Short- and Long-Term Recidivism Prediction of the PCL-R and the Effects of Age: A 24-Year Follow-Up

• Published in: Personality disorders 2015-01, Vol.6 (1), p.97-105
• Main Authors: Olver, Mark E., Wong, Stephen C. P.
• Format: Article
• Language: English
• Subjects: Adult; Age Differences; Age Factors; Antisocial Behavior; Antisocial Personality Disorder - diagnosis; Antisocial Personality Disorder - epidemiology; Criminal Offenders; Criminals - statistics & numerical data; Follow-Up Studies; Human; Humans; Male; Prediction; Prisoners - statistics & numerical data; Prognosis; Psychiatric Status Rating Scales - standards; Psychiatric Status Rating Scales - statistics & numerical data; Psychopathy; Recidivism; Risk Assessment; Time Factors; Violence - statistics & numerical data
• ISSN: 1949-2715; 1949-2723
• DOI: 10.1037/per0000095

We prospectively examined the short- and long-term prediction of several recidivism outcomes as a function of psychopathy and age in a sample of 273 Canadian federal inmates with an average 24 years postrelease follow-up. Offenders were rated using the original 22-item Hare Psychopathy Checklist (PCL: Hare, 1980) based on extensive archival file information, and the ratings were used to compute the Psychopathy Checklist-Revised (Hare, 2003) and the 4 facet scores. PCL-R total scores and the Lifestyle and Antisocial facets, but not the Interpersonal and Affective facets, showed mostly small and some moderate predictive efficacy for general and nonviolent recidivism over 3-, 5-, 10-, and 20-year fixed follow-ups, and predicted violence recidivism at shorter follow-ups. Age at release was negatively correlated with all recidivism outcomes and follow-up periods for both high and low PCL-R rated offenders, and uniquely predicted all recidivism outcomes after controlling for the PCL-R using Cox regression survival analysis. Increased age was consistently linked to recidivism reduction even for psychopathic offenders. The results showed that both PCL-R scores and age contributed to the prediction of recidivism; however, the PCL-R facets made differential contributions that varied with the type of offense (violent vs. nonviolent) and follow-up time (shorter vs. longer). The results have implications for both risk assessment using the PCL-R and potentially for risk reduction interventions.