Use of historical remote monitoring data to determine predictors of CPAP non-compliance in patients with OSA

Purpose: Continuous positive airway pressure (CPAP) is the gold standard treatment for obstructive sleep apnoea. This study aimed to use complete usage data collected remotely from modern CPAP devices to identify compliance trends and clinical predictors of CPAP usage. Methods: Group usage data were analysed for a large cohort at a single tertiary sleep-centre before a detailed review of a 90-day reporting window for each patient was conducted. Individual data were collected for a smaller cohort of patients including demographics, past medical history and diagnostic sleep study results. A zero-inflated negative binomial regression model was used to determine associations between patient characteristics and usage days. Results: Of 6450 patients who were prescribed CPAP and included in the initial service analysis, 476 patients were included in the sub-group. Complete usage data revealed that 46% of patients were fully compliant with CPAP therapy. Compliance fell from 55 to 46% by day 90 and remained at this rate going forward. Significant predictors of CPAP non-compliance included being in the lowest quartile of Index of Multiple Deprivation scores (most deprived) compared with the highest quartile (least deprived) (p =.005), and less severe oxygen desaturation index (ODI) on diagnosis (p =.03). Conclusions: Complete usage data show that compliance at day 90 appears to be a good indicator of future CPAP usage. Predictors of CPAP non-compliance may include lower socioeconomic status, and lower ODI.