National Audit Office Press Releases
Printable version

Improvements needed in NHS England’s modelling for the Long Term Workforce Plan

Weaknesses in the modelling underpinning NHS England’s (NHSE’s) first Long Term Workforce Plan (LTWP) need addressing if the LTWP modelling is to be a reasonable basis for regular strategic workforce planning, according to a new National Audit Office (NAO) report

  • The NAO has assessed NHS England’s modelling for the Long Term Workforce Plan, which sets out projected staffing needs over the next 15 years
  • Creating the modelling is a significant achievement and provides a foundation on which to build
  • Weaknesses include a complex design; manual adjustments; optimistic future assumptions and limited public communication of their uncertainty; and modelling outputs that could not be fully replicated
  • NHS England has committed to improve the modelling as part of regular planned updates

NHSE has for the first time produced modelling that brings together its planning of future NHS health services with its longer-term assessment of the workforce it thinks it will need to deliver them. This is a significant achievement. The LTWP, published in June 2023, estimates that the NHS’s 1.4m full-time equivalent (FTE) staff in 2021-22 will need to grow to between 2.3m and 2.4m FTE workers in 2036-37. The NAO’s report focusses on the modelling behind the LTWP.

From a technical perspective, the NAO found significant weaknesses with the modelling.  The use of manual adjustments and limitations in how modellers documented their work increased the risk of errors. The LTWP modelling took the form of a pipeline: a structured sequence of steps involving a series of distinct models. The NAO was able to replicate the outputs from one part of the modelling, conducted in the Python code, which represented a reasonable technical approach to health workforce modelling. However, it could not replicate the outputs of another part of the modelling, conducted in Microsoft Excel. Overall, this meant the NAO could not replicate the numbers that feature in the published LTWP.

The NAO found that some of the modelling assumptions may be optimistic, including assuming that a planned doubling of the number of undergraduate places in medical school can be achieved by 2031-32. Such a rapid expansion in capacity presents challenges, which the modellers did not factor in. These include the impact of rapid expansion of student numbers on the quality of training, and whether existing staff have capacity to provide on-the-job training to that number of students. The full increase in student numbers is some years away, allowing some time for NHSE to adjust its plans.

NHSE intends that increased domestic education will reverse an historical and growing reliance on the recruitment of professionals trained overseas. In broad terms, NHSE expects the number of international recruits to fall as domestic training grows. NHSE’s modelling projects that there will be no international recruitment of doctors at all from the mid-2030s. In our view, this is not a reasonable modelling assumption and, if the rest of the plan is implemented in full, risks too many medical students being trained from the early 2030s onwards.

The modelling also assumes increases in NHS productivity above the long-term historical average and a significant change in how general practice operates, with much more work done by trainee GPs. The latter means that, while the total number of doctors in primary care will increase substantially, the LTWP foresees there being only 4% more fully-qualified GPs in 2036 than there were in 2021. The number of consultants in the NHS would grow by 49%. The NAO is recommending that future versions of the modelling should do more to explore the uncertainty of these assumptions and what might happen if they do not fully come to pass.

The NAO makes numerous other recommendations aimed at improving NHSE’s modelling, including full integration of the different parts of the modelling pipeline so that manual adjustments can be minimised. While the decisions taken as part of the LTWP are out of scope of our review, we note that government has only committed funding up to 2028-29 and that NHSE plans to make changes in stages. This gives NHSE a built-in opportunity to make adjustments to its workforce plans after it has revisited the modelling.

NHSE has committed to regularly refreshing the modelling and publishing the results, and has recognised the need to improve the modelling. Implementing the NAO’s recommendations can lead to more reliable and transparent modelling next time.

“The creation of the modelling that underpins the Long Term Workforce plan has been a significant undertaking. However, NHS England must strengthen its workforce modelling next time around to make better-informed decisions about the NHS’s future workforce.”
Gareth Davies, head of the NAO

Read the full report : NHS England’s modelling for the Long Term Workforce Plan

Notes for editors

After a request from HM Treasury, the Department of Health & Social Care and NHSE, the Comptroller & Auditor General (C&AG) agreed that the NAO would carry out an independent assessment of the modelling underpinning the LTWP. The scope was to consider whether NHSE constructed its models effectively and whether they operated correctly in a technical sense to generate the projections and other outputs required of them; and to consider whether NHSE’s approach to workforce modelling and the models themselves are a reasonable basis for regular strategic workforce planning.

In 2022, the NAO published Framework to review models, good practice guidance aimed at people who commission analysis, provide analytical assurance and deliver the analysis itself.  Our approach to this assessment was based on this framework.

Channel website: https://www.nao.org.uk/

Original article link: https://www.nao.org.uk/press-releases/improvements-needed-in-nhs-englands-modelling-for-the-long-term-workforce-plan/

Share this article

Latest News from
National Audit Office Press Releases

Webinar: Telford and Wrekin Council’s Award-Winning AI Customer Service Journey