Assessment of obsessions and compulsions: Reliability, validity, and sensitivity to treatment effects

Advances in the treatment of obsessive-compulsive disorder (OCD) require reliable and valid measures of sufficient sensitivity to detect treatment effects. The present article critically reviews the instruments used in OCD treatment-outcome research. Behavioral methods, self-report inventories, and observer-rated scales are reviewed with respect to content, reliability, validity, and sensitivity to treatment effects. The latter was determined by meta-analyses of trials of behavior therapy (exposure plus response prevention) and clomipramine. Little is known about the psychometric properties of behavioral assessment methods, and they are used increasingly less often in outcome research, despite certain advantages. Self-report inventories tend to have acceptable reliability and validity, except for the SCL-90-R OC scale (and its predecessors) which has weak discriminant validity and appears to be essentially a measure of nonspecific distress. Little is known about the reliability and vatiduy of most observer-rated scales, despite the fact that they are popular in treatment outcome research. All measures appear sensitive to treatment effects, although observer-rated scales tend to yield larger effect sizes than self-report measures. For treatment outcome research, the Yale-Brown Obsessive Compulsive Scale (YBOCS) appears to be the best available instrument in terms of range of obsessive-compulsive features assessed, reliability, validity, and sensitivity to treatment effects. Computer-administered and self-report versions of the YBOCS have been developed, which appear promising but require further evaluation. The effects of treatment may be best understood by using measures of specific symptoms rather than relying on global measures of symptom severity. The YBOCS can be readily used for these purposes. The article concludes by considering additional requirements for a comprehensive assessment of obsessions and compulsions.