Suicide Risk and Absconding in Psychiatric Hospitals with and without Open Door Policies: A 15-year Naturalistic Observational Study

Background: Inpatient suicides and absconding of inpatients at risk of self-endangering behavior constitute important challenges for all medical disciplines. Patients at risk are often admitted to locked wards in psychiatric hospitals to prevent absconding, suicide attempts, and suicides. However, there is currently insufficient evidence that treatment on locked wards can effectively prevent these outcomes. Methods: In this 15-year naturalistic observational study, 349,574 cases treated in 21 German psychiatric inpatient hospitals from 1998 to 2012 were examined. Propensity score matching was used to select 145,738 cases for an analysis allowing causal inference on the effect of ward type (i.e., locked, partly locked, open, and day clinic wards) and hospital type (i.e., hospitals with and without locked wards) on suicide, suicide attempts, and absconding, despite the lack of an experimental design. Generalized linear mixed-effects models were used to analyze the data. Findings: We show that suicide, suicide attempts, and absconding are not increased in hospitals with an open door policy. Treatment on open wards is associated with a decreased probability of suicide attempts (OR = 0⋅66, p = 0⋅003), absconding with return (OR = 0·63, p < 0⋅001), and absconding without return (OR = 0⋅71, p = 0⋅01) compared to treatment on locked wards. Interpretation: Locked doors may not be successful in the prevention of suicide and absconding. Funding: No funding except for income from the primary employer was received.