Prevalence and risk factors for seclusion and restraint: results of a retrospective analysis.

Background: The reduction of coercion is a central challenge of inpatient mental health services. Little is known about the use of seclusion and restraint in open-door settings. This work aims to analyse the prevalence and risk factors of coercive measures in open inpatient wards of a Swiss university psychiatric hospital.

Methods: The use of seclusion and restraint was retrospectively analysed between 2017 and 2019. Incidence rate ratios using multivariable Poisson regressions were calculated to determine patient- and hospital stay-related risk factors for coercion.

Results: Of 1764 patients, 293 (16.6%) underwent at least one coercive measure (mostly seclusion) during their stay. The experience of coercive measures was associated with the following factors: male gender (IRR 1.34 [1.26, 1.38], p < 0.001), younger age (IRR 0.97 [0.97, 0.97], p < 0.001), being single or divorced (IRR 1.49 [1.43, 1.60], p < 0.001), diagnoses of organic (IRR 13.43 [13.38, 13.48], p < 0.001), psychotic (IRR 1.79 [1.79, 1.79], p < 0.001), or bipolar (IRR 2.17 [2.16, 2.17], p < 0.001) disorders, receiving disability benefits (IRR 1.92 [1.90, 2.07], p < 0.001), higher number of previous psychiatric hospitalizations (IRR 1.03 [1.03, 1.03], p < 0.001), and higher Health of the Nation Outcome Scales (HoNOS) item 1-scores (overactive, aggressive, or agitated behaviour) at admission 1.27 [1.27, 1.27], p < 0.001). Referrals from the emergency department (ED) (IRR 0.78 [0.77, 0.86], p < 0.001) were associated with a lower risk of coercion.

Conclusions: Young age, being single or divorced, male gender, diagnoses of organic, bipolar, or psychotic disorders, being on disability benefits, and previous psychiatric hospitalizations were identified as risk factors of coercion. The incidence of coercion was lower in ED referrals. Item 1 HoNOS scores at admission might be sufficient for a meaningful assessment of coercion risk. Our results suggest a strong link between aggressive behaviour and the use of coercive measures. Identification of patients at risk should be integrated to coercion reduction programs. The role of institutional and cultural factors influencing the use of coercion should be further investigated.