Beyond "Do No Harm": Re-engineering Safety in Psychiatric Care through Quality Improvement

Background: Safety in psychiatric care involves distinctive risks, self-harm, violence/coercion, medication and diagnostic harms, and unsafe transitions, often intensified in resource-constrained systems. Objective: To synthesize what works, for whom, and under what conditions when quality improvement (QI) is used to reduce harm in psychiatric services. Methods: Narrative review with realist synthesis examining context–mechanism–outcome patterns across QI projects, time-series and before–after studies, implementation reports, and guidance. Data were charted for setting, safety domain, QI strategy, implementation supports, measures (outcome/process/balancing), and equity/rights considerations. Results: Effective programs paired a standardized core (e.g., de-escalation/Safewards, Six Core Strategies, post-discharge follow-up, antipsychotic metabolic bundles, red-flag diagnostic checklists) with locally adaptable workflows and run/SPC-based feedback. Success depended on addressed CFIR determinants (leadership, resources, workflow fit) and on psychological safety and supervision. In primary care/LMIC settings, mhGAP-aligned task-sharing and clear referral rules enabled scale without proportional budgets. Equity-stratified dashboards surfaced disparities. Economic signals favored portfolios combining de-escalation, safer transitions, and metabolic monitoring. Ethical governance (least-restrictive practice, data minimization, accountable review) underpinned trust. Conclusions: Psychiatric safety improves when QI discipline, Safety-II/HRO mindsets, and equity-aware measurement are embedded in everyday work, delivering less harm without new harms elsewhere and enabling sustainable spread across inpatient, community, and primary-care interfaces.