Absolute and relative risk of mechanical restraint, forced medication, and seclusion during involuntary psychiatric hospitalisation: a population-wide cohort study

Mechanical restraint, forced medication, and seclusion are coercive measures that may be used during involuntary psychiatric hospitalisation. However, the extent of their usage and application across patient groups is unclear, limiting efforts to ensure equitable psychiatric care. Aim. To describe the absolute and relative risk of mechanical restraint, forced medication, and seclusion across sociodemographic and clinical patient characteristics. Methods. We conducted a population-wide cohort study of all involuntary psychiatric hospitalisations in Stockholm, 2012-2023 (n=49 873), using linked population-wide Swedish register data. We estimated crude absolute and relative risk adjusted for sex, age, and calendar year (aRR) with generalized estimating equations. Results. Out of all hospitalisations, 12 592 (25%) included at least one of these coercive measures. Mechanical restraint, forced medication, and seclusion were used during 6015 (12%), 9491 (19%), and 5562 (11%) of the hospitalisations, respectively. Across sociodemographic patient characteristics, the highest risk of being subject to any coercive measure was observed during hospitalisation among 18–24-year-olds (aRR 1.32 [1.22-1.43] vs. 65+ years) and patients born outside of Europe (1.20 [1.13-1.27] vs. Swedish-born). The risk was also elevated when patients were female (1.09 [1.04-1.10] vs. male), had low income (1.15 [1.06-1.25] vs. high income), primary education (1.10 [1.02-1.19] vs. higher education), or were unmarried (1.15 [1.08-1.25] vs. married). Clinically, the highest risk of any coercive measure was observed during hospitalisation for bipolar disorder (1.57 [1.42-1.73]), drug-induced psychosis (1.46 [1.31-1.62]), and for the category “other psychiatric disorders” meaning diagnoses not commonly motivating involuntary care, such as ADHD or delirium (1.48 [1.33-1.64]), compared with hospitalisation for alcohol use disorder. The risk was also elevated when patients had a lifetime history of psychiatric hospitalisation (1.26 [1.20-1.33]) or intellectual disability (1.22 [1.10-1.34]), but lower when hospitalisation was preceded by recent self-harm (0.69 [0.62-0.76]). The risk of mechanical restraint specifically was most pronounced during hospitalisation for personality disorder, while the risk of forced medication and seclusion was highest during hospitalisation for bipolar disorder and psychotic conditions. Conclusions. Coercive measures appear concentrated in identifiable patient subgroups, including socially vulnerable groups such as people born outside of Europe and those with intellectual disability. Targeted prevention strategies may help ensure equitable and safe delivery of involuntary psychiatric care. The analytic plan was preregistered at the Open Science Framework (https://osf.io/8rpu4).