文献速递小儿耳鼻咽喉睡眠呼吸障碍

识别扁桃体切除术后可安全同日出院的肥胖伴重度OSA高危儿童

Identifying High-Risk Children Safe for Same-Day Discharge After Tonsillectomy.
PubMed 文献 · Laryngoscope · 2026-05-14

该回顾性队列研究纳入304例2-18岁肥胖且重度OSA的扁桃体切除术患儿,发现仅11.8%发生严重围手术期事件。基于AHI<25次/小时且SpO2最低值>85%的简单临床规则可识别极低风险患儿,其阴性预测值为99.5%,支持个体化出院决策。

要点

  • 多数肥胖伴重度OSA患儿扁桃体切除术后无严重围手术期事件。
  • SpO2最低值是严重事件最强预测因子,其次为III级肥胖和AHI。
  • AHI<25次/小时且SpO2最低值>85%的规则可识别极低风险患儿。

核心结论

  • 肥胖伴重度OSA的多数患儿扁桃体切除术后不发生严重围手术期事件。
  • 简单多导睡眠图标准可识别极低风险患儿,可能适合同日出院计划。
  • 研究结果支持更个体化的术后住院决策,但需前瞻性验证。

研究设计与样本

  • 回顾性队列研究,纳入2021-2024年某三级儿童医院2-18岁肥胖且重度OSA的扁桃体切除术患儿304例。
  • 肥胖定义为BMI≥第95百分位,重度OSA定义为AHI≥10次/小时和/或SpO2最低值<80%。
  • 采用信息性先验的贝叶斯逻辑回归识别预测因子,评估模型风险分层及简化临床规则的敏感性、特异性和阴性预测值。

主要发现

  • 304例患儿中36例(11.8%)发生严重围手术期事件(ICU入住、住院延长>48小时或需高级呼吸支持)。
  • SpO2最低值是最强预测因子(β=-0.668,95%可信区间[-1.164, -0.172],方向概率99.6%)。
  • III级肥胖和AHI的预测方向概率分别为97.0%和95.0%。
  • 模型分层识别出33%的患者严重事件预测概率<5%。
  • 临床规则(AHI<25次/小时且SpO2最低值>85%)敏感性85.7%,阴性预测值99.5%,对应观察事件率<5%。

关键数值与推荐

  • 重度围手术期事件定义为ICU入住、住院延长>48小时或需高级呼吸支持。
  • 肥胖定义为BMI≥第95百分位。
  • 重度OSA定义为AHI≥10次/小时和/或SpO2最低值<80%。
  • 简化临床规则:AHI<25次/小时且SpO2最低值>85%。

临床怎么用

  • 对于肥胖伴重度OSA的扁桃体切除术患儿,可考虑使用AHI和SpO2最低值进行风险分层。
  • 符合AHI<25次/小时且SpO2最低值>85%的患儿严重围手术期事件风险极低,或可考虑同日出院。
  • 实施前需进行前瞻性验证。

局限性

  • 回顾性研究设计。
  • 单中心三级儿童医院数据。
  • 研究结果需前瞻性验证后方可实施。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

OBJECTIVE: Current guidelines recommend overnight admission for children with severe obstructive sleep apnea (OSA) and obesity undergoing tonsillectomy, although most have uneventful postoperative courses. We aimed to identify low-risk subgroups within this high-risk population who may be candidates for same-day discharge. METHODS: Retrospective cohort study of children aged 2-18 years with obesity (≥ 95th BMI percentile) and severe OSA (AHI ≥ 10 events/h and/or SpO2 nadir < 80%) undergoing tonsillectomy at a tertiary children's hospital (2021-2024). The primary outcome was a severe perioperative event: ICU admission, prolonged hospitalization (> 48 h), or need for advanced respiratory support. Bayesian logistic regression using informative priors identified predictors. Model-based risk stratification and simplified clinical rules were evaluated using sensitivity, specificity, and negative predictive value. RESULTS: Among 304 children, 36 (11.8%) experienced a severe event. SpO2 nadir was the strongest predictor (β = -0.668, 95% CrI [-1.164, -0.172], directional probability 99.6%), followed by Class III obesity (97.0%) and AHI (95.0%). Model-derived stratification identified 33% of patients with < 5% predicted probability of severe events. A clinical rule combining AHI < 25 events/h and SpO2 nadir > 85% showed sensitivity 85.7% and negative predictive value 99.5%, corresponding to an observed event rate < 5%. CONCLUSION: Most obese children with severe OSA do not experience severe perioperative events. Simple polysomnographic criteria can identify very low-risk patients who may be candidates for same-day discharge planning. These findings support a more individualized approach to postoperative admission; prospective validation is needed before implementation.

来源:Laryngoscope,查看原文。版权归原作者所有。

引用

  1. Ho A, Kou YF, Liu C, et al. Identifying High-Risk Children Safe for Same-Day Discharge After Tonsillectomy. Laryngoscope. 2026. PMID: 42135615. doi:10.1002/lary.70629

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。