中国实用儿科杂志

中国实用儿科杂志 ›› 2022, Vol. 37 ›› Issue (9): 701-707.DOI: 10.19538/j.ek2022090611

• 论著 • 上一篇    下一篇

重度阻塞性睡眠呼吸暂停患儿临床预测模型研究

  

  1. 国家儿童医学中心 首都医科大学附属北京儿童医院 a儿童耳鼻咽喉头颈外科疾病北京市重点实验室，b呼吸科，c儿童耳鼻咽喉头颈外科，北京  100045 吴云肖与唐瑜芬为本文共同第一作者
  • 出版日期:2022-09-06 发布日期:2022-09-27
  • 通讯作者: 许志飞，电子信箱：zhifeixu@aliyun.com
  • 基金资助:
    国家自然科学基金（82070092）；北京市自然科学基金（7212033）

Clinical prediction model of severe obstructive sleep apnea in children

  1. *Beijing Key Laboratory of Pediatric Otolaryngology，Head & Neck Surgery，Beijing Children's Hospital，Capital Medical University，National Center for Children's Health，Beijing  100045，China 
  • Online:2022-09-06 Published:2022-09-27

摘要： 目的　建立习惯性打鼾患儿中重度阻塞性睡眠呼吸暂停（OSA）的临床预测模型，为临床诊疗提供依据。方法　选择2019年1月至12月就诊于首都医科大学附属北京儿童医院睡眠中心的3～12岁习惯性打鼾患儿。所有患儿完成一般资料收集、OSA-18问卷、PSQ-SRBD量表及多导睡眠监测。应用决策树方法构建重度OSA患儿的临床预测模型。结果　共纳入受试患儿1441例，根据PSG结果，重度OSA 1152例，非重度OSA 289例。重度OSA组年龄、男性比例、体重指数（BMI）、颈围/身高比、腹围/臀围比、SRBD量表呼吸维度、其他维度及总分均高于非重度OSA组（P均<0.01）。OSA-18问卷各个维度得分及总分在两组患儿间比较，差异无统计学意义（P均>0.05）。基于决策树构建的重度OSA患儿预测模型，对非重度OSA患儿预测精确率为90%，召回率76%，F1得分82%，对重度OSA患儿的预测精确率32%，召回率58%，F1得分41%，整体准确率为73%。 结论　该研究构建的重度OSA患儿临床预测模型整体准确率73%，有一定的预测价值，能为临床排除重度OSA患儿提供一定的依据，指导临床决策，但仍需更多的临床资料进一步优化模型。

关键词: 阻塞性睡眠呼吸暂停, 儿童, 临床预测

Abstract: Objective　To establish a clinical prediction model of severe OSA in children with habitual snoring and to provide the evidence for clinical diagnosis and treatment. Methods　The children aged 3-12 years with habitual snoring who visited the sleep center of Beijing Children's Hospital affiliated to Capital Medical University from January to December 2019 were included. All children completed general data collection，OSA-18 questionnaire， PSQ-SRBD scale and polysomnography. The clinical prediction model of children with severe OSA was established based on decision tree method. Results　A total of 1441 children were included and they were divided into severe OSA group（1152 cases） and non-severe OSA group（289 cases） according to PSG results. The age， proportion of boys， BMI， neck circumference/height ratio， abdominal circumference/hip circumference ratio， SRBD breathing dimension， SRBD scale behavior dimension and total score in severe OSA group were higher than those in non-severe OSA group （all P < 0.05）. There was no significant difference in the scores of each dimension or the total score of OSA-18 questionnaire between the two groups （all P> 0.05）. The prediction model of severe OSA based on decision tree had a prediction accuracy of 90%，a recall rate of 76%， and a F1 score of 82% for non-severe OSA children，and a prediction accuracy of 32%， a recall rate of 58%， and a F1 score of 41% for severe OSA children，with an overall accuracy of 73%. Conclusion　The overall accuracy of the clinical prediction model for children with severe OSA constructed in this study is 73%，which has certain predictive value and can provide some evidence for clinical exclusion of children with severe OSA and guide clinical decision-making. However，more clinical data are still needed to further optimize the model.

Key words: obstructive sleep apnea syndrome, child, clinical prediction