肠内营养病人喂养不耐受风险预测模型的系统评价  

A systematic review of risk prediction models for feeding intolerance in patients receiving enteral nutrition

作　　者：陈小颉 段霞 郑微艳[1] 陶丽 CHEN Xiao-jie;DUAN Xia;ZHENG Wei-yan;TAO Li(Nursing Department,Renji Hospital Affiliated to Shanghai Jiaotong University School of Medicine,Shanghai 200127,China;Tongji University School of Medicine,Shanghai 200092,China;Nursing Department,the First Maternity and Infant Hospital Affiliated to Tongji University,Shanghai 200126,China)

机构地区：[1]上海交通大学医学院附属仁济医院护理部,上海200127 [2]同济大学医学院,上海200092 [3]同济大学附属第一妇婴保健院护理部,上海200126

出　　处：《肠外与肠内营养》2024年第2期107-113,共7页Parenteral & Enteral Nutrition

摘　　要：目的:系统分析了肠内营养病人喂养不耐受风险预测模型的研究现况,为医务工作者选择、应用及校正模型,或构建相关预测模型提供参考依据。方法:计算机检索中国知网、维普、万方、中国生物医学文献数据库(CBM)、Cochrane Library、PubMed、Embase、Web of Science、CINAHL中发表的关于肠内营养病人喂养不耐受风险预测模型的文献,并限定检索时间为建库至2023年2月28日。由2名研究者独立查阅文献、提取相关信息并评价纳入研究的偏倚性和适用性。结果:共纳入10个研究,包括14个模型。纳入模型的受试者工作曲线下面积为0.70~0.889。纳入模型的前3位预测因子依次为年龄、机械通气和白蛋白水平,其中白蛋白水平为保护性因素。结论:肠内营养病人喂养不耐受的发生与病人高龄、行机械通气治疗、白蛋白水平较低有关。现有风险预测模型的偏倚风险较高,未来应选择合适的机器学习算法,开展大样本、多中心研究,以构建具有普适性的FI风险预测模型,实施针对性预防措施以降低FI的发生风险。Objective:To systematically review the current status of research on risk prediction models for feeding intolerance(FI) in patients receiving enteral nutrition(EN),and to provide a reference for medical workers to select,apply,and calibrate models,or to construct related prediction models.Methods:A literature search was conducted in the China National Knowledge Infrastructure(CNKI),VIP,WanFang,Chinese Biomedical Literature Database(CBM),Cochrane Library,PubMed,Embase,Web of Science,and CINAHL databases published on risk prediction models for FI in patients receiving EN.The search time was limited from the database establishment to February 28,2023.Two researchers independently reviewed the literature,extracted relevant information,and evaluated the bias and applicability of the included studies.Results:A total of 10 studies were included,involving 14 models.The area under the receiver operating characteristic curve(AUC) of the included models ranged from 0.70 to 0.889.The top three predictors in the included models were age,mechanical ventilation,and albumin level,with albumin level being a protective factor.Conclusion:The occurrence of FI in patients receiving EN is related to advanced age,mechanical ventilation,and low albumin level.The existing risk prediction models have a high risk of bias.In the future,appropriate machine learning algorithms should be selected,and large-sample,multicenter studies should be conducted to construct FI risk prediction models with universal applicability.Targeted preventive measures should be implemented to reduce the risk of FI.

关 键 词：喂养不耐受 风险预测 模型 系统评价 

分 类 号：R459.3[医药卫生—治疗学]