肝性脑病风险预测模型建立与验证的系统评价

目的 系统评价肝性脑病风险预测模型建立与验证情况.方法 采用主题词和自由词结合,计算机检索PubMed、Embase、Cochrane、中国知网、万方医学网、维普等数据库建库至2020年5月发表的肝性脑病风险预测模型相关文献,由2名研究者独立阅读并筛选文献,采用临床预测模型的偏倚风险评估工具CHARMS清单对纳入文献进...

Development and Validation of Hepatic Encephalopathy Risk Prediction Models:A Systematic Review

Objective To systematically review the development and validation of hepatic encephalopathy risk prediction models.Methods The Cochrane Library,PubMed,Embase,CNKI,VIP,and Wan Fang Databases were electronically searched to collect studies on hepatic encephalopathy risk prediction model from inception to May 2020.Two reviewers independently screened and analysis the literature,evaluated the included studies according to the CHARMS checklist.Results A total of 11 studies were included,of which 6 were prospective studies.5 models were externally verified,while 4 models were not internally or externally verified.The methodological quality evaluation of the includ-ed studies had low bias.The most common predisposing and precipitating factors included in the model were the histo-ry of hepatic encephalopathy and the decline in liver function.Albumin was a common protective factor.All studies reported the area under receiver of characteristic curve(AUROC)0.68-0.93.Conclusions The included model has good predictive performance and can help medical staff to identify high risk groups of hepatic encephalopathy in early stage,but the extrapolation of the model's predictive effect has not been effectively evaluated,and further research is needed to verify the model externally in the future.