A Realist Evaluation of a Rapid Response System for Mental State Deterioration in Acute Hospital Settings

Patient mental state deterioration presents significant challenges in acute hospital settings, affecting outcomes, increasing reliance on restrictive interventions, and placing additional strain on healthcare staff. Despite its prevalence, consensus on best practice remains limited. The De-escalation, Intervention, Early Response Team (DIvERT) is a pilot rapid response system introduced to improve early identification, enable timely interventions, reduce crisis incidents, and support ward staff in caring for patients with mental state deterioration.

A realist evaluation approach was used to test, validate, and refine program theories explaining DIvERT's mechanisms. Data collection included a cross-sectional survey, semi-structured interviews, field observations, a medical record audit, and incident report analysis. Analysis was guided by the Context-Mechanism-Outcome framework to explain DIvERT's effective functioning in responding to patient deterioration.

DIvERT facilitated early intervention through multidisciplinary collaboration, though organisational factors such as staffing constraints, workload pressures, and inconsistent assessment practices influenced effectiveness. Key mechanisms included structured escalation pathways, clinical skills, staff training, and interprofessional collaboration. Challenges included limited after-hours availability, reflecting the constraints of a pilot initiative, underreporting of incidents, and hierarchical decision-making. While causation cannot be directly established, trends indicate DIvERT was associated with fewer Code Grey responses, particularly during initial episodes of mental state deterioration.

This realist evaluation highlights the value of structured escalation pathways, multidisciplinary collaboration, organisational support, and tailored training in managing mental state deterioration. Preliminary trends suggest DIvERT may enable more proactive and timely early intervention, whereas traditional reactive hospital emergency response for aggression (Code Grey) was more often linked to repeat incidents. Workforce constraints and inconsistent assessment limited effective functioning, underscoring the need for strengthened training, integration into workflows, and improved after-hours coverage to support scalability and long-term success.

This study demonstrates that a proactive rapid response model (DIvERT) can strengthen the recognition and management of mental state deterioration in acute hospital settings. By formalising escalation pathways, improving interdisciplinary collaboration, and tailoring training to staff needs, the model supports safer and more timely responses to patient deterioration. Embedding such approaches into organisational workflows has implications for patient safety, staff confidence, and system efficiency.

The study addressed the challenge of inconsistent and reactive responses to patient mental state deterioration, which are often reliant on crisis interventions such as Code Grey. DIvERT facilitated early intervention through structured escalation processes, improved interdisciplinary collaboration, and enhanced staff skills. However, organisational barriers such as workforce constraints, after-hours gaps, and inconsistent use of mental state deterioration assessment tools limited its effective functioning. Findings are directly relevant to clinicians in acute hospital settings (particularly nursing and allied health staff), hospital administrators responsible for patient safety and workforce planning, and policymakers overseeing standards for recognising and responding to acute deterioration. The results highlight where investment in training, structured escalation systems, and organisational support can reduce reliance on restrictive interventions and improve both patient and staff safety.

This evaluation adhered to the Realist And Meta-narrative Evidence Synthesis: Evolving Standards (RAMESES) II reporting standards for realist evaluations, as outlined in the EQUATOR Network guidelines.

This study did not include patient or public involvement in its design, conduct, or reporting.