From Requires Improvement to Good: building CQC evidence that holds up

Why one rating hides several separate judgements

Under the Single Assessment Framework, CQC doesn't arrive at Good or Requires Improvement as a single overall verdict — it scores each of the 34 quality statements against evidence, rolls those up into a rating for each of the five key questions (Safe, Effective, Caring, Responsive, Well-led), and only then aggregates to an overall rating. The detail that catches services out: a single key question rated Inadequate can cap the overall rating, regardless of how strong the other four are. A service that's excellent on Caring and Responsive but has a genuinely weak Safe picture won't be rated Good overall — which means an improvement plan built only around your weakest single quality statement, rather than your weakest key question as a whole, can miss the thing actually capping your rating.

Where inspectors actually look first

In practice, a handful of things surface disproportionately often in Requires Improvement findings, because they're both easy for an inspector to check quickly and hard for a service to hide:

• Weak or missing evidence behind a quality statement — a strong practice with no paper trail scores the same as a weak practice, because CQC assesses the evidence, not your own confidence in the service.
• Overdue improvement actions — a previous finding logged but not closed out is one of the fastest ways to confirm a pattern rather than a one-off.
• Expired mandatory training — a simple, objective check that inspectors can verify in minutes, and one of the most common Requires Improvement drivers in practice.
• Care plans past their review date — evidence that the "person-centred, kept current" standard the framework expects isn't actually happening in practice.
• Unsent statutory notifications — CQC expects notification of specific events (safeguarding concerns, deaths, certain injuries, absconding, and more) within set timeframes; a pattern of late or missing notifications undermines Safe and Well-led together.

Notice that none of these require a dramatic care failure — they're administrative and evidential gaps that compound into a rating drop. This is good news for an improvement plan: the fastest route back to Good is usually closing these gaps systematically, not reinventing how care is delivered.

Building the evidence trail, not just the practice

A common and expensive mistake is improving the actual care delivered without improving the record of it. CQC assesses what it can evidence, and "we know we do this well" carries no weight without a trail showing it — the ratings drop from evidence gaps, not necessarily from care gaps. For each of the 34 quality statements, a workable evidence structure records:

• Self-assessment rating — an honest internal view (Outstanding / Good / Requires improvement / Inadequate), not the rating you hope an inspector reaches.
• Where the evidence lives — the specific document, log or system, not a vague "we have processes for this."
• The gap — what's missing between the self-assessed rating and Good, named specifically enough to act on.
• The improvement action — owner and deadline, tracked to actual closure, not just raised and forgotten.

The discipline that separates services that climb back to Good from those that stall is keeping this current continuously rather than reconstructing it under pressure once an inspection is announced (or after one has already happened) — a scramble produces evidence that looks assembled for the occasion, which inspectors are trained to spot.

A practical sequence for the improvement plan

1. Self-assess honestly against all 34 quality statements first, even the ones you're confident about — a gap you don't know exists is the one an inspector will find.
2. Identify which key question is actually capping the rating, not just which individual quality statement scored lowest — fixing the wrong thing wastes the improvement window.
3. Clear the administrative backlog fast — overdue actions, expired training, unreviewed care plans, unsent notifications. These are the quickest wins and the ones inspectors check most reliably.
4. Re-score honestly on a fixed cycle (monthly is common for services actively working back to Good), not just before an anticipated inspection.
5. Keep evidence of the trend, not just the current snapshot — a service that can show steady closure of gaps over months tells a more credible improvement story than one presenting a suddenly-tidy register the week before a visit.

The audit-trail question CQC will ask

Beyond the content of your evidence, inspectors increasingly probe its integrity — can you show a record wasn't quietly tidied up after the fact? An append-only log of who changed what and when, that can't be edited or deleted retrospectively, answers that question directly rather than asking an inspector to take your word for it. This is the standard CareIQ's audit trail is built to — every create and update stamped with the signed-in user, insert-and-read-only at the database level, so the evidence trail itself is part of the evidence.

Common mistakes on the way back to Good

1. Improving the wrong key question because the lowest individual quality statement wasn't the one actually capping the rating.
2. Fixing the practice but not the record of it, then having nothing to show at assessment.
3. Letting the self-assessment slide back into an aspirational rather than honest exercise once the pressure of a recent inspection fades.
4. A visible spike of activity in the weeks before an expected inspection, rather than a sustained trend — inspectors read the pattern, not just the current state.
5. Treating statutory notifications as paperwork rather than as a Safe/Well-led evidence source in their own right.

This guide is general information, not legal, tax or compliance advice. Rules change — always check the current official guidance for your situation.