Post-incident responses (debriefing) after restrictive interventions in mental health inpatient settings: A realist synthesis

Background: Restrictive interventions (restraint, seclusion and rapid tranquilisation) are used widely in adult mental health inpatient settings despite national and international priorities to reduce them. Post-incident responses (sometimes referred to as debriefing), the immediate support and subsequent reflection and learning that should follow restrictive interventions, are recommended by the National Institute for Health and Care Excellence and embedded in restraint reduction frameworks, but their required components and mechanisms of effect remain poorly specified.

Aim: To explain how post-incident responses work, for whom and in what circumstances to reduce the harm and future use of restrictive interventions in mental health inpatient settings.

Methods: A realist synthesis was conducted across three iterative phases. Initial programme theories were developed from core literature and consultation with a patient and carer expert stakeholder group. A comprehensive database search (CINAHL, Embase, MEDLINE, PsycINFO, Web of Science) identified 993 records after deduplication, of which 35 met realist inclusion criteria. Policies were obtained from 46 NHS mental health trusts in England via Freedom of Information requests. Two expert stakeholder groups (patients and carers, n = 8; staff, n = 5) contributed throughout theory development and refinement. Semi-structured realist interviews were conducted with patients (n = 5), carers (n = 3) and staff (n = 7). Data were coded to context-mechanism-outcome configurations in NVivo and iteratively synthesised into programme theories, validated at two workshops at the Restraint Reduction Network conference.

Results: Four programme theories explain how post-incident responses work. Timing and readiness describes how emotional and physiological regulation determine whether individuals can engage; readiness for support differs from readiness for learning and reflection. Relational repair explains how empathy, fairness and acknowledgement rebuild trust following an incident. Mutual learning and accountability describes how shared sense-making and visible follow-through generate change. Team culture and psychological safety explains how team-level openness both enables post-incident responses and is strengthened by them. The four theories operate as a two-stage process, with post-incident support establishing the conditions for post-incident learning and reflection, and team culture functioning recursively across both stages. Contextual conditions including staffing, workload, acuity and leadership consistency shape whether mechanisms can operate.

Conclusions: Post-incident responses are not discrete procedural events but staged, relational processes whose effectiveness depends on patient readiness, staff relational competence and organisational context. Policies mandating fixed timing without regard for emotional readiness risk undermining the mechanisms they are intended to activate. Intervention development should address organisational conditions alongside individual practice.

Study registration: PROSPERO CRD42024579433, registered 18 November 2024.