文章摘要
心脏手术患者术后谵妄风险预测模型的系统评价
Systematic review of risk prediction models for postoperative delirium in cardiac surgery patients
  
DOI：10.12089/jca.2025.03.012
中文关键词: 心脏手术  术后谵妄  预测模型  系统评价
英文关键词: Cardiac surgery  Postoperative delirium  Predictive modeling  Systematic review
基金项目:
作者单位E-mail
郭畔旭 475004,河南省开封市,河南大学护理与健康学院（现在中山市人民医院手术麻醉科）  
阳晓娟 中山市人民医院心胸外科  
李晓婉 475004,河南省开封市,河南大学护理与健康学院  
徐文琪 475004,河南省开封市,河南大学护理与健康学院  
李思思 475004,河南省开封市,河南大学护理与健康学院  
安详 475004,河南省开封市,河南大学护理与健康学院  
王培席 475004,河南省开封市,河南大学护理与健康学院  
聂芳 中山市人民医院护理部 13435714797@163.com 
摘要点击次数: 215
全文下载次数: 60
中文摘要:
      
目的： 系统检索和评价心脏手术患者术后谵妄（POD）的风险预测模型，为临床选择合适的风险评估工具提供参考。
方法：计算机检索Embase、PubMed、Web of Science、Cochrane Library、维普期刊库、中国生物医学文献数据库、中国知网、万方数据库，检索时间为建库至2023年11月，纳入关于心脏手术患者POD的风险预测模型临床研究。研究人员依据纳入与排除标准筛选文献，按照偏倚风险评估工具独立进行数据提取和质量评价。
结果：共纳入12篇文献，涉及15个POD风险预测模型，建模样本量范围为130～57 180，受试者工作特征曲线下面积为0.680～0.932。纳入模型中最常见的预测因子为：年龄、射血分数、ICU住院时间、简易精神状态检查量表（MMSE）、手术时间、机械通气时间、BMI、出血量、术中输血、脑血管疾病史。其中9篇文献整体适用性较好，均具有偏倚风险，偏倚风险主要源于部分文献存在样本量较小、不恰当的变量选择方法、未报道缺失数据处理、缺乏内部验证或外部验证等因素。
结论： 心脏手术患者POD风险预测模型整体具有较好的适用性、校准性和区分度，但也存在显著的方法学缺陷和高偏倚风险。
英文摘要:
      
Objective: To systematically retrieve and evaluate risk of prediction models for postoperative delirium in patients undergoing cardiac surgery, and to provide references for clinical selection of appropriate risk assessment tools.
Methods: Computer searches were conducted in Embase, PubMed, Web of Science, Cochrane Library, VIP database, China biomedical literature database, CNKI, and Wanfang. The search period was from the establishment of the database to November 2023, and clinical studies on risk prediction models for postoperative delirium (POD) in patients undergoing cardiac surgery were included. Researchers independently screened the literature based on the inclusion and exclusion criteria and extracted data and evaluated the quality using the risk of bias assessment tool.
Results: Twelve studies involving 15 POD risk prediction models were included, with modeling sample sizes ranging from 130 to 57 180. The area under the receiver operating characteristic curve ranged from 0.680 to 0.932. The most common predictors included age, ejection fraction, ICU length of stay, minimental state examination (MMSE) score, duration of surgery, duration of mechanical ventilation, BMI, amount of bleeding, intraoperative transfusion, and history of cerebrovascular disease. Overall applicability was good in 9 studies, but all studies had a high risk of bias, primarily due to small sample sizes, inappropriate variable selection methods, unreported handling of missing data, and lack of internal or external validation.
Conclusion: Risk prediction models for postoperative delirium in cardiac surgery patients have overall good applicability, calibration, and discrimination, but significant methodological flaws and high risk of bias exist.
查看全文   查看/发表评论  下载PDF阅读器
关闭