Where UK Care Homes Actually Fail Inspection
The CQC inspects a UK care home against five questions. The one almost every home passes is "Are you caring?" The two that fail at four to five times the rate are "Are you well-led?" and "Are you safe?" When a home is rated Requires Improvement, the cause is almost never the care itself.
Published 23 May 2026 · based on the May 2026 CQC directory + enrichment snapshot
When a UK care home receives a Care Quality Commission inspection, the inspector scores five separate dimensions, each on the same 4-point Outstanding to Inadequate scale: Caring, Responsive, Effective, Safe and Well-led. The five scores are then combined into a single overall rating. The published headline is usually that overall rating. But the five scores beneath it tell a much more specific story about where the sector struggles.
We took the May 2026 CQC enrichment data, filtered to care homes and nursing homes with a published rating on each of the five sub-scales (13,633 sites), and looked at how often each dimension passes.
| Dimension | What it tests | % Good+ | % Below |
|---|---|---|---|
| Caring | Treats people with kindness, dignity, respect, compassion | 96.0% | 4.0% |
| Responsive | Meets people's individual needs and preferences | 93.5% | 6.5% |
| Effective | Care, treatment and support achieve good outcomes | 90.7% | 9.3% |
| Safe | Protects residents from abuse and avoidable harm | 80.7% | 19.3% |
| Well-led | Has leadership, management and governance that delivers high-quality care | 77.6% | 22.4% |
What the gap means
The public mental image of a "Requires Improvement" care home tends to be a place where residents are neglected, isolated, or unhappy. That doesn't match the data. The "Caring" rating, which asks whether residents are treated with kindness, dignity and respect by the people looking after them, passes at 96%. Responsiveness, which asks whether residents' individual needs and preferences are being met, passes at 93.5%.
What fails is structural. "Well-led" asks whether the home has the leadership, management and governance to deliver consistent care. "Safe" asks whether the home has the policies, training, recruitment checks, medication records and incident reporting to keep residents from preventable harm. Both fail roughly one in five inspections. The day-to-day care can be excellent while the paper trail behind it, the rota systems, the audit records, the safeguarding referrals, are inadequate.
This matters for how the sector is understood and improved. It is much easier for a chain to push management discipline across a property than to teach kindness. If the rating data is right about where the failures concentrate, the route to fewer failing homes is operational, not vocational.
The chain effect is real, but narrower than you'd expect
Operators with 100 or more care homes have an Inadequate rate of 0.22% across the group. Single-site independent providers run at 1.38%, six times higher. So scale matters for the floor. But scale doesn't lift the ceiling much: the share of Outstanding ratings is roughly 4 to 5.7% across every size bucket. Big operators are better at avoiding disaster, not at producing exceptional homes.
| Operator size | Homes | % Outstanding | % Good | % Req Imp | % Inadeq |
|---|---|---|---|---|---|
| Independent (1 site) | 5,224 | 4.02% | 75.3% | 19.3% | 1.38% |
| Mini (2-4 sites) | 2,925 | 4.07% | 78.4% | 16.6% | 0.96% |
| Small (5-9 sites) | 1,721 | 5.69% | 78.8% | 15.1% | 0.46% |
| Mid (10-24 sites) | 1,689 | 4.14% | 81.8% | 13.4% | 0.65% |
| Big (25-99 sites) | 1,162 | 5.08% | 82.5% | 11.9% | 0.52% |
| Mega (100+ sites) | 912 | 5.37% | 82.3% | 12.1% | 0.22% |
The biggest operators, by pass rate
| Operator | Sites | % Good or Outstanding | % Req Imp | Inadeq |
|---|---|---|---|---|
| Voyage 1 Limited | 239 | 91.6% | 8.4% | 0 |
| Barchester Healthcare Homes | 194 | 88.6% | 10.8% | 1 |
| Achieve Together | 144 | 80.6% | 19.4% | 0 |
| Anchor Hanover Group | 119 | 85.7% | 14.3% | 0 |
| HC-One Limited | 113 | 84.0% | 15.9% | 0 |
| Care UK Community Partnerships | 103 | 93.2% | 5.8% | 1 |
| Sanctuary Care Limited | 83 | 95.2% | 3.6% | 1 |
| Methodist Homes | 69 | 89.8% | 10.1% | 0 |
| Community Homes of Intensive Care and Education | 66 | 92.4% | 7.6% | 0 |
| Runwood Homes Limited | 59 | 81.4% | 18.6% | 0 |
Half the published ratings are over three years old
CQC inspections were paused through most of 2020 and slowed for years afterward. Combined with a quietly admitted backlog at the regulator, the result is that a lot of the ratings currently in the public window are stale. Of the 13,700 rated UK care and nursing homes in this data, 5,850 were last inspected over 5 years ago. Another 3,129 were inspected 3 to 5 years ago. The 1,989 that have a fresh inspection inside the last 12 months are only 15% of the total.
A 5-year-old Good rating still gives a family some reason for confidence, but the building's manager, ownership, staffing and resident mix may all have changed since. When choosing or reviewing a care home, the inspection date is at least as important as the inspection score.
What this isn't saying
The "Caring" rating passing at 96% is not a claim that every UK care home is delivering excellent care. The CQC's Caring rating is an inspector's judgement on a specific visit, scored against the regulator's published framework. The methodology misses some things and is biased by what an announced inspection can observe.
The chain-by-size finding is not a recommendation. A 100-home corporate operator and a single-site family-run home are different products. Both groups contain excellent and poor examples. The 0.22% vs 1.38% Inadequate rate is a fact about averages, not about any specific home you might be considering.
Methodology
Source: the CQC public directory CSV (May 2026) joined to the CQC API V1 ratings snapshot. We restrict to locations where primary_category is care_home or nursing_home and the overall rating is one of Outstanding, Good, Requires improvement or Inadequate (n = 13,633). KLOE percentages are computed independently per dimension; a home counts in the denominator for each dimension it has a published rating on. Operator-size buckets are computed at the cqc_provider_id level using the same rated-locations population. Inspection recency uses last_inspection_date and the snapshot date for relative bucketing.
CQC data is published under the Open Government Licence v3.0. The CQC's classification of GPs, dentists, hospitals, homecare agencies and other regulated services is excluded from this analysis.
Check any of these for yourself: browse every operator, filter by rating, or search a specific home.