Formulas to predict continuous positive airway pressure level using a home auto-adjusting device for obstructive...

Area of Science:

• Sleep Medicine
• Respiratory Physiology
• Medical Device Technology

Background:

• Obstructive sleep apnea (OSA) affects millions globally, necessitating effective treatment like continuous positive airway pressure (CPAP).
• Accurate CPAP titration is crucial for treatment efficacy and patient adherence.
• Home-based auto-CPAP titration offers a convenient alternative to in-lab studies.

Purpose of the Study:

• To develop and validate predictive equations for therapeutic CPAP (CPAP_T).
• To assess the influence of interface type on CPAP_T prediction.
• To compare manual visual analysis (CPAP_V) and automated 95th percentile (CPAP_P95) methods for predicting CPAP_T.

Main Methods:

• Retrospective analysis of 702 adult OSA patients undergoing home-based autoCPAP titration.
• CPAP_T determined via manual visual analysis (CPAP_V) and automated 95th percentile pressure (CPAP_P95).
• Multiple linear regression and K-fold cross-validation used with AHI, neck circumference, BMI, and mask type as predictors.

Main Results:

• Predictive models for CPAP_V (M1) and CPAP_P95 (M2) were developed, with M1 showing higher accuracy (R²=0.19) than M2 (R²=0.09).
• CPAP_V demonstrated greater precision (±3.2 cm H₂O) compared to CPAP_P95 (±4 cm H₂O).
• M1 achieved 79% accuracy (±2 cm H₂O difference) in predicting CPAP_T, significantly higher than M2 (64%).

Conclusions:

• Predictive equations based on home-based CPAP titration, particularly using visual analysis (CPAP_V), can accurately estimate therapeutic CPAP.
• The visual analysis method (M1) offers a more precise and accurate approach for initial CPAP setting compared to the 95th percentile method (M2).
• These findings support the use of predictive models for optimizing home-based CPAP titration and improving patient management in OSA.