Adelfi Care Services · Domiciliary care · Prestbury, Cheltenham
Good on all five key questions. Three of them would read differently under the rules now in pilot.
Adelfi holds a Good rating on every key question — per CQC’s published assessment, dated 2 October 2025, and a position rebuilt from Requires improvement in 2022. The draft framework CQC is piloting this year removes the averaging that a rating like this quietly leans on. This read shows where that bites, against the published record of 268 rated homecare peers.
The 30-second read
Position
Good on all five key questions, published 2 October 2025. Overall score 74% against a 72.4% average across 268 rated South West homecare agencies.
What’s changing
From end-2026 each key question is judged against written descriptors — on its weakest evidence, not its average. Pilot assessments run June–October 2026.
The pattern
Four statements scored 2 in July’s assessment. On the draft reading, Safe, Effective and Well-led would each read Requires improvement.
The 30 minutes
Where your evidence sits against the new descriptors today is the one thing this document cannot see. We walk the four statements together and narrow six items to your first three.
Taken by Alexey Litvin, GreenM’s CEO. Nothing to prepare — we bring the analysis.
Your published position
Live register · verified 17 Aug 2026
Everything below comes from CQC’s published assessment of 16–29 July 2025 and the live register, checked on 17 August 2026. At the assessment, 12 people were receiving personal care from the service.
Overall rating
Good
All five key questions Good · published 2 Oct 2025
Overall score
74%
Cohort average 72.4% · 268 rated SW homecare agencies
Statements scored
33
27 at Good · 2 at Exceeds · 4 below · 1 unscored
Breach recorded
1
Notifications of other incidents — the report’s one regulation breach
APR 2021
Registered with CQC.
MAY 2022
First inspection: Requires improvement, with a breach of regulation.
APR 2023
Re-inspection of Safe and Well-led: overall rating moved back to Good.
JUL 2025
Assessment under the current framework: Good on all five, 33 of the 34 statements scored. Published 2 October 2025.
NOW
Draft sector frameworks piloting June–October 2026; in force from the end of 2026. The rating is current — the method that produced it is being retired.
The rules this read is scored against
Three changes · six shifts
The six shifts from the guide you downloaded are the rules this analysis applies to Adelfi’s own record — ratings judged rather than calculated, four sector-specific frameworks, and evidence read continuously from your systems rather than assembled for the visit.
Outcomes over processContinuous monitoringInequalities in scopeWorkforce in well-ledJudged, not scoredSector-specific KLOEs
The pattern
Cohort n=268 · scores n=260
Adelfi against the 268 homecare agencies in the South West rated under the current framework (260 carry published scores; cohort re-derived from the CQC register on 17 August 2026). Above the average on three of five key questions — and the average is about to stop mattering.
Safe
69% avg 69.9%
Effective
71% avg 72.6%
Caring
80% avg 75.2%
Responsive
75% avg 73.3%
Well-led
75% avg 71.1%
■ Adelfi ■ cohort average · key-question scores per CQC’s published statement scoring
Judged, not scored: what flips
Under the draft framework each key question is judged against written descriptors, so a weak statement can no longer be averaged away. Applied to Adelfi’s own July 2025 statement scores:
Key question
Published
Weakest scored statement
On the draft reading
Safe
Good · 69%
Involving people to manage risks (2); Medicines optimisation (2)
Would read Requires improvement
Effective
Good · 71%
Consent to care and treatment (2)
Would read Requires improvement
Well-led
Good · 75%
Governance, management and sustainability (2)
Would read Requires improvement
Caring
Good · 80%
Four statements at 3; Workforce wellbeing at 4
Holds, with margin
Responsive
Good · 75%
All seven statements at 3
Holds — no headroom
Illustrative outcome under the draft frameworks — the descriptors are still in draft, and CQC has decided nothing about Adelfi. Rule modelled: a key question is judged on its weakest scored statement.
Well-led scores above the cohort average (75% against 71.1%) — and still flips. The average was never going to be the unit of judgement.
Each with Adelfi’s score, the cohort average, and the share of the 268-agency cohort scoring it below Good. All four sit among the cohort’s five weakest statements — this is where inspectors keep finding the sector short, which is exactly where the new framework points.
Governance, management and sustainability
2 / 4
cohort avg 2.45 · 42% scored below Good
The report records the assessment’s one regulation breach here: 3 safeguarding and 2 serious injury notifications from 2024 onwards had not been submitted to CQC, and the monthly audits had not identified it. Two were submitted retrospectively after the inspection.
Medicines optimisation
2 / 4
cohort avg 2.67 · 27% scored below Good
Inspectors found recording “at times lacked clarity and consistency” — MAR times that did not match how one person’s medicine was actually taken, and PRN records without the reason and effect the provider’s own policy requires.
Involving people to manage risks
2 / 4
cohort avg 2.73 · 25% scored below Good
Accident and incident logs “did not always show the outcome of any injury or hospital admission”, nor whether a notification or safeguarding referral was required — the same thread the governance finding pulls on.
Consent to care and treatment
2 / 4
cohort avg 2.77 · 22% scored below Good
For one person unable to consent, no decision-specific mental capacity assessment or best-interest decision had been completed for their care or the use of bed rails, while the family’s deputyship was still in progress.
From the inspector narrative · 3 strengths · 3 watch items
Against a cohort average of 2.94. Privately funded counselling, monitored workloads, no agency staff. Staff told inspectors, “I feel valued at Adelfi.”
⚠ Watch
Incident follow-through
The report notes initial medical advice “was not always sought” after incidents, and staff-team learning from those events was absent. Already a scored 2 — the same statement the Safe flip turns on.
⌂ Strength
Learning and innovation scored 4
“A strong focus on continuous learning, innovation and improvement” — an emergency lifting chair with a post-falls protocol, and baseline health-monitoring equipment inspectors linked to preventing avoidable admissions.
⚠ Watch
PRN medicines record
After the provider’s post-inspection fix, PRN reasoning was recorded — but the report notes effectiveness still was not, as the provider’s own medicines policy requires. Part of the scored-2 medicines statement.
⌂ Strength
Staffing consistency
No agency use; electronic call data showed staff “arrived promptly and stayed for the required duration”. Recruitment described by inspectors as robust and safe.
⚠ Watch
Notification decision-making
The provider described the missed notifications as an administrative oversight; inspectors recorded a lack of understanding about when to notify. The audit that should catch this is the statement scored 2 in Well-led.
Audit follow-through that stops before resolution; issues identified but not tracked to closure; evidence spread across disconnected systems. Adelfi’s version is specific: the notifications gap sat outside the monthly audit’s field of view.
Medicines — recording, not practice
MAR inconsistencies, missing PRN protocols and follow-up, time-critical medicines unflagged. Most cohort findings are about what the record shows, not what staff did.
Risk involvement — the person’s voice
Generic risk templates, unclear escalation criteria, incident logs without outcomes. Inspectors read the trail from event to decision — where it breaks, the statement scores 2.
Consent — decision-specific or nothing
Capacity assessments that are general rather than per-decision, delayed best-interest records, authority assumed from next-of-kin status. The framework’s bar here is documentary and precise.
The honest blind-spot
What we cannot see
The fieldwork behind every number above finished on 29 July 2025. What the published record cannot show — and what we cannot see from outside — is where Adelfi’s evidence sits today against descriptors that did not exist when it was gathered: which of the four statements has already been closed, and whether the closure reads as connected evidence or as memory. That gap is what the 30 minutes is for.
What we’d work through together
Six items · 30 minutes
1
The four statements scored 2 that decide whether Safe, Effective and Well-led hold
First · time-sensitive
2
The notification decision trail behind July’s breach finding
Time-sensitive
3
Medicines records that read the way the visits actually run
Time-sensitive
4
Where consent evidence has to be decision-specific
Continuous
5
The two statements scored 4 as the margin under the new descriptors
Continuous
6
Responsive’s zero headroom — seven statements sitting exactly on the Good floor
Continuous
Show what each item turns on
1 — which evidence already exists in your systems against each of the four, and which closure is currently only a conversation.
2 — how an incident becomes a notify-or-not decision in your process, and whether that decision leaves a trace an inspector can read.
3 — what your MAR and visit notes show against your own medicines policy on an ordinary day, unprepared.
4 — which people currently lack a decision-specific capacity record, and who holds the authority in each case.
5 — how the workforce-wellbeing and learning evidence keeps carrying weight once averaging is gone and descriptors take over.
6 — which of the seven Responsive statements sits nearest the edge on your current evidence — only your side can see it.
These are the discussion, not a task list — each one turns on something only visible from inside your systems, and they are covered live on the call.
Inspection-ready before the inspector
GreenM are data and AI specialists: we connect the evidence already sitting in your systems — visit records, incidents, audits, training — so it reads cleanly against the new statements on an ordinary day, not assembled the week before. The pilots are running now, June to October 2026, with 9,000+ reports targeted by September and the frameworks in force from the end of the year. Adelfi’s rating is current; the method that produced it is not. In 30 minutes we narrow the six items above to the three that matter first for Adelfi.
Taken by Alexey Litvin, GreenM’s CEO. Nothing to prepare — we bring the analysis.
Prefer email? Reply to the note this came with, or write to hello@greenm.io and we’ll send a few times.
Source: CQC published assessment for Adelfi Care Services, location 1-10717475436 (assessment 16–29 July 2025, report published 2 October 2025), verified against the live CQC register on 17 August 2026. Peer cohort: the 268 homecare agencies in the South West holding a rating under the current framework (260 with published statement scores), re-derived from GreenM’s copy of the CQC register on 17 August 2026. Key-question percentages follow CQC’s published statement scoring. “Judged, not scored” models the draft sector frameworks (consultation closed 12 June 2026) and is illustrative only. Inspector quotations are verbatim from the published report.
