This study 1) describes the prevalence of positive PTSD screens among male, female, and OEF/OIF/OND veterans using various PTSD Checklist-Military version (PCL-M) criteria; 2) evaluates the sensitivity and specificity of various PCL-M criteria; and 3) identifies optimal screening criteria in predicting clinician-documented PTSD diagnoses. VA electronic medical records data from 327,093 veterans during 2008-2012 were analyzed. Receiver operating characteristic curve analyses compared PCL-M scores against clinician-documented PTSD diagnoses. Results showed that different PCL-M scoring criteria resulted in 62.0-84.5% of veterans screening positive for PTSD compared to 40.1% with a clinician-documented PTSD diagnosis. Sensitivity of the PCL-M ranged from 73.7 to 93.5% and specificity ranged from 21.6 to 45.8% for all veterans. The optimal PCL-M cut score according to Youden's index was ≥45 for male veterans, ≥35 for female veterans, and ≥38 for OEF/OIF/OND veterans. Self-report measures like the PCL-M may be a useful screening tool for identifying probable PTSD in VA specialty clinics, but they should be calibrated for different veteran subgroups and followed by structured clinical interviews.
Keywords: Assessment/diagnosis; Epidemiology; Health services; Posttraumatic stress disorder; Trauma.
Published by Elsevier Ireland Ltd.