Abstract
BACKGROUND: Insomnia is common after stroke and is associated with poorer recovery and greater risk of subsequent strokes. Yet, no insomnia measures have been validated in English-speaking individuals affected by stroke. AIMS: This prospective diagnostic validation study investigated the discriminatory validity and optimal diagnostic cut-off of the Sleep Condition Indicator when screening for Diagnostic and Statistical Manual of Mental Disorders-fifth edition (DSM-5) insomnia disorder post-stroke. METHODS: A convenience sample of 180 (60.0% women, mean age=49.61 ± 12.41 years) community-based, adult (≥18 years) self-reported stroke survivors completed an online questionnaire. Diagnosis of DSM-5 insomnia disorder was based on analysis of a detailed sleep history questionnaire. Statistical analyses explored discriminant validity, convergent validity, relationships with demographic and mood variables, and internal consistency. Receiver operating characteristic curves were plotted to assess diagnostic accuracy. RESULTS: Data from the sleep history questionnaire suggested that 75 participants (41.67%) met criteria for DSM-5 insomnia disorder, 33 (18.33%) exhibited symptoms of insomnia but did not meet diagnostic criteria, and 72 (40.0%) had no insomnia symptoms at the time of assessment. The Sleep Condition Indicator (SCI) demonstrated 'excellent' diagnostic accuracy in the detection of insomnia post-stroke, with an area under the curve of 0.86 (95% CI (0.81, 0.91)). The optimal cut-off was determined as being ≤13, yielding a sensitivity of 88.0% and a specificity of 71.43%. CONCLUSIONS: The findings of this study demonstrate the SCI to be a valid and reliable method with which to diagnose DSM-5 insomnia disorder and symptoms post-stroke. However, a lower threshold than is used in the general population may be necessary after stroke.