Staffed, Not Empty: Reading Adult Social Care Data Beyond the Discharge Headline

Every winter, and increasingly every summer, a number emerges from the discharge debate and settles into the news cycle. It usually takes a form like this: thousands of care home beds stand empty while hundreds of patients wait in hospital for somewhere to go. The figure is真实, in the narrow sense that it has been published. What it rarely carries with it is the chain of definitions, collection dates and denominators that produced it.

Adult social care is not a single dataset. It is a patchwork of collections: the monthly Capacity Tracker, the annual Short and Long Term Support return from local authorities, workforce estimates assembled by Skills for Care, and the inspection record maintained by the Care Quality Commission. Each answers a different question. Treating them as interchangeable is the most common error in the discharge conversation, and it is the one that most reliably produces a misleading headline.

Three instruments, one word: capacity

When a report says a care home has a bed available, it usually means a bed is registered and, on the day of the snapshot, not occupied. That is a measure of vacancy, not of capacity. A room may be registered but unusable because it cannot meet the needs of the person being referred — a first-floor room for someone who cannot manage stairs, a shared room for a resident who requires single accommodation, or a unit closed temporarily for infection control. A bed may be physically present and clinically inappropriate.

The Capacity Tracker, which providers complete monthly, captures some of this nuance, but it is self-reported, its response rate varies by region and by provider type, and it is a snapshot rather than a flow. A home that reports one vacancy on the last Thursday of the month may have admitted three residents and discharged two in the preceding fortnight. Occupancy percentages computed from such snapshots are useful for trend-watching and poor for claiming that a specific patient could have been placed.

Staffing hours are not a quality score

Workforce data is often folded into the same sentence as occupancy, as though a home with vacancies must also be short of staff. The two are related but distinct. Skills for Care's annual estimates describe filled posts against establishment — how many roles exist and how many are occupied. They do not describe how those hours are deployed, what proportion is agency cover, or whether the skill mix on a given shift includes a registered nurse.

A care home can be fully staffed on paper and still unable to accept a hospital referral, because the referral requires nursing care and the home is registered for residential care only. It can also have empty beds and no staff to open them. The distinction matters because the policy response differs: one problem is a recruitment problem, the other a commissioning and registration problem. Aggregating both into a single "social care capacity" figure obscures the remedy.

The denominator decides the story

Rates and counts are routinely mixed in public commentary. A count of delayed discharges is a count of days, not of people; a single resident whose discharge is delayed for three months contributes ninety days to a total that may be reported as though it represented ninety individuals. A local authority with a small care home stock can show a dramatic percentage change from a handful of closures. A national rate per 100,000 adults aged 65 and over will look calm beside a regional figure computed on a much smaller base.

None of these choices is dishonest. All of them shape the impression. The useful question to ask of any figure is not only what it counts but what it is divided by, and over what period. A monthly snapshot, an annual return and a rolling inspection cycle cannot be averaged together without producing something that belongs to none of them.

Inspection ratings carry their own calendar

The Care Quality Commission's ratings are frequently used as a proxy for sector health, and in a limited sense they are. But inspections are periodic, ratings persist between inspections, and the distribution of ratings in a local authority area reflects the age of the local care home stock as much as the current quality of care. A cluster of "requires improvement" ratings in one town may describe homes that have not yet been re-inspected after a change of ownership.

Ratings also measure against a framework, not against the needs of the hospital patient waiting for a placement. A well-rated residential home may be entirely unable to meet the needs of a person with complex dementia. The rating is a summary of compliance; it is not a matching service.

Case mix is the hidden variable

Care homes are not a homogeneous stock. Nursing homes, residential homes, dementia-registered units, respite provision and supported living all appear in aggregate counts, yet they serve different populations and carry different staffing requirements. A rise in occupancy across a mixed stock may reflect a shift towards nursing provision rather than a reduction in available places for the people hospitals are trying to discharge. Comparing the two without adjustment is a category error dressed as a trend.

Reading the 2026 numbers sensibly

The care home effect on hospital flow is real, and the data does describe it. But the data describes a set of constraints — registered beds, staffed beds, skill mix, geography, price and contract — rather than a verdict on a sector. A useful reading habit is to ask four questions of any figure: what exactly is being counted, on what date, divided by what, and adjusted for whom. The answers rarely make a headline simpler. They do make it more honest.