Single-Item Measures for Detecting Sleep Problems in United States Military Veterans

Background As many as two-thirds of post-9/11 military veterans complain of sleep problems, including insomnia-like symptoms. Left untreated, chronic sleep problems increase the risk for a range of negative outcomes, including incident mental health disorders. However, sleep problems remain overlooked in primary care settings. To date, no brief sleep screeners have been developed or validated. Items assessing insomnia and poor sleep are often embedded into commonly used psychological assessments, and may serve as a viable first step in screening. Objective The objective of this study was to examine the utility of three single items (i.e., trouble falling asleep, awakening in the early morning, and sleep that is restless or disturbed) embedded into the Symptom Checklist-90-Revised (SCL) for identifying two outcomes of interest, poor sleep and probable insomnia. Design Data were drawn from the cross-sectional Post-Deployment Mental Health Study, hosted by the Mid-Atlantic VA Mental Illness Research, Education, and Clinical Center. Item performance was evaluated using sensitivity, specificity, and predictive value calculations, along with receiver operating characteristic (ROC) curves. Participants Post-9/11 U.S. military veterans with one or more overseas deployments and with no current DSM Axis I mental health disorder ( N = 1118). Main Measures An in-person health and sleep questionnaire, including the Pittsburgh Sleep Quality Index (PSQI) and the Symptom Checklist (SCL). Key Results Using an item response of 1, all three items demonstrated moderate sensitivity (0.70–0.78) and acceptable rates of false positives and false negatives (0.23–0.48 and 0.11–0.42, respectively) in predicting both outcomes, poor sleep and probable insomnia. Conclusions Our initial findings suggest that existing items in the SCL may serve as a first step in screening for sleep problems. Early detection and treatment of sleep problems might prevent or ameliorate several negative outcomes, including incident mental health disorders.