Evidence that stands up: reading CQC’s six categories together

Start with the person, then find the evidence

For an illustrative example, consider whether a person receives support at the time they prefer. The care plan records a preference, but rota records suggest repeated changes and the person describes having to wait. These are competing accounts of one experience. A leader should investigate the disagreement rather than select the document that presents the most reassuring version.

Care Circle suggests choosing a small number of questions rooted in people’s lives: reliability of visits, confidence in support, access to meaningful activity or response to discomfort. For each question, identify whose account is missing. Accessible conversation, advocacy and careful observation may be more useful than another standard questionnaire. Do not assume a relative’s account always substitutes for the person’s own preferences.

Staff and partners show where the system stretches

A useful staff discussion asks what makes good care difficult during an ordinary shift. Include different working patterns and employment arrangements; the manager’s account is not automatically representative of night staff or visiting workers. Record whether employees can identify an escalation route and what happened when they used it.

Partner evidence can test the boundary between services. For example, ask whether an external professional received the information needed to act, whether an agreed referral happened and whether responsibility was clear. This is our suggested assurance exercise, not a claim that CQC requires a particular questionnaire. A constructive relationship remains compatible with documented challenge when a person’s needs are not being met.

Observation tests the process in practice

Policies and training records remain useful, but their meaning depends on how the service operates. Observe a routine interaction or handover with appropriate consent, privacy and professional safeguards. Compare what happens with the intended process and ask staff to explain any departure. Avoid turning observation into a staged performance reserved for a leadership visit.

An illustrative mismatch is a training matrix showing completion while workers remain unsure how to use the relevant equipment. The response should investigate competence, equipment, supervision and working conditions. Reissuing a certificate does not resolve uncertainty in practice. Equally, a single observation cannot establish that every shift has the same problem; record its time, context and limits.

Outcomes need context, not just a favourable direction

CQC describes outcomes in terms of the impact of care on individuals and considers measures in their service context. That does not justify a universal scorecard for every care setting. A meaningful outcome may differ between a person receiving short-term reablement and someone needing long-term support. Explain who agreed the measure and why it matters to them.

Care Circle recommends pairing counts with a denominator, period and account of changing needs. Fewer recorded incidents might indicate improvement, reduced activity or weaker reporting. Better survey results might reflect a different group of respondents. Examine plausible alternative explanations before attributing a movement to a particular intervention. Small services should make the uncertainty especially visible rather than overstate a percentage change.

Build an evidence trail that can be challenged

For one issue, retain the original concern, the competing evidence, the decision, a responsible person and the date of the next check. Keep the information proportionate and access controlled. The aim is a traceable judgement, not a duplicate library of identifiable care records scattered across committees and shared drives.

Where evidence is incomplete, state that explicitly. A missing feedback route is a question to resolve, not a positive assurance finding. Where sources disagree, preserve the disagreement until someone has investigated it. Leaders can then explain both their current conclusion and what new evidence would cause them to change it. That is more useful than a dashboard whose green status cannot be interrogated.

Evidence quality also depends on how the question was asked. If a leader collects feedback only immediately after a positive interaction, the resulting account may miss difficulties at other times. Record the collection method and include inconvenient experiences in the review. Where a person needs support to communicate, explain how that support was provided and whose interpretation appears in the record. This makes the evidence easier to evaluate without pretending that a small local exercise represents every person using the service.