Findings from two studies of an experimental precursor of Pear Therapeutics Inc’s Somryst (called SHUTi) that evaluated effectiveness to improve sleep outcomes and analyzed user journey data to predict intervention dropout were presented at Virtual SLEEP 2020. Somryst is the first FDA-authorized prescription digital therapeutic for the treatment of chronic insomnia. It is a 9-week mobile app program that is available by prescription only.

Findings from a large-scale randomized controlled trial, Web-delivered CBT for Insomnia Intervention Improves Sleep Among Adults with Insomnia and Depressive Symptoms, show that Somryst can produce lasting improvements in sleep outcomes among adults with insomnia and elevated depressive symptoms. These findings further support previous research that validated the utility of sleep diaries as a measure of treatment response. The study evaluated 1,149 Australian adults aged 18-64 years with insomnia and depressive symptoms for the effectiveness of SHUTi versus an attention-matched control with improving sleep outcomes measured by prospectively entered sleep diary data.

Sleep diary data indicate that treatment with the prescription digital therapeutic produced lasting improvements in sleep-onset latency, wake after sleep onset (WASO), sleep efficiency, number of awakenings, sleep quality, and total sleep time among adults with insomnia and elevated depressive symptoms. These findings show not only statistically significant improvements in diary-derived variables but also clinically meaningful change, with both sleep-onset latency and WASO falling below the 30-minute criterion in the treatment arm and total sleep time improving over time.

The digital nature of Somryst also offers new opportunities to probe the user experience at a granular level. A second study, Analyzing User Journey Data in Digital Health: Predicting Dropout from a Digital CBT-I Intervention, found that prediction of individual user dropout was possible early in the intervention, which may eventually help clinicians identify users at risk for dropping out and support continuing engagement with treatment. For this study, a collaboration with lead author Vincent Bremer at Leuphana University in Germany, data from a published randomized controlled trial evaluating effectiveness of Somryst were evaluated at each of six steps for a treatment arm of 151 participants to predict whether a user completed the full course of treatment.

“Research shows that Somryst use correlates with lasting improvements in sleep outcomes at 12 and 18 months for people suffering from chronic insomnia, and prospectively collected sleep diaries over time will allow us to refine therapeutic content and interfaces to support continuing engagement with treatment,” says Yuri Maricich, MD, chief medical officer at Pear Therapeutics, in a release. “These studies provide additional evidence that [prescription digital therapeutics] can play a vital role in the treatment of people with chronic insomnia and break down the barriers to access guideline recommended first line treatment of care.”

Study Results for 0524 Web-delivered CBT for Insomnia Intervention Improves Sleep Among Adults with Insomnia and Depressive Symptoms:

  • prescription digital therapeutic participants demonstrated greater reductions from baseline to end of treatment compared with control for sleep onset latency (LS mean difference [95% CI]=-22.3 min [-29.2, -15.3]; p<.0001), wake after sleep onset (-17.8 min [-23.4, -12.3]; p<.0001), and number of awakenings (-0.38 [-0.68, -0.09]; p=.0113).
  • prescription digital therapeutic participants also showed greater improvements in sleep efficiency (9.18% [7.25%, 11.10%]; p<.0001) and sleep quality (0.41 [0.30, 0.53]; p<.0001) from baseline to end of treatment compared with control.
  • Total sleep time was not significantly different between groups at end of treatment or 6-month follow-up, although it improved over baseline at 12 (18.73 min [7.39, 30.07]; p=.0013) and 18 months (23.76 min [9.15, 38.36]; p=.0015) relative to control.
  • All other significant sleep treatment effects were maintained in the treatment arm at 6, 12, and 18-month follow-up.

Study Results for 1204 Analyzing User Journey Data in Digital Health: Predicting Dropout from a Digital CBT-I Intervention:

  • Factors influencing dropout included aspects like length of time to complete a treatment core, difference between awake time and arising time at baseline, and number of email support exchanges.
  • Accuracy of predicting dropout varied depending on point in time of prediction and the machine learning technique. After model evaluation, a decision tree (yes/no dropout) achieved Area Under Curve values ranging between 0.6-0.9 across six steps over the intervention (how-to-use tutorial through treatment Cores).
  • Additional features (based on clinician input) contributed to prediction performance, including time to complete certain steps of the intervention, number of email exchanges, and time to get out of bed (difference between awake and arise times).