Abstract Introduction Digital Cognitive Behavioral Therapy for Insomnia (dCBT-I) is a patient-led, online, non-pharmacological intervention proven effective for insomnia. Digital programs are cost-effective and convenient, but not all patients respond equally well. Identifying who may not respond as well a priori could ensure timely connection to more appropriate treatment and effectively allocate limited behavioral sleep medicine resources. The present study analyzed differences in treatment response between groups of participants taking hypnotic medications with varying frequency at baseline: 0, between 1 and fewer than 4 times, or 4 or more times per week. Methods Data were drawn from 118 participants in the RCT of Stepped-Care Sleep Treatment in General Practice (RESTING Study) randomized to dCBT-I. For eligibility, participants had to be 50 years or older and meet DSM-5 criteria for insomnia disorder. The sample was divided into three groups based on the frequency of use of prescription sleep medications at baseline: taking no sleep medications (NO-MEDS; N=74), taking medication four or more times per week (HIGH-MED; N=28), and the rest (LOW-MED; N=16). Insomnia was measured via Insomnia Severity Index (ISI). Hypnotic medication use at baseline was measured via the patient self-report. We computed Cohen’s d effect sizes (ES) for change in ISI from baseline to 2-months in the three groups. Results Results showed a moderate effect size for NO-MED (baseline ISI mean=15.8, SD=4.2, 2-mo ISI mean=12.58, SD=5.06, d=.69), a very large effect size for LOW-MED (baseline ISI mean=17.3, SD=4.0, 2-mo ISI mean=12.8, SD=3.9, d=1.13), and a moderate effect size for individuals in the HIGH-MED group (baseline ISI mean=15.32, SD=4.61, 2-mo ISI mean=12.67, SD=5.46, d=.53). Conclusion In the present study, the highest effect size for reduction in ISI from baseline to 2-months was observed among participants taking sleep medication at least one but fewer than 4 times per week at baseline. Relevant to the interpretation of these results is the fact that this group had the highest mean ISI score at baseline. To further understand this result, future work could explore the impact of potential mediating factors, such as the spontaneous reduction is the use of prescription sleep medication. Support (if any) Support: R01AG057500
Osorno et al. (Fri,) studied this question.