系统评价直接前入路(DAA)、经皮穿刺辅助保留关节囊全髋关节(SuperPATH)手术入路(SPA)和后外侧入路(PLA)进行全髋关节置换术(THA)治疗髋部疾病的临床疗效。

应用计算机检索PubMed、荷兰医学文摘数据库(EMbase)、中国期刊全文数据库(CNKI)、万方及维普等数据库建库至2018年10月国内外公开发表的试验组为SuperPATH入路或DAA入路，对照组为PLA的疗效比较的随机对照研究或观察性研究文章。由2位研究者独立完成文章筛选(排除无DAA、SuperPATH、PLA行THA两两对比，总样本量少于30例，未提及分配方法、盲法，非中文、英文的文献)、数据提取，依据Cochrane风险偏倚评估工具进行文章方法学质量评价后，采用RevMan 5.3及间接比较软件(ITC)进行Meta分析。

最初检出501篇相关文献，经严格筛选最终纳入6篇文献，均为随机对照研究文献。纳入的研究对象共478例，其中试验组241例，对照组237例。Meta分析结果显示，直接比较：DAA入路和PLA切口长度的比较，差异无统计学意义[WMD＝－1.30，95%CI(－3.27，0.67)]。SuperPATH入路组的切口长度小于PLA组[WMD ＝ －7.07，95%CI(－8.21，－5.93)]。DAA[WMD＝2.37，95%CI(－30.19，34.93)]、SuperPATH入路[WMD＝12.26，95%CI(－3.22，27.74)]和PLA行THA的手术时间比较，差异均无统计学意义。DAA入路和PLA行THA的术中出血量比较，差异无统计学意义[WMD＝－37.70，95%CI(－91.14，15.75)]。SuperPATH入路组的术中出血量小于PLA组[WMD＝－171.56，95%CI(－252.92，－90.20)]。DAA入路组[WMD＝7.10，95%CI(5.54，8.66)]、SuperPATH入路组[WMD＝5.80，95%CI(0.10，11.50)]的术后1个月的Harris髋关节功能评分高于PLA组。校正间接比较：DAA入路组的切口长度小于SuperPATH入路组[MD＝5.77，95%CI(3.94，8.046)]。2组的手术时间比较，差异无统计学意义[MD＝－9.89，95%CI(－45.943，26.163)]。DAA入路组的术中出血量小于SuperPATH入路组[MD＝133.86，95%CI(36.79，230.93)]。2组术后1个月的Harris髋关节功能评分比较，差异无统计学意义[MD＝1.3，95%CI(－4.61，7.21)]。

在治疗髋部疾病的3种不同手术入路中，DAA仅术后Harris髋关节功能评分要优于PLA，其他方面可能与DAA学习曲线长，股骨近端暴露困难等有关，导致其无明显差异。SuperPATH入路在切口长度、术中出血量、术后Harris髋关节功能评分方面要优于PLA。以PLA为对照，间接比较DAA与SuperPATH入路，DAA在切口长度、术中出血量方面占有优势，但中长期的临床疗效还需要更多的优质量文章进一步研究证实。

To systematically evaluate the clinical outcomes of direct anterior approach (DAA), supercapsular percutaneously assisted total hip ( SuperPATH) approach (SPA) and posterolateral approach (PLA) for total hip arthroplasty (THA) in the treatment of hip diseases.

All the randomized controlled trial ( RCT) articles and observational research articles about the curative effect of THA through SuperPATH approach ( SPA) and direct anterior approach (DAA) versus posterolateral approach (PLA) for treatment of hip diseases that published at home and abroad from database establishing to October 2018 were retrieved from PubMed, Excerpt Medica Database (EMbase), China National Knowledge Infrastructure (CNKI), WanFang, China Science and Technology Journal Database (VIP) and other databases. Two researchers independently completed the article selection (excluding the comparison of THA without DAA, SPA and PLA, the total sample sizes were less than 30 cases, no mention of allocation method, blind method, non-Chinese or English literature), data extraction, and meta-analysis using RevMan 5.3 and indirect treatment comparison (ITC) software after evaluating the methodological quality of articles based on Cochrane risk bias assessment tool.

A total of 501 articles were initially detected, and six articles were included after strict screening, all of which were randomized controlled studies. A total of 478 subjects were included, including 241 in the experimental group and 237 in the control group. The meta-analysis results showed that there was no statistically significant difference in the incision length between the DAA and PLA [WMD= -1.30, 95%CI(-3.27, 0.67)]. The incision length in the SPA group was smaller than that in the PLA group [WMD=-7.07, 95%CI(-8.21, -5.93)]. There was no statistically significant difference in the operative time of THA in DAA [WMD=2.37, 95%CI(-30.19, 34.93)], SPA [WMD=12.26, 95%CI(-3.22, 27.74)] and PLA. There was no significant difference in intraoperative bleeding volume between DAA and PLA in THA [WMD=-37.70, 95%CI(-91.14, 15.75)]. Intraoperative bleeding in the SPA group was less than that in the PLA group [WMD=-171.56, 95%CI(-252.92, -90.20)]. In the DAA group [WMD=7.10, 95%CI (5.54, 8.66)] and SPA group [WMD=5.80, 95%CI (0.10, 11.50)], the Harris hip function score one month after surgery was higher than that in the PLA group. Indirect comparison of correction: the incision length of the DAA group was smaller than that of the SPA group [MD=5.77, 95%CI(3.94, 8.046)]. There was no statistically significant difference in operative time between the two groups [MD=-9.89, 95%CI(-45.943, 26.163)]. Intraoperative blood loss in the DAA group was lower than that in the SPA group [MD=133.86, 95%CI(36.79, 230.93)]. There was no statistically significant difference in Harris hip function scores one month after surgery between the two groups [MD=1.3, 95%CI (-4.61, 7.21)].

Among the three different approaches for the treatment of hip diseases, the hip function of DAA is better than PLA only after the operation. Other aspects may be related to the long DAA learning curve and the difficulty of proximal femur exposure, leading to no significant difference. The SPA is superior to the PLA in terms of incision length, intraoperative blood loss and postoperative hip function. While DAA and SPA are indirectly compared using PLA as control, DAA shows advantages in terms of incision length and intraoperative blood loss, but its medium-and long-term clinical efficacy still needs further studies and confirmation of more high-quality articles.