A CPAP data-based algorithm for automatic early prediction of therapy adherence.

Area of Science:

• Sleep Medicine
• Respiratory Medicine
• Medical Device Technology

Background:

• Continuous positive airway pressure (CPAP) therapy is crucial for obstructive sleep apnea (OSA) management.
• High rates of CPAP non-adherence (up to 40%) lead to treatment failure and poor outcomes.
• Early identification of patients at risk of non-adherence is essential for effective OSA management.

Purpose of the Study:

• To evaluate the predictive accuracy of the Philips Adherence Profiler™ (AP) algorithm for CPAP adherence at 90 days (D90).
• To assess the predictive value of early CPAP usage patterns (at D14) for long-term adherence.
• To determine if telemonitoring algorithms can identify patients likely to discontinue CPAP therapy.

Main Methods:

• Retrospective analysis of 457 OSA patients from the Pays de la Loire Sleep Cohort.
• Utilized CPAP machine data from the EncoreAnywhere™ database to run the AP™ algorithm.
• Assessed CPAP adherence (≥ 4 hours/night) at D14 and D90, with 88% adherence at D90.

Main Results:

• Older age and AP™ prediction at D14 were significant predictors of CPAP adherence at D90 in multivariate analysis.
• The AP™ algorithm at D14 demonstrated a strong predictive value for adherence at D90 (OR 16.99).
• CPAP adherence was not significantly associated with device-derived residual events or pressure levels, except for persistent significant leakage.

Conclusions:

• Automatic telemonitoring algorithms, like the Philips AP™, are valuable for early prediction of CPAP therapy adherence.
• These algorithms enable focused therapeutic follow-up on patients identified as high-risk for non-adherence.
• Early prediction facilitates timely interventions to improve long-term CPAP compliance in OSA patients.