Accurate assessment of sleep-wake patterns is important for sleep researchers and clinicians. Actigraphs are low-cost, non-intrusive, wrist-worn activity detectors used to estimate sleep-wake patterns in a natural environment for several nights. Although actigraphy shows good sensitivity (sleep detection), it has consistently demonstrated poor specificity (wakefulness detection while lying in bed relatively motionless). Because insomnia is characterized by wakefulness in bed, actigraphy may not be a valid objective measure of wakefulness for this group. It is possible that refinement of actigraphy software settings for sleep/wake algorithms might improve specificity. The current studies investigated this hypothesis by comparing wake parameters from 48 combinations of actigraphy settings to determine which sleep/wake algorithms best inform insomnia diagnosis and treatment. In the first study, none of the 48 actigraphy setting combinations consistently discriminated between adults with insomnia (n = 69) and non-insomnia (n = 80) on all three wake parameters, and no setting clearly discriminated between groups for the composite variable, total wake time. Similarly, in the second study, no setting combinations consistently discriminated between adults treated for insomnia (n = 18) and controls with untreated insomnia (n = 26) on all three wake parameters. Although two setting combinations discriminated between groups for the composite …

Suicide is a significant public health issue in the US. Despite national and international prioritization since 1996, little definitive progress has been made in terms of identification and intervention in cases of elevated suicide risk. Forty percent of those who died by suicide attended an emergency department within a year of death. Therefore, universal suicide risk screening in emergency departments could prove a vital component to a national suicide prevention strategy. The present study empirically evaluated the universal suicide risk screening program recently implemented at Parkland Health and Hospital System. The sample consisted of patients over 18 years of age (N=333,855; Mage=42.7, 32% male) screened as part of routine clinical care from May 4th, 2015, through November 3rd, 2015. The Parkland Algorithm for Suicide Screening (PASS) is part of a clinical decision support system for responses to Columbia - Suicide Severity Rating Scale Clinical Practice Screener (C-SSRS) items, leading to an automated clinical response via three suicide risk stratification levels: no action for no risk identified, psychiatric social worker assessment for moderate risk identified, and psychiatrist/psychologist interview for high risk identified. The present study used receiver operating characteristic (ROC) curve analysis, which found the PASS predicted disposition (z=30.46, p<.001, AUC=.78, …