Search

Introduction Initiating treatment with continuous positive airway pressure (CPAP) traditionally relies on in-person visits with trained therapists to provide hands-on instruction regarding CPAP usage and mask fit. To overcome geographic barriers and reduce COVID-19 transmission, health systems increasingly rely on remote set-ups of mailed equipment. Despite a strong rationale for the mailed approach, relative effectiveness is unclear. Methods Our VA medical center shifted from in-person to mailed CPAP dispensation during the COVID-19 pandemic in March 2020. Using VA administrative and wireless CPAP usage data, we assembled a cohort of patients with newly diagnosed obstructive sleep apnea (OSA) who initiated CPAP for the first time from July 2019 to August 2020. Our primary outcome was mean nightly usage over the first 90 days. We compared patients with in-person vs. mailed CPAP dispensation using generalized linear models adjusted for age, gender, race, and Charlson Comorbidity Index. Among patients with >1 hour of overall usage, we compared secondary outcomes of leak, apnea hypopnea index (AHI), and obstructive/central apnea indices. Results We identified 693 patients with newly diagnosed OSA whose CPAP was provided in-person and 296 who had CPAP mailed. Nightly usage in the first 90 days was modest in both groups (in-person: 149.7, mailed: 152.9 min/night), and we did not detect a difference in adjusted models (+7.6 min/night, 95%CI -13.6–28.8). We also did not detect a difference in 95th percentile leak (-1.2 liter/minute, 95%CI -3.3-0.9). Device-detected AHI was relatively low overall (in-person: 3.2, mailed: 4.1 events/hour), but was greater in the mailout group (+1.0/hour, 95%CI 0.2–1.7). AHI differences appeared to be driven by obstructive (+0.5/hour, 95%CI 0.2–0.8) but not central events (-0.1, 95% CI -0.2–0.4). Risk of AHI>5 was comparable between groups (in-person: 17.3%, mailed: 19.0%, OR 1.2, 95%CI 0.8–1.7). Conclusion We were able to switch from an in-person to a mail-based system of CPAP initiation without a change in CPAP adherence or mask leak. While AHI was slightly greater in the mailed group, the clinical significance of this finding is unclear. Future work will need to evaluate the impact of remote CPAP dispensation on patient-centered outcomes. Support (if any) VA Health Services Research and Development, CDA 18–187