胸腔镜治疗食管闭锁的术后并发症发生率的Meta分析

目的 运用Meta分析对胸腔镜治疗食管闭锁术后并发症的发生率进行综合评价.方法 收集近十年国内外胸腔镜治疗食管闭锁术后患儿共407例,检索万方数据库、中国知网、医知网、PubMed数据库,收集所有胸腔镜治疗食管闭锁的相关文献,根据文献纳入和排除标准从中筛选出有记录术后并发症病例数的研究,运用R3.1.2软件中Meta程序包meta进行率的Meta分析,采用Freeman.Tukey双反正弦变换法进行率的合并计算.先进行异质性检验,结果若为P＞0.05,说明纳入文献间是同质性,采用固定效应模型,反之说明文献间是异质性,采用随机效应模型.然后采用秩相关检验法对估计研究的个数进行发表偏倚评估.P≤0.05为差异有统计学意义.结果 共入选相关文献14篇,累计胸腔镜治疗食管闭锁术后患儿407例,合并的并发症发生率为49％(95％CI:37％～61％).其中发生率最高的并发症为吻合口狭窄和吻合口漏.吻合口狭窄的发生率为17％(95％CI:11％～26％),吻合口漏的发生率为11％ (95 ％CI:8％～15％).结论 Meta分析作为一种新的文献研究分析方法,能系统、客观地对多个研究结果进行综合评价和定量分析,可提高检验统计效能.

Incidence of postoperative complications for thoracoscopy of esophageal atresia: a meta-analysis

Objective To evaluate the incidence of postoperative complications for thoracoscopy of esophageal atresia (EA) through a literature review.Methods A total of 407 domestic and foreign eases of thoracoscope of esophageal atresia were retrieved for the last decade.The relevant databases were Wanfang,China Hownet,Chinese Medical Knowledge Network and PubMed.Based upon the inclusion and exclusion criteria,all records of postoperative complications were collected.The Meta R 3.1.2 software package was utilized.And Freeman and Tukey double arc sine transformation method was employed for calculating the merger of rates.Heterogeneity inspection was performed first.If the result was P＞0.05,homogeneity dictated using fixed effect model.If the result was P＜0.05,random effect model was utilized.Then rank correlation test was used for estimating the number of studies for publication bias.If P≤0.05,the difference was deemed as statistically significant.Results There were 14 articles involving 407 cases of thoracoscopy of esophageal atresia with postoperative complications at a rate of 49％ (95％CI:37％-61％).Two most frequent complications were anastomotic stenosis and anastomotic leakage.The rate of anastomotic stenosis was 17％ (95％CI:11％-26％) and the rate of anastomotic leakage 11％(95％CI:8％-15％).Conclusions As a new method of literature researches and analyses,meta-analysis may be used for comprehensive evaluations and quantitative analyses.