The Inventory of Problems-29 (IOP-29): A Systematic Review and Bivariate Diagnostic Test Accuracy Meta-Analysis

The Inventory of Problems-29 (IOP-29) is a 29-item self-administered symptom validity test (SVT) that assesses the credibility of clinical presentations related to posttraumatic stress disorder, depression/anxiety, psychosis, cognitive impairment and combination thereof. To date, no publications have summarized the classification accuracy of the IOP-29 using a bivariate meta-analytical approach that preserves the two-dimensional nature of the estimators. Our objective was to conduct a systematic review and bivariate diagnostic test accuracy meta-analysis of the IOP-29 according to the relevant guidelines. Twenty-one independent samples were included, with a total sample size of 4,163 participants. The results indicated that the IOP-29 is able to discriminate adequately between instructed simulators and healthy controls/clinical patients. Using the recommended cutoff (False Disorder Probability Score [FDS], ≥.50), a sensitivity of 82% was achieved, maintaining specificity at 93% (false positive rate of 7%). The language of the test and the type of comparison group have been identified as possible sources of heterogeneity. Specificity decreases for the non-English version of the IOP-29, for the FDS ≥ .30, and also decreases for studies using clinical controls, for all three cutoff scores. In general, our findings support the usefulness of the IOP-29 as an SVT; however, most of the included studies use a simulation design and have been coauthored by the test authors. Likewise, about half of the studies did not include bona fide patient controls but only nonclinical controls. The results obtained are highly promising, but further research, especially that using the criterion group paradigm, is recommended. Public Significance Statement Following a bivariate meta-analytic model, we have evaluated the discriminative ability of the IOP-29, a novel symptom validity test. The results obtained indicate that it is able to discriminate adequately between instructed simulators and healthy controls/clinical patients.