Primary care workforce quarterly update statistics - methodology statement

In May 2022 we intend to begin releasing quarterly updates to the existing primary care workforce publications. These updates will bring together staff working in general practice with those in primary care networks (PCNs) using data collected by the National Workforce Reporting Service (NWRS). Data will be updated for different staff groups at different points during the quarter as it becomes available.

Where appropriate, additional data sources will also be incorporated for some staff groups. This paper describes the approach to combining the NWRS sourced data and its subsequent release and the rationale and methodology for incorporating these additional data sources into the quarterly updates.

NWRS data provided by general practices and PCNs will be combined to give a complete as possible picture of the workforce in primary care for all staff groups. For GPs, this will allow data on infrequent locums to mature (due to the known time lag between activity and reporting) to be included in the GP figures each quarter. As data on GP locums takes two months to mature, the release of GPs will follow the release for other staff groups.

There are known gaps in the completeness of NWRS data. For example, 84.5% of PCNs are providing valid data using the NWRS tool, which makes triangulating with additional data sources one way of providing a full picture as possible of the primary care workforce. The data we will use to provide this enhanced view for staff in the Direct Patient Care (DPC) group which may help bring reported FTE closer to actual figures is from the Additional Roles Reimbursement Scheme (ARRS).

Data entered into the NWRS by general practices and PCNs are combined and then compared to reported FTE in the ARRS claims data. Through a series of rules, a collated figure will be created for each DPC job role and reported on a quarterly basis under the ‘experimental statistics’ badge.

The data sources listed below provide an overview of the different data sources used in the production of the quarterly publications.

• PCN workforce taken from NWRS quarterly
• General practice workforce taken from NWRS (month to align with final month of PCN NWRS quarter)
• General practice workforce taken from NWRS (Q4 2018/19) to calculate baseline
• Approved ARRS claims data (most recent month data to align with PCN NWRS quarter)
• Latest ePCN ODS reference data
• Role reference table
• Geographic lookup (PCN>CCG>ICS>region)

We plan to expand and include further data sources which would add to our understanding of GPs providing primary care services in other settings as sources become available.

NWRS is used to produce figures for:

• GPs
• nurses
• administrative staff

For the Direct Patient Care (DPC) group, data from the PCN and general practice NWRS extracts are used to create a combined NWRS figure (same approach as used for the staff groups listed above). In addition, for the DPC staff group, ARRS claims data are used to produce a collated figure. However, NWRS contains data on a wider range of job roles than the ARRS scheme data making comparisons only available for roles with the scheme.

Background

PCNs submit claims for ARRS staff using an online portal, which clinical commissioning groups (CCGs) are responsible for approving based on guidance set out in the Network Contract Directed Enhanced Service (PCN DES).

A PCN is entitled to funding as part of the Network Contract DES to support the recruitment of new additional staff to deliver health services. The new additional staff recruited by a PCN or provided under contract as a service from a third-party organisation are referred to in this Network Contract DES specification as “Additional Roles” and this element of the Network Contract DES is referred to as the “Additional Roles Reimbursement Scheme”. To receive the associated funding, a PCN must show that the staff delivering health services for whom funding is requested, comply with the principle of “additionality”.

Additionality is measured on a baseline of staff supporting a general practice as taken at 31 March 2019 against six of the reimbursable staff roles – clinical pharmacists, social prescribing link workers, first contact physiotherapists, physician associates, pharmacy technicians and paramedics. A PCN baseline will not be established for other ARRS roles however, the additionality principles still apply.

There are some limitations on the numbers of staff in particular roles which PCNs can claim for – see the DES for further detail. CCGs take this into account when approving claims.

ARRS roles

The following job roles are covered under the ARRS scheme. Note there are other job roles within PCNs which are not covered by this scheme for example healthcare assistants.

• Clinical pharmacist
• Pharmacy technician
• Social prescribing link worker
• Health and wellbeing coach
• Care co-ordinator
• Physician associate
• First contact physiotherapist
• Dietician
• Podiatrist
• Occupational therapist
• Nursing associate
• Trainee nursing associate
• Paramedic
• Mental health practitioner (bands 5-8a)
• Advanced practitioners (clinical pharmacists, first contact physiotherapists, dietitians, podiatrists, occupational therapists and paramedics)

Further detail on the ARRS claims data

PCNs submit claims via a portal administered by the North of England Commissioning Support Unit (NECS). The claims are then reviewed for approval by CCGs, and PCNs are then reimbursed for approved claims. The Network DES sets out expectations of the responsibilities of PCNs and CCGs in this process (see sections 7.2.5 and 7.2.7 for requirements on PCNs and 10.2 for the role of CCGs). NECS send monthly extracts of the claims data to NHS England and NHS Improvement Primary Care Workforce team. The data is currently sent in Excel format, but work is underway to set up an API to replace this.

There is currently no defined national schedule for the submission and approval of ARRS claims and it is acknowledged that some claims are made and approved retrospectively. For example, ARRS data for March may not be submitted until later months. As such, the figures presented cannot be considered to represent the entire picture of staff in post and should be considered an under-count. Early stage work, led by NHS England and NHS Improvement, is underway to support and encourage timely submission and approval of ARRS claims.

ODS codes are incorrect or missing for approximately 7% of claims and so are not included in the analysis, therefore potentially leading to further undercount. There are also cases where approved claims are attributed to closed PCNs, which may artificially increase the combined figure, but would most likely be substantially outweighed by the claims with no valid code at all.

Baseline

ARRS claims can only be made for additional staff recruited since March 19. Therefore, monitoring tends to consider changes net of the baseline.

General practice workforce data reported through the NWRS are used to establish that baseline.

In summary, NWRS figures are calculated at PCN level (combining practice and PCN data, net of baseline), and compared to the FTE data in the claims submitted by that PCN, taking the higher of the two values.

Unaligned practices

To cover DPC staff at practices in that CCG which are not aligned to a PCN, published data will include one Unaligned category per Job Role and CCG. This will present in the data as an extra PCN which only has practice staff.

NWRS data recorded for practices which were previously part of PCNs which have since closed will also be reported under Unaligned. Where there is Claims FTE data against a PCN which was closed prior to the claim month, that FTE data will be excluded from the analysis.

Role comparisons

NWRS and the claims portal do not use the exact same terminology for all roles. Where this is the case some roles will be mapped, as described in the table below. Not all ARRS roles are currently available in the NWRS, primarily mental health practitioners. The differences are as follows:

Where advanced practitioners (AP) do not appear in NWRS they will be excluded from the collated figure to avoid the risk of double counting (where PCNs may have been recording them in the NWRS under their specific non-AP role). Where they do appear in NWRS they will be combined into a single AP category to align with how they appear in the ARRS data.

Until mental health practitioners appear in NWRS they will be excluded from the collated figure to avoid the risk of double counting (where PCNs may have been recording them in the NWRS under Other DPC).

Release frequency

It will only be possible to publish quarterly to align with the NWRS PCN collection frequency. Data will be made available for September 2021 onwards to coincide with the first data being available from when the NWRS tool was brought in-house at NHS Digital. This is also the same period that maturity of the NWRS PCN was such to allow for published reporting.

Use of estimates

The primary care network FTE figures do not include estimates for staff records with no contracted or worked hours recorded, or where PCNs have not provided usable data for an entire staff group. General practice figures do include estimates where practices have not provided valid data for a staff group, or where hours information is missing. These estimates have not been used in the collated figure.

The use of ARRS data to supplement NWRS data for the purposes of generating quarterly figures is not seen as a long-term solution. The strategic aim is to be able to derive FTE data for the quarterly update publication primarily from the NWRS collection tool. As such we intend to keep the use of ARRS data under close review and will consider ceasing the publication of the collated figure if any one of the three criteria below are met.

1. There is an improvement in the completeness of data submitted to NWRS.
2. The difference in reported FTE between the collated figure and NWRS alone narrows sufficiently that the use of ARRS claims data is no longer warranted.
3. The ARRS scheme comes to an end, meaning that claims data are no longer available.