Residential aged care psychiatry — structured clinical viva

Working problem: aggression during personal care in RAC dementia — likely multifactorial (pain, approach, environment, possible delirium) with sedation and falls from current PRN antipsychotic use. Standing olanzapine 10 mg is excessive as an opening move in a frail older adult. Prioritise assessment, non-drug redesign of care, pain protocol, and only then consider lowest-dose short-term drug if residual severe risk, with deprescribing plan.[1][2][5]

Assessment. Collateral for time course; medical screen (infection, constipation, retention, pain, hypoxia); medication reconciliation; ABC behaviour charting focused on shower triggers; risk (staff injury, resident harm, falls); capacity/proxy for treatment decisions. Exclude Lewy body features before any high-potency antipsychotic.[1]

Non-drug first. Person-centred personal care (two-person calm approach, warmth, privacy, preferred gender of carer, music, shorter showers). Staff training/psychosocial packages: Fossey reduced antipsychotic use without worsening behaviour; WHELD person-centred training and activities improve QoL/agitation metrics. Husebo pain protocol for behavioural disturbance in nursing-home dementia.[3][4][5]

Drug decision. Do not start olanzapine 10 mg standing. If residual severe danger after non-drug measures: document indication, consent/proxy, ECG as indicated, lowest dose (e.g. olanzapine 2.5 mg range if chosen, or alternative lower-risk local formulary agent), short duration, falls monitoring. Name Schneider death risk and Gill observational mortality. Avoid chemical restraint for convenience.[2][8]

Deprescribing. Build review into the plan. DART-AD: continuation not clearly better for many; long-term follow-up raised mortality concerns — trial reduction when stable with monitoring for relapse.[6][7]

Disposition. Optimise in facility with liaison follow-up; hospital if medical instability or unmanageable risk. Educate staff that evidence supports non-drug pathways reducing drug dependence.[1][3]