术中生理事件自动标记:理解其临床后果
本研究利用手术室黑匣子(ORBB)平台前瞻性标记术中低血压、低温和低氧事件,对2875例手术进行回归建模,发现低温和低氧是重返手术室的独立预测因素,低血压是住院时间和患者安全指标事件的独立预测因素,提示术中生理事件应纳入术后质量与安全分析。
要点
- 低温和低氧是重返手术室的独立预测因素(分别p<0.02和p<0.01)。
- 低血压是住院时间(p<0.03)和患者安全指标事件(p=0.03)的独立预测因素。
- 加入术中生理数据后回归模型性能提升。
核心结论
- 术中生理事件是术后并发症的重要考虑变量,应纳入质量与安全分析。
研究设计与样本
- 前瞻性应用ORBB标记系统,对2020年8月17日至2022年9月3日期间2875例手术进行监测,采用回归建模分析术中标记与不良结局的关联。
主要发现
- 2875例手术中567例(19%)被标记。
- 低温和低氧分别是重返手术室的独立预测因素(p<0.02,p<0.01)。
- 低血压是住院时间(p<0.03)和患者安全指标事件(p=0.03)的独立预测因素。
- 合并症、手术时长、BMI、年龄和ASA状态在预测模型中起重要作用。
临床怎么用
- 可利用术中生理事件标记识别术后并发症高风险患者,为术后护理决策提供支持。
局限性
- 本研究为回顾性建模,是术中预测分析的第一步,需进一步识别最影响结局的术中事件和患者因素。
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
INTRODUCTION: Hypotension, hypoxia, and hypothermia tissue have unclear effects on postoperative outcomes. The Operating Room Black Box™ (ORBB) is a novel monitoring platform that automates "tagging" of intraoperative hypotension, hypoxia, and hypothermia. We hypothesized that intraoperative physiologic event tags would serve as predictive variables of adverse postoperative outcomes. METHODS: Using the ORBB, tags were prospectively applied for moderate and severe episodes of hypotension, hypothermia, and hypoxia. Regression modeling was performed to examine the association between intraoperative tags and adverse outcomes of surgical site infection, patient safety indicator (PSI) events, length of stay, readmission, return to OR, and mortality. RESULTS: 2875 cases were performed and monitored with the ORBB tag system between August 17, 2020 and September 3, 2022. 567 cases (19%) were tagged. Regression models had improved performance after adding in intraoperative physiologic data. Hypothermia and hypoxia were independent predictors of increased return to the operating room (p < 0.02, p < 0.01, respectively). Hypotension was an independent predictor of length of stay (p < 0.03) and patient safety indicator events (p = 0.03). Comorbidities, length of case, BMI, age, and ASA status played significant roles in the predictive models. CONCLUSION: Intraoperative physiologic events are important variables to consider regarding postoperative complications and should be incorporated into quality and safety analysis. As real-time OR data monitoring becomes a reality, work must be done to identify the intraoperative events and patient factors that most affect outcomes. This retrospective modeling serves as the first step toward intraoperative predictive analytics that can provide decision support for postoperative care.
引用
- Bain AP, Succar B, Odeh J, et al. Automated tagging of intraoperative physiologic events: understanding the clinical consequences. Surg Endosc. 2025. PMID: 41275004. doi:10.1007/s00464-025-12409-6
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。