INTRODUCING TEMPUS NEXT: AI-ENABLED CARE PATHWAY INTELLIGENCE /// EXPLORE NOW INTRODUCING TEMPUS NEXT: AI-ENABLED CARE PATHWAY INTELLIGENCE ///
06/26/2023

Association Between Mobile-Based Patient-Reported Sleep Questionnaires and Passive Wearable Sleep Sensors in Psychiatry Patients

AANP 2023 PRESENTATION
Authors Joseph Stanton, Mykel Robble, Candice Blacknall, Marcus Badgeley

Purpose: Historically, tracking behavioral health symptoms has been challenging. Mobile health apps for patient-reported outcomes (PRO) can administer measurement-based care to track patients’ symptom severity and be combined with Apple Health kit wearables. These tools can provide unique insights for real-time tracking of patient sleep quality.

Methods: This is a retrospective study of de-identified records of patients treated for psychiatric disease from 2021 to 2022 who received a pharmacogenomic assay and used the PRO app. We compared patient-reported sleep scores from standard psychiatric assessments to device-sensed sleep. Sleep from the 2 weeks prior to an assessment was included when at least 7 nights were
measured, and summarized to average duration and variance. Symptom severity was evaluated with full assessment scores (DSMV-C, PHQ-9, ISI, GAD-7, ADHD). The associations between sensed sleep and patient-reported sleep and cardiac sensor data were tested using Spearman’s and Pearson correlations, respectively.

Results: A subset of 1,503 patients that used PRO between January 2021 and November 2022 were analyzed. 115 completed assessments reviewing the prior 2 weeks’ sleep and sensed sleep was measured by a device most nights. Self-reported sleep difficulty on assessments was negatively associated with device-sensed average duration of sleep (Spearman rho=-0.3, p=5e-10), and positively with the variation in sleep (Spearman rho=0.2, p=5e-7). Across all assessments (scores normalized from 0 to 1), worse symptom severity was associated with less sleep (strongest association for DSMV-C). More device-sensed sleep was associated with higher heart rate variability and lower resting heart rate (p=6e-8 or lower).

Implications: Sensed sleep captured through a wearable device is significantly correlated to patients’ sleep quality and psychiatric symptoms on self-reported clinical assessments, as well as expected correlations between sensed sleep and cardiac activity that reflect the nervous system’s sympathetic and parasympathetic tone (fight or flight).

VIEW THE POSTER