A-232 Head-To-Head Comparison of the Rey Auditory Verbal Learning Test Effort Score and Forced Choice Embedded Performance Validity Indicators Among Patients with and without Verbal Memory Impairment

Abstract

Abstract Objective: This study directly compared the classification accuracies of the Rey Auditory Verbal Learning Test (RAVLT) Effort Score (ES) and Forced Choice (FC) embedded performance validity indicators among patients with and without verbal memory impairment. Methods: This study included data from 268 ethnoracially and diagnostically diverse (41% Black; 36% White; 16 % Hispanic; 5% Asian; 2% other), with mean age of 46.13 (SD=16.08) and mean education of 13.91 years (SD=2.80). Receiver operating characteristic (ROC) curve analyses assessed classification accuracy via areas under the curve (AUCs) and facilitated selection of optimal cut-scores and their associated sensitivities/specificities for the RAVLT ES and FC among the unimpaired verbal memory (≥37T) (n=135) and impaired memory (≤36T) (n=76) subsamples. Results: Among the overall sample, ES has acceptable classification accuracy (AUC=.724), with an optimal cutoff of ≤2 (26% sensitivity/90% specificity). FC also had acceptable accuracy (AUC = .764), with an optimal cut score ≤12 (47% sensitivity/90% specificity). For the unimpaired group, ES (AUC = .863) had an optimal cut score of ≤10 (67% sensitivity/90% specificity) and Forced Choice (AUC = .836) had an optimal cutoff of ≤14 (75% sensitivity/87% specificity). In the memory impaired group, ES no longer had significant classification accuracy (AUC = .477), whereas FC remained significant (AUC = .639), with an optimal cut score of c11 (30% sensitivity/90% specificity). Conclusion: When compared head-to-head, RAVLT FC was a superior performance validity measure for both memory impaired and unimpaired patients compared to the ES, which only worked well as a validity index among those without significant memory impairment.