The SCL-90-R as a screening instrument for severe personality disturbance among outpatients with mood and anxiety disorders

Our objective was to examine whether the SCL-90-R, a widely used self-report measure of distress and psychopathology, could screen for personality disorders in general and for severe personality disturbance (SPD) in particular, at the time when patients seek treatment for a state (mood or anxiety) disorder. The SCL-90-R was administered to 112 consecutive outpatients with various mood and anxiety disorders. The personality severity index (PSI) score, defined as the mean value of the scores on the SCL-90-R subscales of interpersonal sensitivity, hostility, and paranoid ideation, was compared with the current symptom index (CSI) score, defined as the mean value of the scores on the remaining six SCL-90-R subscales. A positive screen was considered if PSI > CSI. SPD was defined as the presence of any DSM-III-R Cluster A and/or Cluster B personality disorder. The accuracy of the screen was verified by means of the Structured Clinical Interview for DSM-III-R Personality Disorders. The sensitivity of the PSI > CSI criterion to screen for SPD was 89.4%, while its sensitivity to screen for any DSM-III-R personality disorder was 72.9%. The SCL-90-R subscales that contributed the most to the screening discriminability of this SCL-90-R-derived screening measure were hostility, paranoid ideation, somatization, and obsessive-compulsive behavior. The SCL-90-R may be used to screen for SPD in routine work with outpatients with mood and anxiety disorders, but the results of the screening need to be verified because of the possibility of false negatives and false positives, although that possibility is apparently low. These findings may have important prognostic and treatment implications.