background Standardised risk assessment is becoming increasingly important in forensic psychiatry. At the moment, however, the strengths and weaknesses of this procedure are largely unknown.
aim To study the predictive validity of standardised risk assessment in forensic psychiatric and psychopathological reports using the hkt-30 (historical-clinical-future) and to compare the results produced by this instrument with the predictive validity of clinical risk assessment.
method The hkt-30 was completed retrospectively for 123 randomly selected forensic assessment reports compiled between 1993 en 1995. The hkt-30 is a standardised risk assessment instrument for criminal recidivism, consisting of 30 items. The predictive validity was calculated by means of an roc-analysis (receiver operating characteristic). A distinction was made between severe and less severe recidivism. The clinical assessment was also subjected to an roc-analysis and tested against actual recidivism (severe or less severe).
results The predictive validity of the hkt-30 seems to be better for severe recidivism than for less severe recidivism. The difference, however, is not significant. The predictive validity of the total score on the hkt-30 with regard to severe recidivism is comparable to the predictive validity of the clinical assessment. Clinical assessment seems to predict less severe recidivism better than the hkt-30. Again, however, the difference is not significant.
conclusion When scores are awarded on the basis of existing forensic psychopathological assessment reports, the predictive validity of the hkt-30 is just as good as the clinical assessment in these reports. The main advantages of a standardised risk assessment instrument are that it provides more insight into the way in which the assessment is made and that high scores on particular items can highlight problem areas that are important in connection with future treatment.