General practice workforce: quality report

Data description 

The general practice workforce statistics release provides information on the number of staff working in general practices across Wales and their characteristics.

Statistics are presented for different staff roles by headcount and full-time equivalent (FTE). Headcounts are unique counts of people and FTEs count the number of hours staff are contracted to work, with one FTE typically equal to 37.5 hours per week.

The release also includes analysis of GPs and wider practice staff by age, gender, ethnicity and Welsh language skills.

Classification system

Staff groupings: GPs

The following explains the individual GP roles included in each GP group.

Fully qualified

• Partner/provider
• Senior partner
• Salaried by practice
• Salaried by other
• Locum – other
• Retainer

Fully qualified permanent

• Partner/provider
• Senior partner
• Salaried by practice
• Salaried by other
• Retainer

GP practitioner

• Partner/provider
• Senior partner
• Salaried by practice
• Salaried by other

GP registrar

• Registrar ST1/4

F2

• Registrar F1/2

Staff groupings: wider practice staff

The following explains the individual staff roles included in each wider practice staff group.

Nurses

• Advanced Nurse Practitioner
• Extended Role Practice Nurse
• Nurse Dispenser
• Nurse Specialist
• Nursing Partner
• Practice Nurse
• Trainee Nurse

Direct Patient Care

• Apprentice - Health Care Assistant
• Apprentice – Other
• Apprentice – Pharmacist
• Apprentice – Phlebotomist
• Dietician
• Dispenser
• Health Care Assistant
• Health Support Worker
• Nursing Associate
• Other
• Paramedic
• Pharmacist
• Pharmacy Technician
• Phlebotomist
• Physician Associate
• Physiotherapist
• Psychological Wellbeing Practitioner
• Social Prescribing Link Worker
• Therapist- Counsellor
• Therapist- Occupational Therapist
• Therapist - Other

Admin/non-clinical

• Administrator
• Apprentice
• Estates and Ancillary
• Finance Director
• Management Partner
• Manager
• Medical Secretary
• Other
• Receptionist
• Senior Manager
• Telephonist

Ethnicity groupings

The following shows the individual ethnicities that are included in the ethnic groups.

White

• White British
• White Irish
• White - Any other White background

Asian or Asian British

• Asian or Asian British – Indian
• Asian or Asian British – Pakistani
• Asian or Asian British – Bangladeshi
• Asian or Asian British - Any other Asian background
• Malaysian
• Chinese

Black or Black British

• Black or Black British – African
• Black or Black British – Caribbean
• Other Black background
• Black Nigerian

Mixed / Multiple Ethnic Group

• Mixed - White and Black African
• Mixed White / Asian
• Mixed White / Black Caribbean
• Other mixed background

Other Ethnic Groups

• Other ethnic group

Not stated/Unknown

• Unknown
• Not Stated
• Declined

Statistical concepts and definitions

Headcount

Headcounts are unique counts of people. This means that if a person works across multiple practices within a health board, they are only counted once in that health board. Similarly, if a person works across multiple staff roles, they are only counted once in the overarching staff group (for example, a person who works as a locum GP and salaried GP is only counted once in the ‘fully qualified GP’ group). 

Full-time equivalent (FTE)

The full-time equivalent (FTE) measure is a measure of the contracted hours of each staff member which is expressed as a full-time equivalent number and is aggregated for the total workforce. 

For all staff other than GP registrars and F2 doctors, 1.0 FTE is equal to full-time work of 37.5 hours per week, 0.5 FTE is equal to 18.75 hours per week. Registrars and F2 doctors have 40 hour per week contracts as standard, so 1.0 FTE is equal to 40 hours for these GP types. 

FTEs for each staff role count all the hours each person is contracted to work, regardless of the number of contracts they have. For example, if the same person works as a salaried GP for 30 hours a week (0.8 FTE) in one practice and for 7.5 hours a week (0.2 FTE) in a different practice, they count as 1.0 FTE in the salaried GP type at the Wales level. 

Contract

A single staff member may have more than one contract type, therefore the number of contracts will differ from the total headcount. This could happen, for example when a person has contracts as a salaried GP and as a GP locum; or if the same person had contracts as a practice nurse and a phlebotomist. 

General practice

General practices provide primary care medical services on behalf of the local health board, with at least one qualified general medical practitioner who is able to prescribe medicine and where patients can be registered and held on a list. For the purpose of this release, the term general practice refers to main surgeries only and excludes prisons, army bases, education establishments, specialist care centres and walk-in centres.

Fully qualified GPs

Includes partners, providers, salaried, retainers and locum GPs only. These are fully qualified GPs registered on the Wales Medical Performers List and licensed by the General Medical Council. It does not include GP registrars. 

Data for headcounts of this staff group are available from 31 June 2021 only and for FTE from 31 December 2021 only. Prior to this data for all the GP types included in the group was not consistently collected and is therefore not comparable. 

Fully qualified GPs can have full-time or part-time contracts. Apart from GP locums, all GPs are permanently contracted staff. GP locums are only counted where they had sessions recorded through Locum Hub Wales in the reference period.

Fully qualified permanent GPs

Includes partners, providers, salaried and retainer GPs only. These are the same as the ‘fully qualified GPs’ GP types except with GP locums removed. 

Data for headcounts of this staff group have been consistently collected and published from 31 March 2020. Headcount data prior to this was collected via different methods (payment systems). When the Wales National Workforce Reporting System was implemented, analysis of comparability between the new and old systems showed that headcounts were not directly comparable and any comparisons of data between the two data collection methods should only be done to show an indication of the scale of the workforce at different points in time. 

FTE data for this staff group is available from 31 December 2021 only. No FTE data collected prior to this date is comparable. 

The comparability guidance above applies to all the GP types included in the ‘fully qualified permanent’ measure (partner, provider, salaried, retainer) when they are analysed as individual GP types.

GP practitioner

Includes partners, providers and salaried GPs only. It does not include GP retainers, GP locums, or GP registrars. Retainers are excluded because they are a small number of GPs who work on specific contracts with reduced hours.

GP practitioners make up the large majority of the workforce and the headcount of this staff group has historically been the most stable measure of the GP workforce. 

GP partner or provider

General practitioners who have entered into a contract with a local health board to provide services to patients. These are typically the most senior GPs at a practice. 

Specifically, a GP partner is a partner in the business partnership that provides GP services. A GP provider is either a sole practitioner; a partner in a partnership and that partnership is the contractor; or a shareholder in a company limited by shares and that company is the contractor.

Salaried GPs

General practitioners who are generally employed and remunerated by salary through the general practice. Salaried GPs can also be employed by health boards directly to work in health board managed practices; in these instances the salaried GP information will be present on the WNWRS. 

GP retainer

A registered GP who has entered the GP Retainer Scheme. This is designed to help keep doctors working in general practice, and usually applies to GPs who are either approaching retirement or who require greater flexibility in their work in general practice. A retainer may only work up to 4 clinical sessions per week in practice.

Locum GP

GP locums are fully qualified GPs who provide services at practices on a temporary basis. In these statistics, only GP locums who actively provided services in the reference period are counted. In practice this is done by only counting GP locums who had sessions recorded through Locum Hub Wales (LHW). 

Nearly all practices will record their locum GP sessions through LHW as this is a requirement for that work to be covered by The Scheme for General Medical Practice Indemnity (GMPI). The only practices which may not do this are local health board managed practices, as locum work would be covered by the health board’s indemnity. GP locums may also use LHW as a scheduling tool, so they may record sessions through LHW, even when working in a local health board managed practice. 

On 30 September 2023 there were 26 local health board managed practices, and 23 of these had locum activity recorded through LHW. At present, there is no formal data collection for locum GP sessions not recorded through LHW; however, it is expected that the amount of locum work not currently captured is likely to be a small percentage of the overall locum GP FTE in each quarter.

Furthermore, any locum GP who is registered on the All-Wales Locum Register, but did not have any sessions recorded on LHW in the quarter, will not be counted in the headcounts or FTE.

GP registrar

Sometimes called a GP trainee, registrars are qualified doctors who are in the process of becoming a GP and undertaking a period of training in general practice and hospitals. Historically, in Wales the training programme consisted of 18 months in approved hospital posts and 18 months in general practice; however, from 2019 the training programme changed to 12 months in hospital posts and 24 months in general practice. This explains some of the increase in registrar numbers since 2019. 

Registrars counted in these statistics are those who are on GP practice placements on the reference date of data extraction, and it does not include those who are on the programme but are on hospital-based placements on the reference date.

A standard contract for a registrar is for 40 hours per week (rather than 37.5 hours), therefore 1.0 FTE for a registrar is equivalent to a 40-hour week contract.

F2 doctors

Doctors with a full General Medical Council (GMC) registration in their second year of postgraduate medical training. They have completed their first foundation year and will be rotating through three specialties. They would typically undertake clinical work under supervision but they are not training specifically to be a GP. A standard contract for an F2 doctor is for 40 hours per week (rather than 37.5 hours), therefore 1.0 FTE is equivalent to a 40-hour week contract.

Nurses 

Qualified nurses who are registered with the Nursing and Midwifery Council (NMC) and provide nursing services. The majority of staff in this group are practice nurses but this group also includes nurse partners and advanced nurse practitioners who perform senior roles in general practices. 

Direct patient care staff

Typically includes anyone who is directly involved in delivering patient care but who is not a nurse or GP. The majority of staff in this group are health care assistants but it also includes staff providing specialist qualified services such as pharmacists and physiotherapists. 

Admin/non-clinical staff

Anyone who is involved in the administration or organisation of the practice. The majority of staff in this group are receptionists but it also includes senior roles such as practice manager and management partner. 

A partnership

A financial arrangement between two or more bodies of which one must be a general practitioner, which provides general practice services to patients.

General Medical Services (GMS)

The contract by which all general practices in Wales provide services on behalf of the local health board.

Statistical unit

Statistics are published for both headcounts and FTE of staff recorded as being employed at general practices through the Wales National Workforce Reporting System on the reference date. 

The number of all active general practices on the reference date is also published.

Statistical population

The statistics refer to all staff employed in general practices in Wales on the reference date. 

Reference area

Most statistics are published at Wales and local health board level, with some high level statistics published at cluster level. 

Time coverage

Data are extracted on 31 March, 30 June, 30 September and 31 December each year by NHS Wales Shared Services Partnership (NWSSP). 

Data for all staff, aside from locums, provides a ‘snapshot’ of the general practice workforce on that specific date. As locum work is temporary, locum data reflects the contracts that were recorded on Locum Hub Wales during the quarter that ended on the reference date. 

Practices are expected to confirm their WNWRS data as being correct during the quarter before the extract is taken. This should provide an accurate reflection of the workforce on the snapshot date.

Source data

Various data sources have been used to produce these statistics. 

Wales National Workforce Reporting System (WNWRS) 

This is the primary source for data on fully qualified permanent GPs and wider practice staff. 

The WNWRS is managed by the NHS Wales Shared Services Partnership (NWSSP). It is a secure web-based system developed to capture practice staff information for all general practices in Wales. In broad terms, practice managers are asked to update the system when staff changes occur and to confirm all records are correct at least once during every quarter. 

Locum Hub Wales (LHW)

As of the 30 June 2021 reference period, data for GP locums have been collected through Locum Hub Wales.

Locum Hub Wales is a service that enables GP practices across Wales to advertise their short-term vacancies, and book a GP locum whose preferences match those of the practice, quickly and efficiently. Locums are not required to book shifts through Locum Hub Wales if they do not wish to do so. However, GP locums must join the All Wales Locum Register (hosted on Locum Hub Wales) and must record details of their shifts worked on in order to be captured by The Scheme for General Medical Practice Indemnity (GMPI) from 1 February 2021.

The Scheme for General Medical Practice Indemnity (GMPI) (NWSSP), is managed by NWSSP’s Legal and Risk Services as part of the Clinical Negligence Scheme for NHS Trusts and Local Health Boards (Administration) (Wales) Directions 2019 and The National Health Service (Clinical Negligence Scheme) (Wales) Regulations 2019.

Electronic Staff Record (ESR) 

ESR is a human resources tool used by the NHS to support the delivery of national workforce policy and strategy. It provides NHS organisations with a range of tools that facilitate effective workforce management and planning. All staff employed by the NHS should have an ESR record as it is used to pay staff. 

All members of the GP speciality training programme (including registrars) are employed by Velindre University NHS Trust and are therefore present on ESR. Registrar data has been sourced from ESR from March 2020 onwards.

As of the 30 September 2021 reference period, data for Foundation Year 2 doctors (F2) on placement in general practices have also been sourced from the ESR. Previously, data for F2 doctors were sourced separately from the Intrepid system. 

Digital Health and Care Wales (DHCW) 

All practices which open, close or merge in Wales need to be recorded by the Welsh Reference Data and Terminology Service (WRTS) in DHCW. This service provides the definitive source of active practices in Wales. 

The statistics in this release refer to the practices which were active during the reference period based on the list of practices provided by the WRTS.

Frequency of data collection

Data is extracted by NWSSP and supplied to Welsh Government every quarter, on these dates each year: 31 March, 30 June, 30 September, and 31 December. 

Data collection

Wales National Workforce Reporting System (WNWRS)

For each staff member, the system includes information on names, staff role, person identifiers (national insurance numbers, GMC number for GPs, NMC number for nurses), contracted and worked hours, start date, termination date, date of birth, gender, ethnicity and Welsh language skills. Once information has been entered, it remains in the system; data only requires amending if staff details change. 

A validated individual level extract is then sent to Welsh Government. All personal information such as names, date of birth and national insurance number are excluded and each individual is allocated a pseudonymised I.D. number which should stay with each individual throughout all data extracts. 

Locum Hub Wales (LHW)

Data for a GP locum is collected once locum worked is arranged through LHW. Unlike the WNWRS, certain data items are not mandatory for GP locums to enter, including age, ethnicity and Welsh language skills, therefore data completeness is usually lower for GP locums than other GP types. 

NWSSP extract data from LHW every month and upload it to their system. Data from LHW is appended to the WNWRS extract when supplied to WG.

Electronic Staff Record (ESR)

NWSSP extract the data for registrars and F2s from the ESR every quarter. ESR contains similar data fields as those on WNWRS and the ESR extract is appended to the WNWRS extract when supplied to WG. 

Data validation

After the data is extracted from WNWRS, validation steps are performed by NWSSP and any quality issues are flagged to the practice for resolution. 

After the initial validation, a data extract is set to Welsh Government where further validation is performed to improve data quality. 

Using automated queries built into the data processing system (using R software), the validation exercise flags anomalous records. The records in question are returned to NWSSP for investigation. NWSSP may then liaise with practices directly and subsequently advise whether to edit or delete the anomalous records. After this process, a final ‘clean’ dataset is created and used for producing official statistics. This final dataset is also loaded back into WNWRS on the same day as the official statistics are published, so that practices and health boards can see their own data and certain aggregated data which matches the published statistics. 

Any records which have any of the following issues are flagged as part of the validation exercise: 

• active practices not present in the dataset
• inactive practice present in the dataset
• GP records with zero contracted hours and/or zero worked hours
• inconsistency between the GMC number and unique pseudonymised ID for GPs with contracts in multiple practices or roles (for example, same two records with the same GMC number but different unique pseudonymised ID)
• staff who joined after reference date
• staff who left before reference date
• GPs with missing or invalid GMC numbers
• nurses missing NMC numbers
• new staff roles which have not provided in previous extracts
• duplicated GMC numbers between WNWRS and ESR sources, to ensure trainee doctors (registrars and F2 doctors) are not double counted
• duplicate staff records within a practice
• practices without GP practitioners or locums
• staff with FTE greater than 1.5
• fully qualified GPs younger than 27 years of age
• headcounts and FTEs that are not broadly in-line with previous quarters
• multiple records with the same pseudonymised ID but different values for age, gender, ethnicity, or Welsh language skills
• records with the same pseudonymised ID across subsequent extracts where age has increased by more than one, or where ethnicity has changed
• missing data for age, gender, ethnicity, Welsh Language, or date joined

Data compilation

NWSSP compiles the data from the WNWRS, LHW and ESR together as one individual level extract for Welsh Government based on the data items in WNWRS. Each record in the extract represents a contract, so for example, there would be two records attributed to a single GP who has two contracts. 

Any staff on maternity or paternity leave are retained in the data.

Adjustment

During the validation exercise, where an error is confirmed and the proposed correction is provided by NWSSP, adjustments are made to the original dataset supplied by either amending a value or removing an entire record. For example, if the incorrect GMC number was recorded for a GP, this is overwritten with the correct GMC number as advised by NWSSP; if there are duplicate records for the same person within a single practice, then a record is removed following advice from NWSSP. 

Any records where a GP locum is recorded through WNWRS are removed. This is because locum work is not recorded consistently through WNWRS and all locum data used in these statistics are collected through LHW. 

The completion of these adjustments results in a final ‘clean’ workforce dataset used for producing statistics for publication. 

The FTE is calculated using the ‘contracted hours’ data item. If this data item is missing or zero, the ‘hours worked’ data item is used instead. If the ‘hours worked’ data item is also missing or zero, the individual is counted as 1 in the headcount but 0 in the FTE measure.

Frequency of data collection

Data is extracted by NWSSP and supplied to Welsh Government every quarter, on these dates each year: 31 March, 30 June, 30 September, and 31 December.

Quality assurance

Quality assurance is undertaken in line with the quality strategy and the Quality pillar of the Code of Practice for statistics.

The GP workforce data is individual-level and allows for detailed validation checks on every record for different data items, which is not possible with aggregated datasets. An initial range of validation checks were performed on the first data extract in 2020 and further validations have been added to the process over time. 

Quality assessment

The first eight cycles of quarterly data collections produced GP workforce statistics labelled as ‘experimental statistics’. This status was necessary as the data items and processes were being developed and understood by all organizations involved. 

As of the 31 December 2021 reference period the experimental label was removed as the statistics were deemed to be robust and reliable, following quality assessments. This reference date coincided with the first publication of FTE statistics.

These statistics are used in a variety of ways. Some examples of these are: 

• advice to Ministers
• to inform workforce planning decisions for practices, health boards and government
• to inform debate in the Welsh Parliament and beyond
• to provide publicly available data on general practice workforce in Wales

These statistics are useful both within and outside the Welsh Government. Some of the key users are:

• Ministers and the Members Research Service in the Welsh Parliament
• local health boards
• local authorities
• GP collaboratives and primary care clusters
• general practices
• the department for Health and Social Services in the Welsh Government
• other areas of the Welsh Government
• National Health Service
• Health Education and Improvement Wales
• Public Health Wales
• the research community including students, academics and universities
• GPC Wales and other representative bodies
• individual citizens, third sector organizations and private companies
• other UK government departments

Completeness

The incompleteness of each data item used to produce the statistical release is presented in Table 1.

[Note 1] Missing data includes records where no data was provided (null entry).

[Note 2] Missing gender data includes records where no data was provided (null entry) and records where the gender is entered as ‘unknown’, ‘not provided’ or ‘other/unknown’.

[Note 3] Missing ethnicity data includes records where no data was provided (null entry) and records where the ethnicity is entered as ‘declined’, ‘unknown’, ‘not provided’, ‘not specified’, ‘unspecified’ or ‘not stated’.

[Note 4] Missing Welsh language skills data includes records where no data was provided (null entry) and records where the Welsh language ability is entered as ‘unknown’, ‘not stated’ or ‘not provided’.

The rate of completion for staff characteristics (age/gender/ethnicity/Welsh language) are noted for each staff type in the statistical release.

Information about missing contract and working hours data is shown in Table 3.

Overall accuracy

For every extraction date, practices are expected to confirm their WNWRS data as correct during the quarter that ends in the reference date of extraction. This should provide an accurate reflection of the workforce on the snapshot date. If a practice neither logs in to the WNWRS to view their data nor modifies data during the latest quarter, then they are counted as not confirming their data as being up to date.

Table 2 shows that since 31 March 2020 the percentage of active practices which did not confirm their data prior to the reference date of extraction ranged between 0 and 11%.

While NWSSP and WG perform a range of data validation checks, data quality is affected by the accuracy of the data inputted by general practices. There may be occasions where not all staff are recorded or incorrect information is inputted, but these are likely to be rare given the extensive validation process and engagement work both NWSSP and WG do with general practices and local health boards. 

There are some known potential quality issues which are being worked on. For example, there is a high degree of confidence in the accuracy of GP records as the GMC number provided is validated against GMC data to confirm the registration. The same check is not yet applied on nurse registration data with the Nursing and Midwifery Council (NMC); however, work is ongoing to try and implement this check in future periods. 

FTE is calculated using the ‘contracted hours’ data item first, and ‘hours worked’ if ‘contracted hours’ is missing or zero. Completeness of these two data items is very high across all staff roles; however, the validation checks and engagement work are not always successful in obtaining corrected data. If we are unable to obtain data for those fields, there is a possibility that the FTE calculated may slightly underestimate the true FTE at a point in time. 

There is also uncertainty on how agency and nurse bank staff are recorded on WNWRS and these issues will be investigated in the future.

[Note 1] no hours include 0 hours or a null entry.

Table 3 shows the percentage of GP records with no hours recorded was high during the early periods of the data collection, ranging between 4% and 20% from 31 March 2020 to 30 September 2021, but decreased in subsequent quarters to less than 1% following investigations by NWSSP. 

The percentage of wider practice staff records with no hours recorded has been very low since the data collection started but has increased marginally in more recent quarters. Some of these staff may genuinely have zero contracted hours, but some are likely to be incomplete records.

Age, gender, ethnicity and Welsh language skills information is not recorded for every person working in general practice as the data items are not mandatory in the reporting system. Staff characteristic statistics are therefore based on partial coverage of staff whose information is available.

GP locums are also not required to enter this information when recording work through Locum Hub Wales, and therefore account for a large proportion of fully qualified GPs with unknown information.

There are cases where characteristics are not consistent from quarter-to-quarter. Following NWSSP investigation, this may result in the correction of a value for the latest reference date, but no adjustments are made for the previous period. This includes cases where a person’s age has changed by more than 1 from one quarter to the next or a person’s recorded ethnicity changes. This affects both GPs and wider practice staff and should be borne in mind when using staff characteristics statistics.

As the extract is compiled in a spreadsheet by NWSSP, this may be subject to transpose errors.

‘Nursing associate’ is an option for the staff role data item on the system; however, this is not an available role for nursing staff in Wales. Every quarter a small number of staff are recorded as this staff role which is likely to be an error and is currently being investigated.

Data revision

The published statistics are not routinely revised, and any marginal quarter-to-quarter changes may be the result of improvements in data quality.

If an error is discovered after publication, statistics will be revised and clearly noted.

Timeliness

Statistics are planned for publication 4 to 5 months after the data reference date.

Punctuality

The majority of publications have been published as planned on the expected publication date. Since 2020, two publications were released later than initially planned due to resource constraints. In these instances, statistics for two quarters of data were published at the same time to immediately resume the usual publication schedule.

Geographical comparability

Central guidance provided by NWSSP for entering information on the WNWRS provides assurance that data is inputted consistently across practices and health boards. There are no known reasons which limit comparisons within any geographic breakdown in Wales.

All four UK nations publish GP workforce data but there may be differences in definitions, methodologies and data collection processes. 

In England, NHS England publish statistics on a monthly basis. The Wales National Workforce Reporting System was originally based on the same tool (National Workforce Reporting Service) (NWRS)) originally used by NHS England (formally NHS Digital), so there are some close similarities in data collection processes. These similarities include: 

• data is extracted on the same reference dates
• definitions for staff roles are consistent
• FTE definition is consistent for all fully qualified GP staff (but not trainees) and all wider practice staff
• both countries have nearly total coverage of all general practices in the respective countries

However there are known differences which include:

• how locums are counted in both countries is not consistent and therefore the measure ‘fully qualified GPs’ is not comparable
• Welsh GP practices are asked to log into the WNWRS and confirm their data in the quarter before data is extracted, in England there is no formal process for this
• NHS England use an estimation model for missing data but Wales does not
• FTE for registrars/trainee doctors is based on 37.5 hours a week in England but 40 hours a week in Wales

Further work is planned to understand the differences between GP workforce data for Wales and England. 

In Scotland, GP workforce data is published annually by Public Health Scotland. 

In Northern Ireland, statistics are published by the Family Practitioner Services (FPS) Information Unit within the Business Services Organisation.

Comparability over time

Workforce data can vary from quarter-to-quarter because of natural workforce flows. This is particularly affected by the availability of new staff being linked to the end dates of training programs. Therefore, any comparisons made between subsequent quarters may only show seasonal variations and not underlying changes in workforce levels. For these reasons, any time series analysis should focus on the same reference dates in each year, for example comparing 30 September 2021 with 30 September 2022. 

Prior to the WNWRS being implemented, data for general practice workforce was collected through the Exeter payment system. This was a fundamentally different method for collecting data. The last publication from this data source was for reference date 30 September 2018. 

As part of the quality assurance process for the newly introduced WNWRS, a final extract was taken from the Exeter payment system in March 2020 and compared with WNWRS data for the same reference date. The headcount of GP practitioners (partners, providers and salaried only) was 5% higher in the Exeter data than in the WNWRS data and only 42% of practices had the same headcount of GP practitioners recorded in both sources. As a result, data from the two sources is not directly comparable, and analyzing headcount measures between the two sources should only provide an indication of the scale of the workforce at different points in time. 

This applies to GP practitioner, fully qualified permanent GPs, partners, salaried and retainer headcounts only. No comparisons of headcount can be made for other GP types between the two data sources. 

No comparisons of FTE between the two sources can be made. 

One of the main reasons why the WNWRS was developed was to remedy a number of data quality concerns with the previous data collection method. A non-exhaustive list of quality issues included: large volumes of missing data for hours worked; missing values for GMC numbers; incomplete list of registration statuses; invalid practice codes; incorrectly coded trainee staff; unrealistic values for age; no data on locums; and limited characteristic information.

The validation processes implemented for WNWRS data provide far greater certainty over the accuracy and reliability of workforce data than the previous data collection method. These validation processes are under constant review and new processes will be added when quality issues are identified. 

Locum data was first collected on a consistent basis from 30 June 2021, so time series analysis for this GP type and any groups of GP types which include GP locums starts from this date. This includes the headline measure for ‘fully qualified GPs’.

Publication

The statistics are pre-announced and then published at 9:30am on the date of publication on the Welsh Government website. 

For publications relating to the 31 March, 30 June and 31 December reference periods, a headline covering the main points is published on the website. Accompanying data tables are also published on the StatsWales website. 

For publications relating to the 30 September reference period, an in-depth HTML report is also published, providing additional analysis and insights using charts and commentary. Data tables for all analysis covered in the HTML report are published on StatsWales. 

The data available on StatsWales can be viewed and manipulated online or downloaded into spreadsheets. Additionally, a connection can be made to the datasets from a software package (for example, Microsoft Excel) using the OData feed of the StatsWales cube.

All statistical releases are published in English and Welsh.

Online databases

Each StatsWales data table includes the full, consistent time series and are published on the ‘general practice workforce’ section of StatsWales.

Documentation on methodology

• Data collection guidance for the WNWRS is available on the NWSSP website.
• Information about LHW 

The WNWRS is supplied by a third-party organisation who are contracted by NWSSP to provide the system. 

Practices are expected edit their data as staff changes arise and to log in to the system every quarter to confirm their data as being up to date. They are also expected to liaise with NWSSP to resolve any data quality issues. 

The WNWRS has a reporting function for practices and health boards, so in addition to providing data for official statistics, it is used as a helpful workforce planning tool by practices.

Confidentiality: policy

The Welsh Government’s statistics and research statement on confidentiality and data access describes our approach to data confidentiality and conforming with the data governance principle of the Trustworthiness pillar in the Code of Practice for Statistics.

Confidentiality: data treatment

The published statistics on StatsWales include small numbers but are likely to be very low risk following a disclosure control assessment. Therefore, we do not apply suppression to small values. 

Currently data are not published for small areas such as individual practices, but this is under review.

Statistician: Sabir Ahmed
Email: stats.healthinfo@gov.wales

Media: 0300 025 8099