Abstract
BACKGROUND: We continue to have limited success in identifying Veterans at high risk of suicide, due in part to reticence of disclosing suicidal ideation. AIMS: This study used a simulated groups experimental design to evaluate the MMPI-3's ability to assess suicide risk underreporting in Veterans with past-month death/suicide ideation. MATERIALS AND METHODS: Thirty-nine Veterans (53.85% men) were randomized to standard and simulated underreporting groups and provided valid data on the MMPI-3 and collateral measures. We examined (1) whether simulated underreporting on the MMPI-3 (indexed by L and K scale scores) impacts SUI scale scores, (2) if these effects generalize to underreporting on extratest suicide and non-suicide measures, and (3) if MMPI-3 L and K scales incrementally predict and differentiate between Veterans with recent death/suicide ideation who were instructed to answer honestly and those instructed to underreport. RESULTS: Groups scored significantly differently on K (g = 0.99: M(simulation) = 57.83, M(standard) = 43.72), but not L. Underreporting captured by K generalized to lower MMPI-3 SUI scale scores (g = 2.00; M(simulation) = 46.33, M(standard) = 66.81) and collateral measures of suicide risk (g = 0.69-0.79). K scores significantly predicted group membership and added incrementally to L. DISCUSSION: The limitations and clinical implications of these findinga are discussed. CONCLUSION: MMPI-3 K, but not L, scale scores most reliably capture defensive reporting of suicidal ideation and intent and psychopatholpgy, more boradly. However a signifcant amount of underreporrting of suicidal ideation and intent may go undetected. TRIAL REGISTRATION: ClinicalTrials.gov identifier: NCT04896151.