct direct patient care workforces and data relating to these staff was collected for the first time on 31 March 2020. There is a helpful overview of PCNs and a new series of experimental statistics.

It is possible that some individuals previously working in a GP practice may transfer some or all of their working hours to their new PCNs. In particular:

• All PCNs are required to have a suitably qualified accountable person to act as the clinical director. This post is usually filled by a GP, nurse, or another direct care professional and typically requires a commitment of 0.25 FTE. It is therefore possible that FTE counts of staff in these job roles will decrease in the general practice workforce statistics as some of their working hours are transferred to the PCNs.
  PCNs employ GPs or nurses only in the capacity of clinical directors - they do not employ GPs or nurses to deliver patient care.

• Some clinical pharmacists and pharmacy technicians formerly working in general practice will transfer some or all of their working hours to their local PCN which will reduce the FTE and headcount figures in the General Practice Workforce statistics for staff in these roles.
• There may also be decreases in the FTE counts for some other roles such as administrative staff and other direct patient care professionals if they begin to work full or part-time for the PCN.

We are working closely with data providers to ensure that the reported hours of staff working on both general practices and PCNs are recorded correctly to prevent double counting.

We have included information about each GP practice’s PCN in the practice level CSV file. When considering these general practice statistics by PCN, please take note of the following points:

• aggregating from GP practice to PCN to CCG will not replicate the CCG figures
  This is because we estimate for missing records at CCG level but not at a lower level meaning the CCG totals would be higher than the sum of the PCN figures.

• because membership of a PCN is not mandatory, some practices have chosen not to join a PCN and information about their workforces will not be included in any PCN totals

• while PCN boundaries are generally expected to align with CCG boundaries there are some exceptions, notably where effective and successful cross-CCG collaborative working was already in place. In this instance, a responsible CCG is identified for PCN-data reporting purposes even if some GP practice members belong to a different CCG

• the PCN structure is not static and some PCNs have already closed, merged, or recently opened. Similarly, some GP practices have changed their PCN membership – a few have done so several times – since the PCNs were first formed in July 2019, and some GP practices that were originally PCN members have left the local PCN but not joined another

• we have no plans to retrospectively re-map PCN membership in the practice-level CSVs as practice membership changes

Information about PCNs, the GP practice members and the relationships to CCGs are available from the Organisation Data Service (ODS) which receives monthly updates about PCNs and their member organisations.

This publication includes a standalone table of counts of PCN Clinical Directors. We began collecting information on PCN Clinical Directors during 2019 when the PCNs were first set up, and before the PCN module of NWRS was available.

The role of Clinical Director is only available for use in PCNs and we have asked practices to transfer details of all staff holding PCN roles - including Clinical Directors - to the PCN module in the NWRS collection. Therefore, these Clinical Directors do not constitute part of the General Practice workforce which is why they are being reported separately.

We anticipate that all practices will transfer Clinical Directors to the applicable PCNs over the next few months and that this standalone table will be withdrawn.