Development and validation of a diagnostic model for obstructive sleep apnea.

STUDY

OBJECTIVES

Obstructive sleep apnea (OSA) is a prevalent, underdiagnosed disorder linked to serious health risks. This study developed a diagnostic model for OSA.

METHODS

Clinical data were analyzed from 3,038 adults who underwent polysomnography at the Second Affiliated Hospital of Xinxiang Medical University between 2015 and 2024. Sixteen candidate predictors were initially considered and a 70:30 split was used to divide participants into training and validation groups. A multivariate logistic regression model was developed to predict OSA risk. Model performance was assessed using the area under the receiver operating characteristic curve, calibration plots, Brier score, and decision curve analysis.

RESULTS

The model included four predictors of OSA (age, longest apnea time, lowest oxygen saturation, and oxygen desaturation index). The model demonstrated high accuracy with the area under the receiver operating characteristic curve values of 0.957 (95% CI: 0.945–0.969) in the training group and 0.945 (95% CI: 0.923–0.967) in the validation group. Brier scores were 0.071 and 0.081, with calibration plots showing excellent agreement between predicted and observed probabilities. Decision curve analysis confirmed clinical usefulness.

CONCLUSIONS

This study developed a highly accurate OSA diagnostic model to support clinical decision-making following polysomnography. Further validation in diverse populations is warranted to optimize its broad applicability.