Factors associated with involuntary admissions: a register-based cross-sectional multicenter study

Despite the EU recommendations on mental health, involuntary admission has been under researched in Italy for a long time and the overall picture of involuntary admission still appears fragmentary. The aims of this study are to evaluate involuntary admission rates in the Piedmont Region (Italy) and to investigate clinical and service-related variables associated with involuntary admission.

This is a cross-sectional retrospective multicenter study involving all psychiatric inpatients units of the general hospitals of Piedmont Region. Data on hospitalizations during 2016 were collected by consulting hospital discharge registers. The analyses were performed on two samples: 6018 patients (data analysis was run on first hospitalization during the study period for those with multiple admissions) and 7881 inpatient episodes. The association between involuntary admission and socio-demographic and clinical characteristics was examined through t-test for continuous variables, and Pearson's Chi-square test for categorical variables. Multilevel modeling was applied in logistic regression models with two levels: for the first model center and participants and for the second model center and inpatient episodes.

Of 6018 inpatients, 10.1% were admitted involuntarily at first hospitalization, while the overall compulsory treatment rate was slightly lower (9.1%) in the inpatient episodes sample (n = 7881). The involuntary admission rates ranged from 0.8 to 21% among study centers. Involuntary admissions were primarily associated with younger age, diagnosis of schizophrenia or substance use disorders, longer duration of hospital stay, mechanical restraint episodes, and fewer subsequent hospitalizations during the study period.

The rate of involuntary admission in the Piedmont Region was lower than the mean rate across countries worldwide. There were noteworthy differences in rates of involuntary admission among psychiatric units, although no relationship was found with characteristics of the psychiatric wards or of the areas where hospitals are located.