Evaluating the gold standard: a review and meta-analysis of the Structured Interview of Reported Symptoms

The Structured Interview of Reported Symptoms (SIRS; Rogers, Bagby, & Dickens, 1992) is often touted as the gold standard of measures of feigning. This label likely arises in part out of the impressive accuracy rates reported in the extensive validation research that preceded its publication. However, since its publication, researchers not only have continued to investigate the measure's utility but have expanded the study of the SIRS to include novel populations, different study methodologies, and abbreviated versions. The current review examines 26 studies using the SIRS to identify feigning, evaluating both its effectiveness at differentiating feigners from genuine responders and the potential impact of moderating variables. Meta-analyses revealed that research published since the initial validation studies demonstrate higher sensitivity but lower specificity rates than those reported in the SIRS manual. Studies in which feigners were composed of simulators yielded higher classification rates than studies sampling actual suspected malingerers. Furthermore, genuine patient samples were significantly more likely than nonclinical samples to be misclassified as feigning. Abbreviated versions of the SIRS also demonstrated equivalent accuracy with the standard measure. The implications of these findings for clinical practice are discussed.