An Empirical Study of Questionnaires for the Diagnosis of Pediatric Obstructive Sleep Apnea

Area of Science:

• Pediatric Sleep Medicine
• Medical Informatics
• Data Mining

Background:

• Pediatric Obstructive Sleep Apnea (OSA) is a growing health concern.
• Current diagnostic methods like polysomnography are costly and invasive.
• There is a need for accessible and accurate screening tools for pediatric OSA.

Purpose of the Study:

• To evaluate the diagnostic and stratification efficacy of five common questionnaires for pediatric OSA.
• To identify the most informative questions for diagnosing pediatric OSA using data mining.
• To develop and validate a novel questionnaire for improved pediatric OSA screening.

Main Methods:

• Analysis of responses from five established pediatric OSA questionnaires.
• Application of data mining techniques to identify significant diagnostic questions.
• Development of a new questionnaire based on identified informative questions.
• Training and evaluation of machine learning models using the new questionnaire data.

Main Results:

• Existing questionnaires demonstrated insufficient diagnostic accuracy for widespread clinical use.
• Data mining identified key questions that are highly informative for OSA diagnosis.
• Machine learning models trained on the new questionnaire showed enhanced accuracy in stratifying pediatric OSA.

Conclusions:

• Current questionnaires are inadequate for reliable pediatric OSA screening.
• A novel, data-driven questionnaire combined with machine learning offers a more accurate approach to pediatric OSA stratification.
• This approach could lead to more cost-effective and accessible screening for pediatric OSA.