Required CPAP usage time to normalize AHI in obstructive sleep apnea patients: a simulation study

Adherence to continuous positive airway pressure (CPAP) is often limited. The aim of the current work was to create a simulation tool to enable determination of the individual CPAP therapy time required to normalize apnea-hypopnea index (AHI) (<5 events h^-1) in a cohort of OSA patients.

Polygraphic studies of 1989 consecutive patients were analyzed. CPAP therapy was simulated in 5 min intervals starting from the beginning of the night and continuing until the end. In simulation, events inside the simulated CPAP therapy periods were considered to be prevented. The cutoff points where AHI reached a normal level of  <  5 events h^-1 were determined for mild, moderate and severe OSA categories.

The median values of the required simulated CPAP usage times to normalize the AHI values were 3.3 h, 5.6 h and 6.5 h in the mild, moderate and severe OSA categories, respectively. However, there were great differences between individuals in the CPAP usage times required to normalize AHI.

An arbitrary threshold for CPAP adherence (e.g. 4 h) leaves many OSA patients with a significant residual AHI, which could explain why some clinical trials fail to show significant benefits. Thresholds for adherence should be adjusted based on the patient-specific distribution of obstruction events during the night.