粪便钙卫蛋白对新生儿坏死性小肠结肠炎诊断作用的Meta分析

Role of fecal calprotectin in the diagnosis of neonatal necrotizing enterocolitis: a Meta analysis

摘要:

目的 通过Meta分析的方法探讨粪便钙卫蛋白（FC）对新生儿坏死性小肠结肠炎（NEC）的诊断作用。方法 检索Web of Science、Cochrane Library，PubMed、Embase、中国知网（CNKI）、维普期刊数据库、万方数据库、中国生物医学文献数据库，并结合手动检索作为补充，搜索时间为建库至2020年5月。采用QUADAS标准评估纳入文献的质量；采用Meta-Disc 1.4和Stata 15.0软件行Meta分析，包括评价特异度、敏感度、似然比和诊断比值比，进行敏感性分析及异质性测试，绘制综合受试者工作特征（SROC）曲线及Fagan图。结果 共纳入15篇文献，包括1 719例新生儿。其中，低质量文献4篇，高质量文献2篇，其余均为中等质量。纳入研究间异质性较高，不存在阈值效应及发表偏倚。应用随机效应模型分析结果显示：FC对NEC诊断的汇总特异度、灵敏度分别为0.80（95% CI：0.78~0.82）、0.86（95% CI：0.83~0.89），阴性似然比、阳性似然比、诊断比值比分别为0.19（95% CI：0.14~0.26）、4.71（95% CI：3.57~6.23）、29.56（95% CI：17.98~48.61）。SROC曲线下面积为0.9131，Q^*指数为0.8456。Fagan图显示FC不增高时NEC发生概率为13%，FC增高时NEC发生概率提高为86%。Meta回归分析发现异质性来源于其他非协变量因素。结论 FC对NEC的早期诊断具有较高的潜力及效能。FC测定可用于NEC的诊断，但应与临床表现及其他相关实验室检查相结合。

Abstract:

Objective To study the value of fecal calprotectin(FC) in the diagnosis of neonatal necrotizing enterocolitis(NEC) through a Meta analysis. Methods Web of Science, Cochrane Library, PubMed, Embase, China National Knowledge Infrastructure, Weipu Periodical Database, Wanfang Data, Chinese Biomedical Literature Database were searched for related studies published up to May 2020, with manual search as supplementation. The QUADAS criteria were used to evaluate the quality of the articles included. Meta-DiSc 1.4 and Stata 15.0 software were used to perform the Meta analysis, including the evaluation of specificity, sensitivity, likelihood ratio, and diagnostic odds ratio. The sensitivity analysis and heterogeneity testing were performed, and the summary receiver operating characteristic(SROC) curve and Fagan diagram were plotted. Results A total of 15 articles were enrolled, involving 1 719 neonates. Among these articles, 4 had low quality, 2 had high quality, and the rest had medium quality. There was high heterogeneity between studies, and there was no threshold effect or publication bias. The random effects model analysis showed that FC had a pooled specificity of 0.80(95%CI: 0.78-0.82) and a sensitivity of 0.86(95%CI: 0.83-0.89) in the diagnosis of NEC, with a negative likelihood ratio of 0.19(95%CI: 0.14-0.26), a positive likelihood ratio of 4.71(95%CI: 3.57-6.23), and a diagnostic odds ratio of 29.56(95%CI: 17.98-48.61). The area under the SROC curve was 0.9131 and the Q^* index was 0.8456. The Fagan diagram showed that the post-test probability of NEC indicated by negative FC was 13%, while that indicated by positive FC was 86%. The Meta regression analysis showed that the heterogeneity came from other non-threshold factors. Conclusions FC has high potential and efficiency in the early diagnosis of NEC. FC measurement can be used for the diagnosis of NEC, but it should be combined with clinical manifestations and other related laboratory examinations.