Clinical prediction of violence as a conditional judgment

Area of Science:

• Forensic Psychiatry
• Clinical Psychology
• Public Health

Background:

• Previous violence prediction research focused on institutional confinement.
• Community-based violence management requires detailed assessment of precipitating and inhibiting factors.
• Understanding community violence risk in mentally ill individuals is crucial.

Purpose of the Study:

• To examine a conditional prediction model for violence in mentally ill patients.
• To assess clinicians' predictions of violence-related factors in a community setting.
• To compare clinician predictions with actual violent incidents over a 6-month follow-up.

Main Methods:

• A conditional prediction model was used, where clinicians assessed factors associated with violence for individual patients.
• Data were collected from 712 patients in an urban psychiatric emergency room.
• Patients were followed for 6 months using interviews and official records.

Main Results:

• Clinicians demonstrated general accuracy in predicting the seriousness and location of violence.
• The study found an overestimation by clinicians regarding the role of medication compliance in violent incidents.
• Clinicians also overestimated the influence of drug use on violent behaviors.

Conclusions:

• Conditional prediction models offer insights into violence risk assessment in the community.
• Clinical judgment in predicting violence requires refinement, particularly concerning medication compliance and substance use.
• Further research is needed to improve the accuracy of predicting community violence in mentally ill populations.