News - Construction News

NAO analysis of new Hospital programme

The government’s New Hospital Programme (NHP) has experienced delays and is expected to deliver 32 of the intended target of 40 new hospitals by 2030, according to a new National Audit Office report (NAO).

In 2021-22, 43% of the NHS estate was built before 1985, and the maintenance backlog was £10.2Bn. Seven entire hospitals are structurally unsound and in need of urgent replacement owing to the use of reinforced autoclaved aerated concrete (RAAC). Surveys have found 41 buildings at 23 hospital trusts containing the material.

The Department of Health and Social Care (DHSC) created a Health Infrastructure Plan (HIP) in 2019 to modernise the NHS estate. Under HIP, DHSC planned 27 new hospital schemes by 2030. In October 2020, the Government announced an expansion of DHSC’s plans stating that 40 new hospitals would be built by 2030. It identified and provided information on the types of improvements at 32 of these new hospitals, while another eight would be selected later. Alongside eight other hospitals that DHSC had already approved for construction and which it was not counting towards the 40 new hospitals commitment, this meant a total portfolio of 48 hospital schemes by 2030.

DHSC set up the New Hospital Programme (NHP) to deliver this commitment. The programme was also tasked with improving efficiency, quality and standardisation in hospital construction – using modern methods of construction (MMC) – and a centralised approach to contracting.

31 of the 32 new hospital schemes detailed in the October 2020 announcement had previously been part of a HIP scheme. Responding to NAO requests, DHSC was unable to document fully the process originally used to select HIP schemes. DHSC, supported by NHS England, used clear, evidence-based criteria to create a shortlist but then adjusted this shortlist substantially, a part of the process for which no further documentation is available. The failure to document this part of the process is an omission which means there is no basis for the NAO to determine why DHSC selected these schemes. For large capital programmes, the NAO expects government to use clear, defensible criteria to select schemes and to maintain records of its decisions.

Overall, DHSC has adopted a broad definition of a ‘new hospital’, including: completely new hospitals; complete rebuilds of existing hospitals; major new buildings at existing sites; and major refurbishments of existing buildings. Excluding the eight pre-existing schemes, the NAO’s analysis of the 32 new hospitals detailed in October 2020 suggests 11 (34%) represented whole new hospitals, with another 20 meeting other elements of DHSC’s definition, while one scheme does not meet the definition. To address this, NHP now intends to split another scheme, which was for two hospitals, into two separate schemes.

In 2020, DHSC estimated it needed between £19.8Bn and £29.7Bn of capital funding to build 48 hospitals by 2030. This included between £3.7Bn and £16Bn for the programme’s first four years up to 2024-25. HM Treasury’s decision in the 2020 Spending Review to provide £3.7Bn up to 2024-25 meant most of the larger hospital schemes would need to be delivered towards the end of the decade. In its options appraisal, DHSC labelled this “maximum risk and policy compromises”. This approach would likely result in many schemes being simultaneously under construction, making it harder to find builders and potentially increasing costs.

The October 2020 announcement stated that all 40 named schemes (cohorts 1 to 4) would be ‘fully funded’. Yet the £3.7Bn allocated in the 2020 Spending Review was intended mostly for cohorts 1 and 2. Government had still not made funding and scoping decisions about later cohorts as DHSC had not yet developed its new centralised, standardised approach to hospital construction. This means there was inherent uncertainty about whether specific schemes announced for later cohorts and the additional schemes not yet announced were affordable and achievable.

In its first three years, NHP made slow progress. By June 2023, three of the eight schemes in cohort 1 had opened or part opened against an expectation of five. Forecast costs for schemes in cohorts 1 and 2 increased by 41% between 2020 and 2023.

NHP has been planning for schemes in cohort 3 and later to use the innovative standardised hospital design – known as Hospital 2.0 – which incorporates MMC. It estimates that by cohort 4 hospital construction will be 25% cheaper and 20% quicker compared with traditional approaches. The approach can bring efficiencies and other advantages but remains unproven.

The design for a standardised hospital is delayed. There have been shortages of technical staff and it will not now be completed until May 2024. The delays have constrained NHP’s ability to engage with industry. Furthermore, the NAO finds that NHP’s “minimum viable product” version of Hospital 2.0, which is intended to achieve key objectives at the lowest possible cost, may result in hospitals that are too small. This is because NHP is currently modelling hospital sizes using assumptions that may be unrealistic about the extent to which care in future will be provided outside of hospitals and about the transformational benefits of switching to wards with single rooms only.

NHP has also faced difficulties in hiring skilled staff. By February 2023, the team had only filled 109 posts with full time staff, compared with 223 posts filled through consultancy services (see Notes to Editors). The use of consultancy services on large construction programmes is normal. However, relying on consultants, particularly in a long-term programme, brings risks of a lack of continuity and loss of knowledge.

DHSC reset the programme in May 2023, after HM Treasury agreed an indicative funding envelope and scope. The reset fundamentally changed the content and timing of NHP’s schemes. Five hospitals made of RAAC that were not already in the programme will now be rebuilt by 2030 (in effect becoming NHP’s cohort 5), and DHSC will also count three mental health hospital schemes approved outside of NHP towards the commitment. However, completion of eight of the schemes announced in 2020 will now be delayed until the 2030s. By our analysis, this means that DHSC’s plans will now lead to 32 new hospitals by 2030, according to the definition it used in 2020.

Government has not achieved good value for money with NHP so far. NHP had spent £1.1Bn by March 2023. Progress has been slower than expected on individual schemes and central activities, including developing Hospital 2.0. It can improve value for money through to 2030 but needs to manage substantial risks, including the risk of building hospitals that are too small and rising costs resulting from hospitals being built simultaneously.

NAO recommendations to DHSC, NHS England and government more widely include:

  • Announcements about major capital programmes extending over more than one spending review period should fully reflect known uncertainties.
  • DHSC should choose future hospital construction projects transparently and keep full records.
  • Further bunching of construction schemes should be avoided.
  • NHP should reconsider assumptions that may result in ‘minimum viable product’ future hospitals being too small.

Gareth Davies, the head of the NAO said: “The programme has innovative plans to standardise hospital construction, delivering efficiencies and quality improvements. However, by the definition the government used in 2020 it will now deliver 32 rather than 40 new hospitals by 2030.

“Delivery so far has been slower than expected, both on individual schemes and in developing the Hospital 2.0 template, which has delayed programme funding decisions.

“There are some important lessons to be drawn for major programmes from the experience of the New Hospital Programme so far. These include strengthening the business case process to improve confidence on affordability and delivery dates, and improving transparency for key decisions.”

If you would like to read more stories like this, then please click here