摘要
目的评价ESRS对脑梗死和TIA患者再发“脑梗死”和“联合血管事件”的预测效果。方法系统检索CNKI、VIP、万方、PubMed、Web of Science、Embase数据库,搜集ESRS的外部验证研究,搜索时限为2010年1月至2021年1月,研究者按纳排标准独立筛选文献、提取资料并评价偏倚风险。用随机效应模型的meta分析汇总C统计量结果,再通过系统综述的方式探讨异质性来源。结果“脑梗死”结局分析中,纳入8篇文献,meta分析结果显示:ESRS有一定预测性能(C统计量=0.62,95%CI 0.59~0.69),研究间存在较大异质性(PI 0.49~0.78);此结局纳入的文献中,提到患者接受二级预防治疗的有2篇文献,预测效果较差(C统计量均低于0.60);纳入TIA和轻型脑梗死人群的有2篇文献,有较好的预测效果(C统计量分别为0.69和0.70)。“联合血管事件”结局分析中,共纳入8篇文献,meta分析结果显示:ESRS具有一定预测能力(C统计量=0.60,95%CI 0.58~0.66),研究间存在异质性(PI 0.56~0.70);纳入文献中,4篇文献的患者接受二级预防治疗,但预测结果均较差(C统计量在0.55~0.59),对TIA和轻型脑梗死人群的预测中,仅1篇文献取得较好的预测效果(C统计量=0.65),其它文献报道的预测效果不佳(C统计量在0.56~0.61)。结论现有证据表明,ESRS对于“脑梗死”、“联合血管事件”的结局均有一定的预测能力,但总体区分度仅在0.60~0.62之间,接受二级预防治疗的人群、TIA和轻型脑梗死人群可能影响其预测效果,ESRS的预测能力有待更多高质量研究的验证。
Objective To systematically evaluate the predictive performance of ESRS on the outcomes of recurrent“ischemic stroke”and“cardiovascular events”.Methods A systematic search was conducted on CNKI,VIP,PubMed,Web of Science,Embase,to collect external validation studies of ESRS.The search time was from 2010.1 to 2021.1.Then two researchers independently screened and evaluated the included literatures.The discrimination of the prediction models was measured by C-statistic in random effects meta-analysis.And,examining performance in systematic review.Results 8 studies with“ischemic stroke”as the outcome were included.Meta-analysis results showed that ESRS had rather poor discrimination(C-statistic=0.62,95%CI 0.59~0.69),and there was heterogeneity among studies(PI 0.49~0.78).In the analysis of heterogeneous sources,there were 2 literatures in which patients received secondary preventive treatment,and the prediction effect was poor(C statistics were lower than 0.60);There were 2 literatures included in TIA and mild cerebral infarction population,which had good prediction effect(C statistics were 0.69 and 0.70,respectively).There were 8 studies with“cardiovascular events”as outcome.Meta-analysis results showed that ESRS had helpful predictive performance(C-statistic=0.60,95%CI 0.58~0.66),and there was heterogeneity among studies(PI 0.56~0.70).In the analysis of heterogeneous sources,the patients in 4 literatures received secondary preventive treatment,but the prediction results were poor(C statistic was 0.55~0.59).In the prediction of TIA and mild cerebral infarction population,1 literature achieved good prediction effect(C statistic was 0.65),while the prediction effect reported in other studies was poor(C statistic was 0.56~0.61).Conclusions Existing evidence shows that ESRS has a certain ability to predict the outcome of“ischemic stroke”and“cardiovascular events”,but the overall discrimination is only between 0.60 and 0.62,and the prediction effect is not satisfactory.People receiving secondary preventive treatment,TIA and mild cerebral infarction may affect the predictive effect of ESRS.The prediction ability of ESRS needs to be verified by more high-quality studies.
作者
张淑君
李礼
王舒
ZHANG Shujun;LI li;WANG Shu(Department of Acupuncture,First Teaching Hospital of Tianjin University of Traditional Chinese Medicine,Tianjin 300193,China)
出处
《中风与神经疾病杂志》
CAS
2022年第7期624-630,共7页
Journal of Apoplexy and Nervous Diseases
基金
天津市科技计划项目(18PTLCSY00040,20ZYJDSY00020)
长江学者和创新团队发展计划项目(IRT1167)。
