30-second briefing
Read the rating in three layers
- Rating: CQC’s most recently published judgement – Good means the service was judged to be performing well and meeting expectations.
- Recency: the interval since publication of that judgement; it is an assurance-age measure, not evidence of deterioration, safety failure or present quality.
- Current evidence: what people experience now, how risks and outcomes are changing, and whether the provider can demonstrate learning, governance and improvement since the rating.
Two services can both display Good while one rating was published in 2026 and the other in 2019. The regulatory word is identical. The elapsed time since the underlying judgement is not. That difference should influence the questions a board, commissioner, family, lender or prospective employee asks – but never become a shortcut to a negative quality judgement.
CQC defines Good as performing well and meeting its expectations. Regulation 20A requires providers that have received a performance rating to display the most recent rating. Neither the rating scale nor the display duty turns the publication date into an expiry date. An older Good rating remains the latest published CQC judgement unless and until it changes; age tells us only how recently that judgement was established or refreshed.
The assurance gap is therefore informational. The further the publication date recedes, the more work current evidence must do: recent experiences, outcomes, incidents, safeguarding, workforce stability, medicines, complaints, audits and completed improvement. CQC’s new return to selected aged-Good services confirms that the regulator also sees value in refreshing part of this picture.
| Measure | Value |
|---|---|
| Outstanding | 6.6 years |
| Good | 4.7 years |
| Requires improvement | 3.0 years |
| Inadequate | 0.5 years |
Descriptive of this snapshot, not a causal relationship and not a current-quality comparison. Aged ratings are more common among higher published ratings in a risk-informed assessment system.
The central finding is reproducible – with a small snapshot caveat
The source article reports a Care Circle Provider Intelligence Observatory snapshot at 9 August 2026: 29,545 active adult social care locations, of which 22,786 had a usable overall rating of Outstanding, Good, Requires Improvement or Inadequate. It reports 12,047 usable ratings at least four years old (52.9%), and 7,693 Good ratings at least six years old (42.1%). Those exact house counts are not CQC-published statistics.
For this rebuild, the calculation was independently repeated on CQC’s official 4 August 2026 active-locations file. Filtering to Adult social care, Social Care Org and not dormant produced 29,528 locations and 22,783 usable overall ratings. At a 9 August reference date, 12,045 were at least four years old (52.9%); among 18,277 Good ratings, 7,692 were at least six years old (42.1%). The variance from the five-days-later house snapshot is 17 locations, three usable ratings, two four-year records and one six-year Good record.
That near-replication supports the headline proportions. It does not make the exact house method fully reproducible: the 9 August input extract, transformation code and row-level output were not supplied. The small variance is consistent with a timing or transformation difference, but its cause remains unresolved. Publication should preserve that distinction.
Rating age is not a proxy for current quality or safety
Fact: the publication date records when the latest rating was published. It does not measure what has happened on every day since, whether CQC has received information about the service, or whether commissioners and providers have carried out other assurance work. CQC says assessment frequency depends on information received and evidence collected; risk can trigger regulatory attention without a fixed reinspection interval.
Analysis: an older rating creates greater uncertainty about how well that historic judgement describes today’s organisation because managers, ownership, workforce, people supported, systems and expectations may have changed. That is an evidence-recency problem, not evidence that quality declined. A provider may have improved, maintained quality or deteriorated; rating age alone cannot identify which.
Interpretation: use age as a prompt for proportionate due diligence. Never label an aged Good service poor, unsafe, expired or out of date solely because of the publication date. Equally, do not treat Good as proof that every present-day control remains effective.
Risk-led regulation explains part of the pattern; capacity explains another part
The audited file shows median ages of approximately 6.6 years for Outstanding, 4.7 for Good, 3.0 for Requires Improvement and 0.5 for Inadequate. This is consistent with a risk-informed regulator returning sooner where evidence indicates greater concern. It should not be interpreted as higher ratings causing delay or fresher ratings indicating worse current care.
CQC’s own performance data also records a sharp fall in output across health and social care: 3,825 assessment reports published in 2024/25, compared with 6,230 in 2023/24 and 10,356 in 2022/23. CQC attributes the reduction to technology problems, changes to its assessment framework and operational restructure. In May 2026 it said it remained on track to publish reports for at least 9,000 assessments across all sectors by September 2026, while prioritising very-high-risk services, services unassessed for more than a year and ratings over six years old.
The 9,000 commitment is an all-sector activity target, not a promise to refresh every aged adult social care rating. The size of the age profile means current provider assurance remains necessary while regulatory capacity rebuilds.
CQC’s return to Good services is deliberately selective
On 30 July 2026, CQC published its ‘Returning to good adult social care services’ approach. It says the purpose is to confirm whether a Good rating remains accurate, with greater emphasis on people’s experiences and outcomes, observation and targeted record checks. The route covers selected care homes, supported living and homecare services; it is not a universal programme for every aged Good rating.
Eligibility includes Good overall and in all five key questions, a rating at least six or seven years old, a registered manager, no indication of a rating change and no ongoing enforcement. CQC also excludes services providing care or treatment to children and applies a condition where the learning-disability-and-autism service-user band was added after the last Good rating. If planning identifies concerns or outstanding practice, CQC says it will expand the assessment and revert to its usual published approach.
Operational implication: an old Good rating may increase the prospect of contact, but age alone does not establish eligibility, timing or outcome. Providers should not market themselves as selected, cleared or likely to retain Good unless CQC has said so.
Build a living quality picture between published ratings
A current assurance layer should connect evidence rather than assemble a ceremonial inspection folder. Boards and registered managers need trends in people’s outcomes and experience, safeguarding, incidents and near misses, complaints, medicines, staffing sufficiency and stability, competence, care-plan review, environmental risk, audits and the completion and effectiveness of actions.
The stronger question is not ‘Are we still rated Good?’ but ‘What current evidence would allow us to explain why care is Good today, where confidence is weaker and what we are improving?’ That framing leaves room for honest exceptions. A credible assurance view shows deterioration and uncertainty as well as positive performance, names accountable owners and demonstrates whether interventions worked.
Suppliers can help integrate data, run independent audits, improve evidence retrieval or support action tracking. They should not promise a CQC rating, invent a proprietary ‘current CQC score’ or substitute document volume for people’s lived experience and outcomes.
Commissioners, families and the market should read rating plus recency plus evidence
The rating remains a major public signal and a lawful display requirement. Recency adds context. Current evidence adds assurance. A proportionate reader should inspect the latest report and publication date, understand what service or location was rated, check later CQC information and ask what has materially changed since. CQC’s directory warns that some multi-service locations no longer receive one location-level rating under the newer approach, so service-level ratings may be the relevant unit.
For commissioners, investors and lenders, age can trigger deeper review but should not become an automatic exclusion rule. For families, it should prompt questions about current leadership, staffing, incidents, complaints, experience and improvement – not a conclusion that the service is unsafe. For providers, explaining what has changed since the report is more informative than hiding the date or relying on the certificate alone.
The sector does not need an unofficial replacement rating. It needs clearer provenance: latest CQC judgement, date and scope of that judgement, and transparent current evidence from the organisations responsible for care.
Questions leaders should ask now
- 01
What exactly was rated – and when?Confirm the service or location, overall and key-question scope, publication date, inherited status and any changes in registration or ownership.
- 02
What has materially changed since?Map leadership, workforce, people supported, clinical responsibility, digital systems, premises, operating model and major incidents or improvements.
- 03
What evidence is current?Use recent experience and outcome evidence alongside safeguarding, incidents, medicines, complaints, staffing, audits and completed actions.
- 04
Where is assurance weakest?Name missing, old or contradictory evidence and unresolved actions rather than filling a dashboard with only positive measures.
- 05
Are we describing provenance honestly?Separate CQC’s published rating, Care Circle analysis, commissioner assurance and provider self-assurance; do not merge them into one implied regulatory judgement.
The Care Circle view
The Care Circle view
A six-year-old Good rating is not a warning label. It is CQC’s latest published judgement plus a date. The judgement deserves to be stated accurately; the date deserves to be visible. Combining them produces a more informed question about assurance, not a verdict on current care.
Strong providers should be able to show what happened after the report: what people experience now, where risks changed, what leaders learned and whether improvement delivered better outcomes. The aim is not to manufacture a fresh rating in-house. It is to make present quality visible, testable and honest while the regulator refreshes its own evidence.
Continuing coverage
Follow the question into the later editions.
After the rating: the evidence a board needs this month · 9 October 2026
Develop the analysis
Read the connected flagship reports.
Workforce & delivery: turning sector improvement into dependable care
Provider resilience: the capacity, cash and care behind the headline
October cost controls: turn funding, learning and energy changes into a usable plan
Sources, method & limitations
How to read this analysis
The source article states that its Care Circle Provider Intelligence Observatory snapshot covered 29,545 active adult social care locations on 9 August 2026 and calculated age from the latest overall-rating publication date. This rebuild independently audited the central claim using CQC’s ‘Care directory with filters’ dated 4 August 2026. Filters were Location Inspection Directorate = Adult social care; Location Type/Sector = Social Care Org; Dormant = N. Usable ratings were Outstanding, Good, Requires improvement and Inadequate. Age thresholds were inclusive: publication on or before 9 August 2022, 2021 and 2020 for four, five and six years respectively. The official-file reconstruction returned 29,528 locations; 22,783 usable ratings; 12,045 aged at least four years; 9,667 at least five; 8,558 at least six; and 7,692 Good ratings at least six years. It also returned 14,741 care homes and 484,219 beds; 7,961 homes with no usable rating or a rating at least four years old accounted for 238,030 beds, or 49.2%. These closely corroborate, but do not exactly reproduce, the five-days-later house figures.
- The exact 9 August 2026 Observatory extract, row-level result and transformation code were not supplied. Exact house counts therefore remain only near-replicated, not independently reproduced.
- The 4 August CQC file is a monthly snapshot, while CQC says its API updates daily. Timing can change active status, ratings and capacity; the observed variance cannot be attributed without the house extract.
- CQC warns that migration to a new directory system has delayed some updates, including registration cancellations. ‘Active’ and ‘not dormant’ do not guarantee current service delivery.
- Multi-service locations may have service-level rather than location-level ratings under CQC’s newer approach. The active-locations file’s latest overall location rating may not capture every service-level rating.
- Publication date is used as rating age. It is not the inspection visit date, a measure of all CQC contact, or a measure of current quality, safety or compliance.
- Care-home bed figures are registered capacity in the directory, not occupied beds, available beds or a quality-weighted measure.