Sensitivity and specificity of a screening test to document traumatic experiences and to diagnose post-traumatic stress disorder in ARDS patients after intensive care treatment

Abstract

Objective: Many survivors of critical illness and intensive care unit (ICU) treatment have traumatic memories such as nightmares, panic or pain which can be associated with the development of posttraumatic stress disorder (PTSD). In order to simplify the rapid and early detection of PTSD in such patients, we modified an existing questionnaire for diagnosis of PTSD and validated the instrument in a cohort of ARDS patients after long-term ICU therapy. Design: Follow-up cohort study. Setting: The 20-bed ICU of a university teaching hospital. Patients: A cohort of 52 long-term survivors of the acute respiratory distress syndrome (ARDS). Interventions and measurements: The questionnaire was administered to the study cohort at two time points 2 years apart. At the second evaluation, the patients underwent a structured interview with two trained psychiatrists to diagnose PTSD according to Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) criteria. The reliability and validity of the questionnaire was then estimated and its specificity, sensitivity and optimal decision threshold determined using receiver operating characteristic (ROC) curve analyses. Results: The questionnaire showed a high internal consistency (Crohnbach's α = 0.93) and a high test-retest reliability (intraclass correlation coefficient α = 0.89). There was evidence of construct validity by a linear relationship between scores and the number of traumatic memories from the ICU the patients described (Spearman's ρ = 0.48, p < 0.01). Criterion validity was demonstrated by ROC curve analyses resulting in a sensitivity of 77.0% and a specificity of 97.5% for the diagnosis of PTSD. Conclusions: The questionnaire was found to be a responsive, valid and reliable instrument to screen survivors of intensive care for PTSD.