Running Dementia Review Cycles That Actually Work: Supervision, Audits and Governance That Prevent Drift
Dementia services often have good assessments and well-written plans that slowly lose impact over time. This “practice drift” happens when staffing changes, routines become pressured, and review actions are not reinforced through supervision and governance. The result is inconsistency: families feel reassured in meetings, but day-to-day delivery looks different on the floor. This article sits within Assessment, Review & Changing Needs and connects to Service Models & Care Pathways, because the mechanisms for holding review actions in place must fit the service model.
Why review cycles fail in dementia services
Most review failures are operational rather than clinical. Typical causes include:
- Actions agreed in review meetings are not translated into shift guidance
- Staff are not briefed, or the briefing is not checked for understanding
- Audits exist, but do not test whether the plan is being delivered
- Escalation routes are unclear, so concerns are not raised early
- Family feedback is received but not systematically acted on
A strong review cycle is a system: review → translate → embed → test → learn → review again.
What a “working” dementia review cycle looks like
A defensible cycle usually includes:
- Defined review triggers (falls, distress, health change, safeguarding, complaints)
- Time-bound action plan with named responsibility
- Staff briefing plus follow-up supervision check
- Audit sample that tests delivery (not just documentation)
- Family/service user check-in where appropriate
- Formal follow-up review date and governance reporting line
Operational example 1: Care home embedding review actions across shifts
Context: A resident’s review introduces a new approach to morning personal care to reduce distress. Day shift understands it, but night staff continue to wake the resident early, triggering agitation and refusals.
Support approach: The manager treats this as a review-cycle failure. The plan remains correct, but implementation is inconsistent.
Day-to-day delivery detail: The service adds a one-page “shift criticals” instruction: preferred waking time window, reassurance wording, and explicit “do not” prompts. A short night-shift huddle is held and repeated across three nights to capture all staff.
How effectiveness or change is evidenced: Distress episodes reduce and personal care refusals drop. Supervision notes show staff can explain the approach. A quality walkround confirms consistency.
Supervision: the missing link between review and delivery
Supervision should test whether review actions have become practice. In dementia services, supervision questions should be practical, for example:
- “Talk me through how you support X at tea time when they get anxious.”
- “What are the escalation triggers in the plan and what do you do first?”
- “What does least restrictive look like for this person day to day?”
This moves supervision from generic wellbeing conversations to governance-driven assurance.
Operational example 2: Homecare preventing drift in risk controls
Context: A homecare package includes a risk control that staff must confirm the person has eaten before prompting medication. New staff skip the meal check under time pressure.
Support approach: The provider runs a “micro-review” focusing on one risk control and the reasons it is non-negotiable.
Day-to-day delivery detail: The plan is updated to include an explicit checklist item with rationale. Rotas are adjusted slightly to reduce rushed calls. A supervisor completes unannounced spot checks twice weekly for a month, focusing on delivery rather than paperwork.
How effectiveness or change is evidenced: Spot check records confirm compliance, and medication timing issues reduce. The provider can evidence that governance responded quickly to drift.
Audits that test reality, not just records
Dementia audits should sample lived practice. Effective audit questions include:
- Do staff describe the same approach as the care plan?
- Do daily notes evidence the key actions introduced at review?
- Is risk management least restrictive and consistently applied?
- Are family concerns logged and linked to review actions?
Audit findings should feed into learning, training and follow-up reviews.
Operational example 3: Supported living escalation and crisis prevention
Context: A person with dementia begins leaving the property at night. Staff log incidents but do not escalate until a near-miss occurs on a road.
Support approach: The provider identifies a governance gap: escalation thresholds were unclear and staff did not feel confident acting early.
Day-to-day delivery detail: The review introduces explicit escalation triggers (two attempted exits in 48 hours prompts manager review). Staff are trained on immediate safety actions, family contact protocols, and how to update risk assessments promptly. A follow-up review is booked within two weeks.
How effectiveness or change is evidenced: Early escalation occurs on subsequent concerns, interventions are applied sooner, and risk reduces without immediate restrictive practices.
Safeguarding, restrictive practices and review-cycle assurance
Review cycles must include safeguards against informal restriction. If staff begin using locked doors, constant observation, or coercive prompts without formal review, the cycle has failed. Providers should ensure:
- Restrictions are recorded, justified, and time-limited
- Least restrictive options are evidenced
- DoLS/LPS interfaces are understood and escalated appropriately
- Reviews include family/advocate input where relevant
Commissioner expectation: stable delivery and early intervention
Commissioner expectation: Commissioners expect providers to demonstrate that review actions are embedded and maintained through supervision, audit and escalation systems, reducing crisis-driven cost and safeguarding risk.
Regulator / Inspector expectation: governance that prevents repeat failures
Regulator / Inspector expectation (CQC): Inspectors will look for evidence that governance systems identify drift, respond to emerging risks, and embed learning into day-to-day practice, not just into policy documents.
Governance checklist: what to report and review
- Number of review triggers and time to review completion
- Audit findings on implementation (not just documentation)
- Supervision sampling of “critical plan actions” understanding
- Safeguarding themes linked to review-cycle failures
- Repeat incidents for the same person and why they recurred
When these measures are visible at leadership level, dementia reviews stop being episodic and become a reliable improvement system.