Raising the bar: Preliminary investigation of alternative MSVT cutoffs for adults referred for ADHD.

Abstract

This study investigated the classification accuracy of alternative cutoffs for three Medical Symptom Validity Test (MSVT) indices, Immediate Recall (IR), Delayed Recall (DR), and Consistency (CNS). Study 1 included 83 college students randomized into two simulated ADHD groups (Incentivized simulator n = 25; Non-incentivized simulator n = 28) and one control group (n = 30). Study 2 included cross-sectional data from 114 ADHD referrals at a university-based clinic (n = 91 valid group; n = 23 invalid group) who completed the MSVT and at least two additional performance validity tests. Receiver Operator Characteristic (ROC) curves in study 1 revealed significant areas under the curve for IR, DR, and CNS (AUC = .81 - .86), high to perfect specificity (.93 - .10), and moderate to high sensitivity (.45 - .75) using cutoffs of ≤85, ≤90, and ≤95. ROC curve analyses in study 2 yielded significant areas under the curve for IR (AUC = .73), DR (AUC = .74), and CNS (AUC = .78). An optimal cut score of ≤90 across all three subtests yielded modest sensitivity (.35 - .52) while still maintaining adequate specificity (.91 - .96). The manual-recommended cutoff for invalidity (IR, DR, or CNS ≤85) demonstrated high specificity (95%) but only modest sensitivity (35%) whereas the modified cutoff (IR, DR, or CNS ≤90) increased sensitivity substantially (56%) while maintaining adequate specificity (90%). This study provides preliminary support for alternative MSVT cutoffs in adult ADHD assessments. However, additional research is warranted before utilizing these alternative cutoffs in clinical contexts.