用于预测肥胖合并睡眠呼吸障碍患者肥胖低通气综合征的分步临床与复合风险评分模型
Stepwise clinical and composite risk score model for predicting obesity hypoventilation syndrome in patients with obesity and sleep-disordered breathing.
文献信息
| PMID | 42934019 |
|---|---|
| 原文 | 在 PubMed 查看原文 ↗ |
| 发表日期 | 2026 |
| 作者 | Navaponn Phuttasuatta |
| 作者单位 | Department of Otolaryngology Head and Neck Surgery, Faculty of Medicine, Prince of Songkla University, Songkhla, Hat Yai, 90110, Thailand. |
| 期刊 | Obesity research & clinical practice |
| SCI 分区 | Q3 |
| IF | 3 |
| 研究类型 | 临床研究 · 临床 |
| 所属专科 | 鼻科 |
中文摘要
目的： 识别与肥胖低通气综合征（OHS）相关的因素，并在肥胖合并睡眠呼吸障碍患者中开发及内部验证分步临床模型和复合风险评分。
方法： 本回顾性横断面研究纳入213例接受I型多导睡眠监测（PSG）和动脉血气（ABG）检测的肥胖患者（体重指数≥30 kg/m²）。OHS定义为日间PaCO₂≥45 mmHg。构建三级分步模型：第1步——临床因素（男性、Epworth嗜睡量表[ESS]评分≥16、清醒SpO₂<92%）；第2步——多导睡眠监测严重程度（呼吸暂停低通气指数≥60次/小时、最低SpO₂<70%或T90≥20%）；第3步——生化参数（血清碳酸氢盐[HCO₃-]≥27 mmol/L）。复合风险评分由多因素logistic回归得出，并采用受试者工作特征分析进行评估。
结果： OHS的独立预测因素包括男性、ESS≥16和血清HCO3-≥27 mmol/L。分步算法显示各层级特异性递增，当三步均为阳性时特异性达99%，阳性似然比（LR+）为18.10。复合风险评分（范围0-12）表现出良好的区分度，并将患者分为低、中、高风险类别（OHS患病率分别为6.2%、26.1%和61.9%）。高风险截断评分≥7分时，识别OHS的特异性为95%，阳性预测值为62%，LR+为5.88。
结论： 所提出的三步模型和复合风险评分提供了一种无创工具，用于识别肥胖患者中的OHS。该方法有助于有针对性地进行ABG检测和早期干预，尤其在资源有限的环境中。
英文摘要
OBJECTIVES: To identify factors associated with obesity hypoventilation syndrome (OHS) and to develop and internally validate a stepwise clinical model and composite risk score in patients with obesity and sleep-disordered breathing.
METHODS: This retrospective cross-sectional study included 213 patients with obesity (body mass index ≥30 kg/m²) who underwent type I polysomnography (PSG) and arterial blood gas (ABG) testing. OHS was defined as a daytime PaCO₂ ≥ 45 mmHg. A three-tier stepwise model was constructed: step 1-clinical factors (male sex, Epworth Sleepiness Scale [ESS] score ≥16, awake SpO₂ <92%); step 2-polysomnography severity (apnea-hypopnea index ≥60/h, minimum SpO₂ <70%, or T90 ≥ 20%); and step 3-biochemical parameter (serum bicarbonate [HCO₃-] ≥27 mmol/L). A composite risk score was derived from multivariate logistic regression and evaluated using receiver operating characteristic analysis.
RESULTS: Independent predictors of OHS includes male sex, ESS ≥ 16, and serum HCO3- ≥ 27 mmol/L. The stepwise algorithm demonstrated increasing specificity across tiers, reaching 99% specificity and a positive likelihood ratio (LR+) of 18.10 when all three steps were positive. The composite risk score (range: 0-12) exhibited good discrimination and stratified patients into low-, intermediate-, and high-risk categories (OHS prevalence rates: 6.2%, 26.1%, and 61.9%, respectively). A high-risk cutoff score ≥ 7 points yielded 95% specificity, 62% positive predictive value, and LR+ 5.88 for identifying OHS.
CONCLUSIONS: The proposed three-step model and composite risk score provide a noninvasive tool for identifying OHS in patients with obesity. This approach facilitates targeted ABG testing and early intervention, particularly in resource-limited settings.