A national audit finds only 16% of people diagnosed with severe frailty in England had the medication review the GP contract requires

This report was published on 5 December 2025 and is not new this week. It is carried because it is the national audit of the community end of the frailty pathway in England, and the baseline against which any whole-system measure of frailty care has to be read Audit/Data (National Audit Office, HC 1518, 5 December 2025).

In 2024-25, 226,000 people were diagnosed with severe frailty, 12% of those assessed for frailty that year. The GP contract requires a clinical review after that diagnosis, including an annual medication review and a discussion of falls. The audit found that only 16% (37,000) of those patients had a medication review, 18% (41,000) had a falls risk assessment and 29% (66,000) had given consent for an enriched summary care record. Thirty-two of 106 NHS local areas assessed fewer than 10% of their registered patients aged 65 or over for frailty. Among care home residents, 44% had a personalised care and support plan agreed or reviewed in 2024-25, down from 76% in 2022-23.

The audit’s own conclusion is that the Department of Health & Social Care and NHS England do not yet have a single clear joined-up strategy for frailty, and that neither knows accurately how many people with frailty may need support. It notes that the national dataset it relies on, the NHS England core contract GP data, gives no information on age, sex or socio-economic factors such as ethnicity or deprivation. The audit also cites earlier research, using data from 2006 to 2017, indicating that people with severe frailty are nearly six times more likely to be admitted to hospital and have average hospital costs nearly nine times greater than people without frailty. That is the audit citing research, not a finding of its own analysis.

The audit measures whether required steps happened, not what happened to patients afterwards, and it covers England only; this source carries no international comparator. Because the dataset has no deprivation breakdown, the figures cannot show whether the shortfall falls more heavily on deprived or rural populations. The finding corroborates the published pattern on community settings being structurally under-equipped, now checked against the report itself. It also sets the practical context for the electronic frailty index meta-analysis: identification is advancing, and the national audit shows the follow-up that should come after it is largely not recorded as happening.