Logo image
Back
Functional, Contextual, and Interventional Drivers of Formal Coercion in Acute Mental Health Units: A Feature Analysis
Journal article   Open access   Peer reviewed

Functional, Contextual, and Interventional Drivers of Formal Coercion in Acute Mental Health Units: A Feature Analysis

Esario IV Daguman, Marie Hutchinson, Jacqui Yoxall and Richard Lakeman
International Journal of Mental Health Nursing, Vol.35(1), pp.1-17
24/02/2026
DOI: https://doi.org/10.1111/inm.70240
PMID: 41732105
Appears in  Recent Faculty of Health Publications
pdf
Functional, Contextual, and Interventional Drivers1.92 MBDownloadView
Published (Version of record)CC BY V4.0,  Open Access
url
Functional, Contextual, and Interventional DriversView
Published (Version of record)CC BY V4.0,  Open

Related links

Metrics

1 Record Views

Abstract

Boruta algorithm coercion mental health services predictors psychiatric nursing random forests
Minimising formal coercion, such as seclusion, physical restraint, and forced medication, remains a global priority in acute mental health units. However, key drivers beyond individual-level features are poorly understood. This exploratory analysis was intended to identify the top functional, contextual, and interventional features linked to formal coercion in three Australian acute adult mental health inpatient units. Nested within a mixed concurrent control study, this feature analysis examined nurses' reports of 2955 de-escalation events over 324days, from March 2024 to April 2025, including nurses' commentaries on the behavioural functions that triggered de-escalations. Fifteen inductively coded functional features were analysed alongside 15 contextual and 16 interventional features. Studied target variables included seclusion and physical restraint events and their durations, as-needed intramuscular psychotropic events, physical injury events, and Code Black activations. Features were analysed using bivariate statistics and machine learning techniques, including the Boruta algorithm for feature selection and random forest regressions for predictive modelling. Top drivers for the use of formal coercion included behavioural ‘Responses to Challenging, Physical and External Stimuli,’ incidents of self-harm, incidents directed towards nurses, and the application of specific deescalation techniques. A hierarchy of behavioural functions is proposed as a by-product of this analysis. These findings provide nuanced insights into the drivers of formal coercion and the underlying value arrangements, as well as elevate the merit of ecological, bottom-up approaches in early warning signs work.

Details

Logo image