CQC inspections: the evidence ADHD and autism services need

The Care Quality Commission (CQC) inspects ADHD and autism services that prescribe, titrate or treat in other ways. The government has asked CQC to regulate services that only diagnose as well, once new regulations are in place; our explainer on the change covers who has to register and when. This guide covers what inspectors ask registered services to show. Which problems they raise most often is set out in our analysis of 1,520 CQC inspection reports.

The main points are:

• CQC rates ADHD and autism services on the same five key questions as other health services. There is no separate framework for neurodevelopmental services.
• CQC is testing new frameworks for each sector, with no date yet for them to apply, but what inspectors ask to see is unlikely to change.
• Inspectors ask for evidence on staff training, clinical governance, prescribing and titration, feedback and complaints, records, and data security.
• This evidence is easiest to show when it is recorded as work happens, with one person responsible for each area.

1. CQC rates ADHD and autism services on the same five key questions as other health services#

CQC's single assessment framework rates every health and care service on five key questions: is it safe, effective, caring, responsive to people's needs, and well-led? Each key question gets one of four ratings: Outstanding, Good, Requires improvement or Inadequate. There is no separate framework for neurodevelopmental services, so an ADHD or autism service is assessed against the same statements as any other clinic. This is the current framework, which still applies while CQC tests its replacement.

Under the key questions sit 34 quality statements, each describing what a good service does. Inspectors judge each statement on evidence from six evidence categories, from what patients say to the service's own records. Which categories count most depends on the type of service and on whether it is being assessed at registration or as an existing service.

CQC judges every service on 5 key questions, through 34 quality statements and 6 kinds of evidenceThe five key questions, how many quality statements sit under each, and where inspectors get their evidenceSafeAre people protected from avoidable harm?8 quality statementsFor example: Safe and effective staffing; Medicines optimisationEffectiveDoes care follow guidance and get good results?6 quality statementsFor example: Assessing needs; Consent to care and treatmentCaringAre people treated with kindness and respect?5 quality statementsFor example: Kindness, compassion and dignityResponsiveDoes the service meet people's needs?7 quality statementsFor example: Listening to and involving people; Equity in accessWell-ledDo leaders know what is happening and act on it?8 quality statementsFor example: Governance, management and sustainabilityInspectors gather evidence in 6 waysPeople's experienceFeedback from staff and leadersFeedback from partnersObservationProcessesOutcomesInspectors judge each quality statement on evidence from these six categories. Which categories count most depends on the type of service.Source and notes

2. CQC is testing new frameworks with no start date, but what inspectors ask to see is unlikely to change#

Dr Penny Dash's review of how well CQC operates, published in October 2024, recommended changing the framework. After its Better regulation, better care consultation closed on 11 December 2025, CQC published draft frameworks on 24 March 2026 for four sectors: adult social care, mental health care, primary care and community services, and hospitals. Feedback on the drafts closed on 12 June 2026, and CQC piloted them with volunteer providers from June to October 2026, with a final evaluation due in November.

CQC had planned to publish the final frameworks in summer 2026 and start using them at the end of the year. Its August 2026 update gives no date, so services are still assessed under the current single assessment framework.

The drafts keep the five key questions and the four ratings. They replace the 34 quality statements with key lines of enquiry, written as questions about what inspectors will look for, and they drop scoring, so ratings are given directly for each key question. They do not change the regulations services must meet. In our view, the evidence that shows a service meets them (training, governance, prescribing safety, feedback, records and data security) is unlikely to change.

CQC is testing new frameworks, with no date yet for them to applyOct 2024Dash review of how well CQC works is published11 Dec 2025Better regulation, better care consultation closes24 Mar 2026Draft frameworks for 4 sectors publishedJun to Oct 2026Drafts piloted with volunteer providersNov 2026Final evaluation of the pilotsNo date setFinal frameworks start to applyUntil the final frameworks apply, services are assessed under the current single assessment framework. CQC had planned to publish them in summer 2026; its August 2026 update gives no date. Source: CQC updates of 24 March, 4 June and 27 August 2026.

3. Inspectors ask for evidence on training, governance, prescribing, feedback, records and data security#

Each area of evidence feeds named quality statements:

Staff training must be shown in one up-to-date record#

Safe and effective staffing asks whether there are enough skilled, competent people to give safe care. Inspectors ask who is trained on which tools (such as ADOS-2, ADI-R, QbTest and structured ADHD interviews) and how the service knows they are still competent. They ask for:

• a training matrix that is kept up to date
• certificates with dates
• competency sign-offs for new starters and locums
• DBS checks
• evidence that clinical supervision happens as often as the service says it does.

Clinical governance evidence should show what changed after each incident or audit#

Governance, management and sustainability (under Well-led) and Learning culture (under Safe) look for governance on a regular cycle:

• a clinical audit programme with results and repeat audits
• governance meetings with minutes and tracked actions
• an incident log showing investigation and learning
• risk registers that are reviewed.

Inspectors may pick an incident or audit finding from six months ago and trace what changed because of it. For adult ADHD services, auditing against a published standard makes this evidence stronger (see our AQAS audit checklist).

Titration needs NICE NG87 checks at baseline, each dose change and every six months#

Medicines optimisation matters for ADHD services that treat. Stimulant and non-stimulant titration needs a recorded baseline and ongoing checks, as set out in NICE guideline NG87.

Inspectors ask:

• how monitoring is scheduled and what happens when a patient misses it
• how prescribing decisions are recorded
• how patients are handed over to GPs under shared care, including what happens when a GP declines.

Services that only diagnose should show how they hand over safely to whoever does prescribe.

Feedback evidence needs a complaints log with themes and changes made#

Listening to and involving people needs evidence of what patients say about the service. Inspectors look for a complaints log with dates, responses, outcomes and themes, and examples of changes made because of feedback. Routine outcome and experience measures across the pathway also count as evidence of outcomes.

Inspectors sample records for consent, questionnaires and traceable report conclusions#

Under Assessing needs and Consent to care and treatment, inspectors sample clinical records and check that each has:

• referral information
• consent, including for recording assessments and for sharing information with schools and GPs
• patient and informant questionnaires
• the assessment itself
• a diagnostic report whose conclusions can be traced back to the evidence behind them.

Services that use AI to draft reports can expect to be asked where the content came from; our guide to AI clinical reports that hold up at audit sets out what to keep.

Providers with NHS patient data, including Right to Choose, must complete the DSPT#

Data security evidence sits under Well-led. For NHS work the main document is the Data Security and Protection Toolkit (DSPT), an annual self-assessment against the National Data Guardian's ten data security standards. NHS England says every organisation with access to NHS patient data and systems must complete it, including Right to Choose providers. As well as DSPT status, inspectors ask:

• who can access clinical records and how access is controlled
• whether staff log in with a second check (multi-factor authentication)
• where data is stored
• what the audit trail records
• how a breach would be spotted and reported.

Assessment services hold very sensitive data, such as developmental histories, school reports and recordings of children, so a claim to use a secure system needs evidence behind it.

4. Evidence is easiest to show when it is recorded as work happens, with one person responsible for each area#

Which problems inspectors raise most often, and how often, is set out in our analysis of 1,520 CQC inspection reports. Gaps like these do not always mean care is poor, only that it cannot be shown. Each one closes with a habit that produces the evidence as work happens:

Each area also needs one person responsible for it, and governance needs a dated cycle, so the evidence does not depend on one person remembering.