系统评价玻璃体切割手术联合空气填充治疗特发性黄斑裂孔的疗效。

利用计算机检索Pubmed、Web of Science、Cochrane Library、Embase、中国知网及万方数据库，搜集玻璃体切割手术联合空气填充与全氟丙烷(C3F8)填充治疗特发性黄斑裂孔的相关文献。采用Cochrane系统评价的方法，由两位评价者独立进行文献检索、筛选、质量评价及数据提取。从选取的文献中提取手术时间、术后俯卧时间、裂孔闭合率及术后最佳矫正视力。采用Review Manager 5.3软件进行Meta分析，以评价玻璃体切割手术联合空气填充治疗特发性黄斑裂孔的治疗效果。采用均数差(MD)及其95%置信区间(CI)表示计量资料的效应量。各文献研究间的异质性采用Q和I²检验。如P<0.05或I²>50%则认为异质性较大，采用随机效应模型进行分析；反之，采用固定效应模型。

在各数据库初步检索到文献84篇。去除重复、内容不相符或者主要结局指标不完善及数据不能提取的文献，最终本研究纳入文献6篇。其中，1篇为英文文献，5篇为中文文献。6篇文献共纳入505只眼。其中，消毒空气组255只眼，C3F8组250只眼。3篇文献在随机分配序列生成中被评为高风险，其余纳入文献均为低风险，高风险文献进行Minors评分均为合格。纳入文献在随访偏倚、报告偏倚及其他偏倚中均被评为低风险。消毒空气组手术时间低于C3F8组。各研究间患者的手术时间组间异质性较小(I²＝0)。经固定效应模型分析，两组患者手术时间的比较，差异有统计学意义(MD＝-6.34，95%CI：-8.98~-3.71，P<0.05)。各研究间患者的裂孔闭合率组间异质性较小(I²＝0)。经固定效应模型分析，两组患者裂孔闭合率的比较，差异无统计学意义(MD＝0.91，95%CI：0.45~1.83，P>0.05)。消毒空气组患者的俯卧时间低于C3F8组。各研究间患者的俯卧时间组间异质性较大(I²＝99%)。经随机效应模型分析，两组患者术后俯卧时间的比较，差异有统计学意义(MD＝-18.90，95%CI：-24.78~-13.01，P<0.05)。各研究间患者的术后最佳矫正视力组间异质性较大(I²＝66%)。经随机效应模型分析，两组患者术后最佳矫正视力的比较，差异无统计学意义(MD＝0.00，95%CI：-0.16~0.16，P>0.05)。

玻璃体切割手术联合空气填充治疗特发性黄斑裂孔在减少手术时间及术后俯卧位时间方面的效果较好。

The aim of this study was to evaluate the effect of vitrectomy with sterile air in treatment of idiopathic macular hole (IMH) by using Meta-analysis.

The comparison of vitrectomy with C3F8 and sterile air in treatment of IMH from Chinese and English literature was identified after searching of Pubmed, Web of Science, Cochrane Library, Embase and Wanfang medical database. According to the inclusion and exclusion criteria, the data were extracted and divided into two groups: C3F8 group and sterile air group. Then Meta-analysis was performed to evaluate the postoperative vision, operation time, postoperative prone posturing period and macular hole closure rate using the ReviewManager 5.4 software. The mean difference (MD) and a confidence interval (CI) of 95% were used to describe the effect sizes of continuous data. Heterogeneity was examined by Q test and I² statistics. When P<0.05 or I²>50%, the heterogeneity was considered to be greater, and performed with a random effects model; otherwise, the fixed effects model was performed.

In all databases, 84 literatures were retrieved. After excluding the repetitive literatures, the literatures with in consistent contents or incomplete main outcome indicators and the literatures with data that cannot be extracted, there were 6 studies recruited for meta-analysis, including 505 eyes (255 eyes in air group and 250 eyes in C3F8 group). After the random bias risk assessment, 3 literatures were rated as high risk and then Minors assessment was performed. These were rated as qualifited. 6 literatures involved were rated as low risk after using follow-up bias, report bias and other biases analysis. There was non-significant heterogeneity between the air group and C3F8 group in operation time (I²=0) and significant heterogeneity in postoperative prone posturing period (I²=99%). After fixed effects model analysis, it was found that there were statistical significance difference in operation time between them (MD=-6.34, 95%CI: -8.98 to -3.71; P<0.05). After a random effects model analysis, there also was shown the significance difference in postoperative prone posturing period bewteen them (MD=-18.90, 95%CI: -24.78 to -13.01; P<0.05). There was non-significant heterogeneity between the air group and C3F8 group macular hole closure rate (I²=0) and significant heterogeneity in postoperative vision (I²=66%). However, after a random effects model analysis, there was no significant difference between the air group and C3F8 group in postoperative vision (MD=0.00, 95%CI: -0.16 to 0.16; P>0.05). After fixed effects model analysis, there was no significant difference in macular hole closure rate between them (MD=0.91, 95%CI: 0.45 to 1.83; P>0.05).

Vitrectomy with sterile air was more effective in reducing operation time and postoperative prone posturing period in the treatment of idiopathic macular hole.