Maximum Security Psychiatric Patients

Ratings of the dangerousness of 360 men in a maximum security psychiatric institution were obtained from experienced forensic clinicians. These ratings were well predicted (R = .61) from historical, offense, and in-hospital data. Seriousness of outcome scores based upon an 11-year minimum follow-up time were obtained on 85 additional men released from the same institution; this outcome sample was constructed so that the entire range of outcomes, from good to the commission of very serious crimes against persons, was represented. The seriousness of outcome scores were well predicted from the same variable sets used to predict the clinicians' ratings (R = .55); however, the regression model based upon the clinicians' ratings of dangerousness did not predict the outcome scores. The lack of relation between the clinical judgment model and outcome was maintained when the data were analyzed separately according to whether the patients were personality disordered or psychotic. The most important variables associated with dangerous outcome were seriousness of the index offense, an economic index offense, criminal as opposed to civil commitment, young age, and number of previous correctional institutionalizations.