消化道肿瘤病人术后肠内免疫营养支持对感染性并发症及住院时间影响的系统综述和Meta分析

[目的]系统评价术后肠内免疫营养支持对术后感染性并发症和住院时间的影响。[方法]计算机检索中文全文期刊数据库（CNKI ）、万方数据库、维普数据库、PubMed,Medline,Cochran Library,由2位研究者采用Cochrane系统评价法进行评价和数据提取,采用 RevMan5.1软件进行Meta分析。[结果]最终纳入9篇随机对照研究,共800例病人。Meta 分析显示肠内免疫营养支持组和肠外营养组肺部感染发生率[OR=0.61,95%置信区间为（0.35,1.06）]、切口感染发生率[OR=0.71,95%置信区间为（0.37,1.36）]、尿路感染发生率[OR=0.39,95%置信区间为（0.14,1.14）]、腹腔感染发生率[OR=0.48,95%置信区间为（0.22,1.07）]差异均无统计学意义（P〉0.05）；两组住院时间加权均数差为-3.13,95%置信区间为（-5.54,-0.71）,差异有统计学意义（P〈0.05）。[结论]消化道肿瘤病人术后应用肠内免疫营养支持可能不会降低术后常见感染性并发症的发生率,但可能减少病人的住院时间；由于本研究的局限性,可能需要更多高质量的随机对照研究和 Meta分析结果来验证。

Systematic review and meta analysis of influence of postoperative intestinal immune nutritional support on infectious compl ications and hospital ization time in gastrointestinal cancer patients

To systematically evaluate the influence of postoper-ative intestinal immune nutritional support on postoperative infectious complications and hospitalization time.Methods：The CNKI,Wangfang datebase,VIP datebase,PubMed,Medline and Cochran Library were searched by computers.Two investigators used Cochrane system evalua-tion method to evaluate and extract data,and all the data were analyzed with RevMan5 .1 .Software.Results：A total of nine randomized controlled trials involving 800 patients met the inclusion criteria.The results showed that there were no significantly statistical differences in the rate of pulmo-nary infection[OR=0.61,95%CI（0.35,1.06）]、incision infection[OR=0.71,95%CI（0.37,1.36）]、urinary tract infection[OR=0.39,95%CI（0.14, 1.14）]、abdominal infection[OR=0.48,95%CI（0.22,1.07）]be-tween both groups （P〉0.05）.There was significantly statistical differ-ence in length of stay in hospital[WMD=-3.13,95%CI（-5.54,-0.71）] （P〈0.05）.Conclusion：The postoperative application of enteral im-mune nutrition support for gastrointestinal cancer patients may not reduce the incidence of postoperative common infectious complica-tions,but it maybe shrink the length of hospital stay.Limited to the study,more high quality RCTs and Meta analysis are required to ver-ify the effects.