Diagnostic accuracy and psychometric performance of two self-report measures of posttraumatic stress disorder in older veterans

Abstract

The Primary Care PTSD Screen for DSM-5 (PC-PTSD-5) and PTSD Checklist for DSM-5 (PCL-5) have demonstrated high levels of diagnostic accuracy and strong psychometric properties across various samples and settings. However, the impact of age on these measures has been underinvestigated. This is problematic, as without accurate measurement, older adults with posttraumatic stress disorder (PTSD) may be vulnerable to misdiagnosis and inappropriate treatment planning. This study examined the diagnostic accuracy of these measures in older (≥ 65 years; n = 192) versus younger (< 65 years; n = 188) veterans to determine whether precision in identifying PTSD was equally strong between groups in a sample of veterans receiving primary care services in a U.S. Department of Veterans Affairs setting. Results indicated that, despite a lower PTSD prevalence among older veterans (9.3% vs. 25.5%), the PC-PTSD-5 performed better among older veterans, AUC = .960, 95% CI [.928, .992], than younger veterans, AUC = .897, 95% CI [.846, .949], z = 2.04, p = .042. A PC-PTSD-5 cutoff score of 4 was optimal for both groups. PCL-5 performance was excellent for both older, AUC = .925, 95% CI [.880, .970], and younger veterans, AUC = .894, 95% CI [.847, .942], z = 0.92, p = .358. Although different optimal cutoff scores were found for older (36) versus younger (34) veterans, these cutoffs were not significantly different from each other or the standard cutoff (33), McNemar tests ps = .125–1.00. Both measures maintain their robust psychometric properties in veterans across the lifespan.