Reviewing Dementia Risk Assessments: How to Update Falls, Wandering and Self-Neglect Controls Without Over-Restricting

February 5, 2026

In dementia services, risk is dynamic. A risk assessment that was accurate three months ago can become misleading after an infection, a medication change, a bereavement, or a shift in cognition and judgement. The operational challenge is to update risk controls quickly enough to prevent harm, while avoiding a default move toward restriction “just in case.” This guidance sits within Assessment, Review & Changing Needs and connects to Service Models & Care Pathways because different service models (homecare, residential, supported living) require different control measures and escalation routes.

Start with a simple principle: reviews must follow evidence

A dementia risk assessment should not be reviewed “because it is due” and then re-saved unchanged. It should be reviewed against evidence from the last period, including:

Incidents and near-misses (falls, wandering episodes, medication errors, self-neglect indicators)
Daily notes patterns (refusals, sleep disruption, appetite, hygiene, hydration, mood)
Safeguarding concerns, complaints, or family feedback
Health changes (UTIs, delirium, pain, new mobility issues, sensory loss)
Observations from staff who know the person’s baseline

Operationally, the fastest way to do this well is a “review bundle” template: a one-page summary of the last 4–8 weeks plus the updated controls and who is responsible for them.

Falls risk: update the control plan, not just the score

Falls reviews go wrong when they become a numeric exercise. A practical dementia falls review asks: what has changed about mobility, judgement, supervision, environment, and routines?

High-value controls to review

Environment: lighting, clutter, footwear, bed/chair height, trip hazards, signage and cues.
Routines: toileting prompts, hydration prompts, fatigue periods, timing of personal care.
Support: escorting rules, transfer method, use of equipment, staff competency.
Escalation: when to refer to falls team/OT/GP, how repeated falls are analysed.

Operational example 1: Falls cluster in a care home corridor

Context: A person with moderate dementia falls twice in one week, both times late evening when heading toward the bathroom. The existing risk assessment states “requires prompts” but offers little practical control detail.

Support approach: The service triggers an immediate falls review and adds a 72-hour observation window to identify patterns (time, location, footwear, lighting, toileting frequency, fatigue).

Day-to-day delivery detail: Staff introduce a consistent evening toileting prompt at set times, add a night-light route, and adjust staffing allocation so a named staff member checks in during the known risk window. The care plan specifies transfer support and what to do if the person refuses assistance (including re-approach timing and escalation to senior on duty).

How effectiveness is evidenced: Falls are tracked for four weeks with “near miss” logging. The manager completes a short thematic review at week two and week four, documenting what has reduced risk and what remains unresolved. Staff competency is confirmed via spot checks.

Wandering and missing episodes: separate “walking with purpose” from unmanaged risk

Dementia-related walking may be purposeful (seeking familiarity, reducing anxiety, meeting sensory needs). The review task is to enable safe movement while managing the risk of getting lost, exposure, traffic danger, or exploitation.

Controls to review in wandering risk

Baseline mapping: known routes, known triggers, “safer” destinations.
Support options: planned escorted walks, meaningful activity alternatives, reassurance routines.
Response plan: time thresholds, who is contacted, police protocol where relevant, recording requirements.
Least restrictive measures: signage, environmental design, engagement first; restrictions only when justified.

Operational example 2: Supported living night-time leaving and disorientation

Context: A person in supported living leaves the property at night after waking confused. No prior history exists; the risk assessment is outdated and assumes independent community access.

Support approach: The provider completes an urgent review focusing on time-of-day risk. Capacity is reviewed for night-time decisions, and a best-interests decision process is used where the person cannot weigh relevant risks.

Day-to-day delivery detail: Staff introduce a “wake and reassure” routine, improve environmental cues (clear toilet signage, gentle lighting), and schedule a calming pre-bed routine. The plan clarifies the escalation pathway: on-call contacted after defined triggers, and welfare checks documented. If any restrictive measure is proposed (e.g., monitored exits), it is time-limited, reviewed weekly, and recorded with rationale.

How effectiveness is evidenced: Episodes of night-time leaving reduce, sleep improves, and staff record ABC-style notes to confirm triggers are being addressed rather than simply blocked.

Self-neglect: treat it as an early-warning system, not a “choice” label

In dementia care, self-neglect can be subtle: reduced washing, wearing the same clothes, missed meals, unopened medication packs, spoiled food, unpaid bills, or unsafe heating use. Reviews should focus on what has changed in executive function and sequencing.

Controls to review for self-neglect risk

Prompting design: step-by-step prompts, visual cues, simplified choices.
Medication safety: MAR process, blister packs, administration support, missed-dose escalation.
Home safety: cooker and appliance safety, fire risks, hoarding, hygiene risks.
Safeguarding interface: when concerns become a safeguarding referral and who makes it.

Operational example 3: Homecare review after repeated missed meals and hygiene decline

Context: Homecare staff observe a gradual decline: missed meals, spoiled food, and reduced personal hygiene. Family reports the person is “fine” on the phone, and the current plan does not reflect the risks.

Support approach: The care coordinator triggers a review and gathers a two-week evidence pack: visit notes, meal prompts completed, refusal patterns, and photos of unsafe food storage where consent and policy allow (or written descriptions where not).

Day-to-day delivery detail: Staff introduce a consistent morning routine: hydration, toileting, breakfast, and a “one task at a time” personal care prompt. Shopping support is added with agreed lists, and a safe-food check becomes part of the visit. Missed-meal triggers are defined: if two consecutive meals are declined, the on-call is informed and the family is contacted per plan.

How effectiveness is evidenced: The provider tracks meal uptake and hygiene prompts; where capacity is in doubt, a formal capacity assessment is initiated for key decisions (nutrition, medication, home safety). The review outcome is recorded and re-signed with clear responsibilities.

Commissioner expectation: demonstrable risk management and escalation

Commissioner expectation: Commissioners expect providers to evidence active risk management: timely reviews after incidents, clear escalation pathways, and demonstrable controls that reduce harm without collapsing independence. They also expect risks to be reflected in staffing plans (skill mix, supervision) and to see that repeated incidents lead to learning rather than repetition.

Regulator / Inspector expectation: least restrictive practice and clear decision trails

Regulator / Inspector expectation (CQC): Inspectors will test whether restrictions are necessary, proportionate, time-limited and reviewed, and whether mental capacity and best-interests processes are used appropriately. They will also look for staff understanding: can staff explain the person’s risks, triggers, and what they do differently today compared with last month?

Governance: make risk reviews visible at service level

A defensible approach uses light-touch but consistent governance:

Risk review log showing due dates, triggers, and completion status.
Incident thematic reviews (falls, wandering, medication, safeguarding) with action tracking.
Quality checks on care plan accuracy versus real delivery (spot checks, observations, audits).
Board/leadership oversight via a dashboard: falls rates, missing episodes, safeguarding, PRN use, hospital use.

If you can show what changed, why it changed, and how you know it worked, your dementia risk assessment process becomes both safer for people and stronger for commissioning and inspection.