0772 Clusters of Upper Airway Stimulation Adherence Patterns in the First 90 Days. (25th May 2022)

• Record Type: Journal Article
• Authors: Soose, Ryan; Araujo, Mattheus; Faber, Kevin; Roy, Asim; Lee, Kent; Ni, Quan; Srivastava, Jaideep; Strollo, Patrick
• Is Part Of: Sleep. Volume 45(2022)Supplement 1
• Page Start: A336
• Page End: A336
• Publication Date: 2022-05-25
• Subjects: Sleep -- Physiological aspects -- Periodicals; Sleep disorders -- Periodicals; Sommeil -- Aspect physiologique -- Périodiques; Sommeil, Troubles du -- Périodiques; Sleep disorders; Sleep -- Physiological aspects; Sleep -- physiological aspects; Sleep Wake Disorders; Psychophysiology; Electronic journals; Periodicals; 616.8498
• DOI: 10.1093/sleep/zsac079.768
• Languages: English
• ISSNs: 0161-8105

Abstract

Abstract: Introduction: Upper airway stimulation (UAS) therapy is effective for a subset of obstructive sleep apnea (OSA) patients with CPAP intolerance. While overall adherence is high, some patients have suboptimal adherence to UAS, which limits effectiveness. Our goal was to identify UAS therapy usage patterns during the first three months of therapy that affect adherence. Methods: We retrieved anonymized UAS therapy usage data from 2, 091 individuals stored in a cloud-based monitoring system during the first three months after device activation. We aggregated adherence data including mean and standard deviation (SD) of nightly hours of use, therapy pauses, hours from midnight when the therapy was turned ON and OFF, and percentage of missing days. We computed the difference of the stimulation amplitude between the first and last day. We performed cluster analysis with Gaussian mixture models and computed the centroids of each cluster highlighting their main differences. Results: We identified six distinct clusters of UAS usage patterns. Clusters 1A (34% of the total cohort) and 1B (23%) had excellent therapy usage with 7.23h and 7.14h on days of use, respectively; with 1B distinguished by increased night-to-night variability. Clusters 2A (16%) and 2B (12%) had good mean therapy use of 6.63h and 6.21h, respectively, but their usage patterns were distinguished by a higher percentage of missing days (8% missing days in 2A and 23% in 2B) and less favorable therapy timing with an average therapy ON time after midnight. Clusters 3A (8%) and 3B (7%) were characterized by the lowest nightly use at 6.16h and 5.50h, respectively, and the highest night-to-night variability. 3A was further distinguished by the highest percentage of missing days (34%) while 3B was characterized by the frequent therapy pauses (mean 4.1 pauses per night) and the least increase in stimulation amplitude across the first 90 days. Conclusion: Cluster analysis of UAS usage patterns identified six distinct groups that may enable custom interventions for improved long-term management. Differentiation of these groups may have clinical implications on conditions (e.g. therapy discomfort, comorbid insomnia, poor sleep hygiene) that impact adherence. Support (If Any):