Despite multiple ethical issues and little evidence of their efficacy, compulsory admission and treatment are still common psychiatric practice. Therefore, we aimed to assess potential differences in treatment and outcome between voluntarily and compulsorily admitted patients.

We extracted clinical data from inpatients treated in an academic hospital in Zurich, Switzerland between January 1, 2013 and December 31, 2019. Observation time started upon the first admission and ended after a one-year follow-up after the last discharge. Several sociodemographic and clinical characteristics, including Health of the Nation Outcome Scales (HoNOS) scores, were retrospectively obtained. We then identified risk factors of compulsory admission using logistic regression in order to perform a widely balanced propensity score matching. Altogether, we compared 4,570 compulsorily and 4,570 voluntarily admitted propensity score-matched patients. Multiple differences between these groups concerning received treatment, coercive measures, clinical parameters, and service use outcomes were detected.

Upon discharge, compulsorily admitted patients reached a similar HoNOS sum score in a significantly shorter duration of treatment. They were more often admitted for crisis interventions, were prescribed less pharmacologic treatment, and received fewer therapies. During the follow-up, voluntarily admitted patients were readmitted more often, while the time to readmission did not differ.

Under narrowly set circumstances, compulsory admissions might be helpful to avert and relieve exacerbations of severe psychiatric disorders.

Coercion in psychiatry is a controversial issue. Identifying its predictors and their interaction using traditional statistical methods is difficult, given the large number of variables involved. The purpose of this study was to use machine-learning (ML) models to identify socio-demographic, clinical and procedural characteristics that predict the use of compulsory admission on a large sample of psychiatric patients.

We retrospectively analyzed the routinely collected data of all psychiatric admissions that occurred between 2013 and 2017 in the canton of Vaud, Switzerland (N = 25,584). The main predictors of involuntary hospitalization were identified using two ML algorithms: Classification and Regression Tree (CART) and Random Forests (RFs). Their predictive power was compared with that obtained through traditional logistic regression. Sensitivity analyses were also performed and missing data were imputed through multiple imputation using chain equations.

The three models achieved similar predictive balanced accuracy, ranging between 68 and 72%. CART showed the lowest predictive power (68%) but the most parsimonious model, allowing to estimate the probability of being involuntarily admitted with only three checks: aggressive behaviors, who referred the patient to hospital and primary diagnosis. The results of CART and RFs on the imputed data were almost identical to those obtained on the original data, confirming the robustness of our models.

Identifying predictors of coercion is essential to efficiently target the development of professional training, preventive strategies and alternative interventions. ML methodologies could offer new effective tools to achieve this goal, providing accurate but simple models that could be used in clinical practice.

Involuntary admission (IA) for psychiatric treatment has a history of controversial discussions. We aimed to describe characteristics of a cohort of involuntarily compared to voluntarily admitted patients regarding clinical and socio-demographic characteristics before and after implementation of the new legislation.

In this observational cohort study, routine data of 15’125 patients who were admitted to the University Hospital of Psychiatry Zurich between 2008 and 2016 were analyzed using a series of generalized estimating equations.

At least one IA occurred in 4’560 patients (30.1%). Of the 31’508 admissions 8’843 (28.1%) were involuntary. In the final multivariable model, being a tourist (OR = 3.5) or an asylum seeker (OR = 2.3), having a schizophrenic disorder (OR = 2.1), or a bipolar disorder (OR = 1.8) contributed most to our model. Male gender, higher age, prescription of neuroleptics (all OR < 2.0) as well as having a depressive disorder, prescription of psychotherapy, prescription of antidepressants and admission after implementation of the new legislation (all OR > 0.6) were also weakly associated with IA.

Besides schizophrenic or bipolar disorders, a small group of patients had an increased risk for IA due to non-clinical parameters (i.e. tourists and asylum seekers). Knowledge about risk factors should be used for the development of multi-level strategies to prevent frequent (involuntary) hospitalizations in patients at risk. On the organizational level, we could show that the new legislation decreased the risk for IA, and therefore may have succeeded in strengthening patient autonomy.