Abstract
BACKGROUND: Posttraumatic Stress Disorder (PTSD) is associated with increased health care utilization, medical morbidity, and tobacco and alcohol use. Consequently, screening for PTSD has become increasingly common in primary care clinics, especially in Veteran healthcare settings where trauma exposure among patients is common. OBJECTIVE: The objective of this study was to revise the Primary Care PTSD screen (PC-PTSD) to reflect the new Diagnostic and Statistical Manual of Mental Disorders (DSM-5) criteria for PTSD (PC-PTSD-5) and to examine both the diagnostic accuracy and the patient acceptability of the revised measure. DESIGN: We compared the PC-PTSD-5 results with those from a brief psychiatric interview for PTSD. Participants also rated screening preferences and acceptability of the PC-PTSD-5. PARTICIPANTS: A convenience sample of 398 Veterans participated in the study (response rate = 41 %). Most of the participants were male, in their 60s, and the majority identified as non-Hispanic White. MEASURES: The PC-PTSD-5 was used as the screening measure, a modified version of the PTSD module of the MINI-International Neuropsychiatric Interview was used to diagnose DSM-5 PTSD, and five brief survey items were used to assess acceptability and preferences. KEY RESULTS: The PC-PTSD-5 demonstrated excellent diagnostic accuracy (AUC = 0.941; 95 % C.I.: 0.912- 0.969). Whereas a cut score of 3 maximized sensitivity (κ[1]) = 0.93; SE = .041; 95 % C.I.: 0.849-1.00), a cut score of 4 maximized efficiency (κ[0.5] = 0.63; SE = 0.052; 95 % C.I.: 0.527-0.731), and a cut score of 5 maximized specificity (κ[0] = 0.70; SE = 0.077; 95 % C.I.: 0.550-0.853). Patients found the screen acceptable and indicated a preference for administration by their primary care providers as opposed to by other providers or via self-report. CONCLUSIONS: The PC-PTSD-5 demonstrated strong preliminary results for diagnostic accuracy, and was broadly acceptable to patients.