Exploring the impact of a multilevel intervention focused on reducing the practices of seclusion and restraint in acute mental health units in an Australian mental health service.

CORPORATE culture; MEDICAL protocols; MENTAL health services; CONTROL (Psychology); INTERPROFESSIONAL relations; T-test (Statistics); PATIENT safety; INTERVIEWING; LEADERSHIP; SECLUSION of psychiatric hospital patients; RESTRAINT of patients; JUDGMENT sampling; DESCRIPTIVE statistics; THEMATIC analysis; PHYSICIAN practice patterns; RESEARCH; COMMUNICATION; PROFESSIONAL employee training; QUALITY assurance; HOSPITAL health promotion programs; HEALTH facilities; HEALTH care teams

Reducing and eliminating seclusion and restraint in inpatient settings has been a key area of focus in mental health policy and research for many years. To address this issue, numerous programmes aimed at minimising the use of these practices have been developed over the past two decades, with varying degrees of success. This article reports on research focused on the implementation of a localised, multilevel complex intervention that targeted both organisational and individual factors related to the use of seclusion and restraint. The researchers followed the impact of the intervention by interviewing medical, nursing and allied health staff who worked within the service (N = 12) and analysing the rates of seclusion and restraint over an 18‐month period. Post‐adoption, participants identified that there were clear changes in practice culture. Seclusion clearly became a practice of last resort and other options became prominent in staff's practice. Participants identified that there was a sense of shared purpose across the multidisciplinary team. The clinical environment was viewed as being more therapeutic for service users and less frightening for staff. There was a significant difference in the total number of seclusion events between pre‐ (Mean = 6.22, SD = 5.82) and post‐implementation (Mean = 2.55, SD = 2.44, p = 0.002, d = 0.94), demonstrating a significantly lower number of seclusions was observed after the intervention. Similarly, a significant difference in restraint events between pre‐ (Mean = 5.50, SD = 3.77) and post‐implementations (Mean = 3.38, SD = 3.21, p = 0.037, d = 0.62) was observed. [ABSTRACT FROM AUTHOR]