关节置换与保留关节修复肱骨近端移位3，4部分骨折的Meta分析

背景：目前的肱骨近端骨折治疗方法包括保留关节治疗（保守治疗、切开复位内固定等）和关节置换，如何选择存在争议。目的：收集肱骨近端移位3，4部分骨折保留关节与关节置换治疗的随机对照试验，对两种治疗方法的疗效差异进行Meta分析。方法：检索Medline（1966年1月至2013年12月）、Pubmed（1980年1月至2013年12月）、Embase（1990年1月至2013年12月）、Science（1990年1月至2013年12月）、Springer（1990年1月至2013年12月）、CNKI（1994至2013年）、万方数据库（1982至2013年）等数据库中关于肱骨近端3，4部分骨折保留关节与关节置换治疗的随机对照试验，纳入符合标准的文献，提取相关数据输入 Comprehensive Meta-Analysis Software （CMA）软件进行Meta分析。结果与结论：共7篇文献符合纳入标准，共计320例患者（165例行保留关节治疗，155例行关节置换治疗），均为英文文献，经改良的Jadad评分判定均为高质量研究。Meta分析结果显示，总Constant平均值为55.9（95%CI：50.7-61.1，P〈0.001），保留关节治疗组的Constant评分值高于关节置换组（P〈0.01）。研究的异质性显著（I2=88%，Q统计量=107.6，Q=13，P〈0.001），在Meta分析时肩关节Constant评分值随着男女性别比增长、肱骨结节骨吸收率增加而下降。提示保留关节肱骨近端移位3，4部分骨折患者治疗后期可较关节置换得到较高的Constant评分值。此外年龄、骨折类型、性别比和并发症发生率是影响Constant值的显著因子。鉴于研究的异质性和方差，需要有更多更好的随机对照试验进一步证明。

Arthroplasty versus joint preservation for displaced 3-and 4-part proximal humeral fractures：a meta-analysis

BACKGROUND：Currently, the treatment of proximal humeral fractures mainly contains joint preservation （conservative treatment, open reduction and internal fixation） and arthroplasty, but how to choose the treatment is stil controversial. OBJECTIVE：To evaluate and compare the clinical outcomes of joint preservation versus arthroplasty in the treatment of displaced 3-or 4-part humeral fractures in randomized control ed trials using meta-analysis. METHODS：Medline （January 1966 to December 2013）, PubMed （January 1980 to December 2013）, Embase （January 1990 to December 2013）, Science （January 1990 to December 2013）, Springer （January 1990 to December 2013）, China National Knowledge Infrastructure （1994 to 2013）, and Wanfang database （1982 to 2013） were searched for randomized control ed trials addressing joint preservation and arthroplasty for 3-or 4-part proximal humeral fractures. Articles meeting the inclusion criteria were included. The related data were extracted and loaded onto Comprehensive Meta-Analysis Software for meta-analysis. RESULTS AND CONCLUSION：Seven articles with 320 patients （165 patients undergoing joint preservation and＆amp;nbsp;155 patients receiving arthroplasty） were accepted in this mete-analysis and al of them were high-quality English researches by modified Jadad Scale. Meta-analysis results displayed that the random-effects mean Constant score across al types was 55.9 （95%CI：50.7-61.1;P〈0.001）. Constant score was higher in the joint preservation group than in the arthroplasty group （P〈0.01）. The study displayed significant heterogeneity （I2 0.001）. In the meta-analysis, Constant scores were decreased with increasing rate of male to female and rate of tuberosity resorption. These results suggested that 3-or 4-part proximal humeral fractures demonstrate improved Constant scores when treated with joint-preserving options compared with arthroplasty. Moreover, age, fracture pattern, gender ratio and complication rate are significant