What does this research mean for the field?

Data-driven personalization of positive airway pressure (PAP) comfort settings using causal machine learning is associated with improved therapy adherence compared to default device settings. Novelty: ClaimNovelty.INCREMENTAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

To evaluate whether causal machine learning-driven personalization of positive airway pressure comfort settings improves device adherence compared to standard default settings.

August 24, 2026Frontiers in SleepOpen Access

Personalized comfort settings in positive airway pressure machines using a causal inference approach: a retrospective observational analysis

View Full Paper

Ask AI

Bookmark

Authors

CNChinh Nguyen HJHimani Jayawardane HEHussein Emami

Discussion

Member takes

Overview

Retrospective observational analysis finds personalized comfort settings improve positive airway pressure adherence, indicating machine learning optimization enhances sleep therapy usage.

Key Points

To evaluate whether causal machine learning-driven personalization of positive airway pressure comfort settings improves device adherence compared to standard default settings.
Developed a causal forest model using AirSense 10 data and validated it across temporally separated AirSense 10 and AirSense 11 cohorts with 90-day and 1-year follow-ups.
Assessed CMS-defined adherence (device usage ≥ 4 hours per night on ≥ 70% of nights across 90 days) using backdoor adjustment methods and propensity score matching.
Personalized comfort settings produced an estimated average treatment effect increase of 2.9 percentage points in CMS-defined adherence (p < 0.001).
Propensity score-matched patients matching model recommendations demonstrated sustained usage improvements across age, gender, apnea-hypopnea index, and mask types without worsening mask leak or residual apnea-hypopnea index.

Cite This Study

Nguyen et al. (2026) studied this question.

synapsesocial.com/papers/6a8c005bbca056c88e6df00f https://doi.org/10.3389/frsle.2026.1883557

View Full Paper

Ask AI

Bookmark

August 24, 2026Frontiers in SleepOpen Access

Personalized comfort settings in positive airway pressure machines using a causal inference approach: a retrospective observational analysis

View Full Paper

Ask AI

Bookmark

Authors

CNChinh Nguyen HJHimani Jayawardane HEHussein Emami

Discussion

Member takes

Overview

Retrospective observational analysis finds personalized comfort settings improve positive airway pressure adherence, indicating machine learning optimization enhances sleep therapy usage.

Key Points

To evaluate whether causal machine learning-driven personalization of positive airway pressure comfort settings improves device adherence compared to standard default settings.
Developed a causal forest model using AirSense 10 data and validated it across temporally separated AirSense 10 and AirSense 11 cohorts with 90-day and 1-year follow-ups.
Assessed CMS-defined adherence (device usage ≥ 4 hours per night on ≥ 70% of nights across 90 days) using backdoor adjustment methods and propensity score matching.
Personalized comfort settings produced an estimated average treatment effect increase of 2.9 percentage points in CMS-defined adherence (p < 0.001).
Propensity score-matched patients matching model recommendations demonstrated sustained usage improvements across age, gender, apnea-hypopnea index, and mask types without worsening mask leak or residual apnea-hypopnea index.

Cite This Study

Nguyen et al. (2026) studied this question.

synapsesocial.com/papers/6a8c005bbca056c88e6df00f https://doi.org/10.3389/frsle.2026.1883557

View Full Paper

Ask AI

Bookmark