Are NHS Buildings Falling Apart?

The NHS estate in England comprises over 1,300 properties with a total floor area of 26 million square metres — making it one of the largest property portfolios in Europe. Much of it is old: over 40% of the hospital estate predates 1978, and some buildings still in clinical use date from the Victorian era. The condition of these buildings has been deteriorating for years, and the data now shows a system in which the physical infrastructure is failing faster than it is being repaired. The maintenance backlog — the estimated cost of returning NHS buildings to an acceptable standard — reached £13.8 billion in 2024, more than three times the £4.3 billion recorded in 2014.

Capital spending on the NHS has been systematically deprioritised in favour of day-to-day revenue spending. In real terms, annual NHS capital expenditure fell from £5.8 billion in 2014 to a low of £5.0 billion in 2017 before recovering during the COVID emergency (Nightingale hospitals, ventilator procurement). By 2024 it had settled back to £5.9 billion — roughly flat in real terms over a decade, while the backlog tripled. Trusts have repeatedly been required to surrender capital budgets mid-year to cover revenue deficits, a practice known as capital-to-revenue transfers that plugs immediate operational gaps at the cost of long-term infrastructure decay.

The RAAC concrete crisis brought the issue into sharp public focus. Reinforced autoclaved aerated concrete, a lightweight building material widely used in the 1960s to 1980s, has been found to be structurally unsound in buildings across the public sector. In the NHS, 42 trusts have confirmed RAAC in their buildings as of 2024 — though the survey is acknowledged to be incomplete, with some trusts having not yet fully assessed all structures. Remediation ranges from propping and monitoring to full rebuilding. Several hospital wards and departments have been closed or relocated at short notice after structural assessments identified imminent collapse risk.

The consequences of the capital deficit extend beyond buildings. NHS IT infrastructure is equally antiquated: many trusts still operate legacy clinical systems, and interoperability between trusts is limited. Diagnostic equipment — MRI scanners, CT scanners, linear accelerators for radiotherapy — ages beyond optimal replacement cycles, increasing downtime and reducing scan quality. The Lord Darzi review estimated that the combined capital deficit across buildings, equipment, and digital infrastructure is significantly higher than the headline ERIC maintenance backlog figure, because ERIC captures only the building fabric element.

The ERIC (Estates Returns Information Collection) data used here is itself imperfect. Trusts self-report their maintenance backlog through a six-facet survey, and there are known incentives to underreport: acknowledging a large backlog implies a financial liability that affects trust accounts. The NAO has flagged concerns about data quality repeatedly. The true scale of infrastructure decay is very likely larger than the £13.8 billion headline. International comparisons show that the UK spends less per capita on health capital than the OECD average — a gap that has widened, not narrowed, over the past decade. Building new hospitals is essential, but maintaining existing ones is the more immediate and less politically attractive challenge.