摘要：
Objective To assess the risk of subsequent femoral shaft fractures of Gamma nails versus dynamic hip scews fixation implants for treating intertrochanteric fractures in adults. Methods All randomised or quasi-randomised trials compa-ring Gamma nails with dynamic hip scews fixation implants for treating intertrochanteric fractures in adults were searched. Two investigators independently graded study quality and abstracted relevant data. We abstracted information on subsequent femoral shaft fracture rates from studies. We pooled data using a random-effects model and tested for heterogeneity using the 12 test. We conducted subgroup analyses by date and by generation of the Gamma nail. Results We identified 14 relevant randomized trials from 1991 to 2006. Pooled data from 13trials shows the incidence of operative fracture of the femoral diaphysis is signifi-cantly increased when the Gamma nail is used(RR=2. 72,95% CI,1. 17~6. 31,P=0. 02). When the trials were subgrouped according to the date and generation of the Gamma nail,in earlier studies( before 2000 ),Gamma nails increased the risk offemoralshaftfracture3timescomparedwithdynamichipscews(RR =3. 00,95%CI,1. 02 ~8. 85,P =0. 05). However, among the most recent studies( 2 0 0 0—2 0 0 6 ),Gamma nails did not significantly increase femoral shaft fracture risk( RR =2. 33,95% CI,0. 61~8. 92,P=0. 22). Pooled data from 13 trials shows Gamma nails increased the risk of subsequent frac-ture of the femur around the implant compared with dynamic hip scews(RR=3. 62,95%CI,1. 31~10. 02,P=0. 01). In ear-lier studies( before 2000),Gamma nails increased the risk of femoral shaft fracture 4 times compared with dynamic hip scews ( RR=4. 03,95%CI,1. 20~13. 54,P=0. 02). However,among the most recent studies(2000—2006)found no difference in the rates of subsequent femoral fracture(RR=2. 80,95%CI,0. 43~18. 23,P=0. 28). Conclusion With the improvements of the Gamma nail design and learning curves with the device,the risk of the operative and subsequent femoral shaft fractures has been significantly reduced. Earlier meta-analyses and randomized trials should be interpreted with caution in light of more recent evidence.%目的：运用Cochrane系统评价的方法，对不同设计的Gamma钉进行亚组分析，比较Gamma钉与动力髋螺钉治疗股骨粗隆间骨折继发股骨干骨折的发生率。方法纳入比较Gamma钉与动力髋螺钉内固定治疗成人股骨粗隆间骨折的随机和半随机对照试验，由两名评价员独立评价文献质量并提取术中及术后股骨干骨折的患者人数，采用随机效应模型进行Meta分析，对评价指标计算相对危险度（ relative risk，RR）及其95％置信区间（ confidence inter-val，CI），将纳入文献分为2000年之前研究组和2000年之后研究组，进行亚组分析。结果共纳入14个研究，其中13个研究提供了发生术中股骨干骨折的患者资料，Meta分析结果提示Gamma钉固定治疗股骨粗隆间骨折明显增加了术中股骨干骨折的发生率（RR=2.72，95％CI，1.17~6.31，P=0.02）。2000年之前亚组包括8个研究，Meta分析结果提示，Gamma钉固定治疗股骨粗隆间骨折术中发生股骨干骨折的危险是动力髋内固定的3倍（ RR=3.00，95％CI，1.02~8.85，P=0.05）。2000年之后亚组包括5个研究，Meta分析结果提示，与动力髋内固定相比，Gamma钉固定没有增加发生术中股骨干骨折的风险，两者之间差异无统计学意义（ RR=2.33，95％CI，0.61~8.92，P=0.22）。13个研究提供了发生术后股骨干骨折的患者资料，Meta分析结果提示，与动力髋内固定相比，Gamma钉固定治疗股骨粗隆间骨折明显增加了术后股骨干骨折的发生率（ RR=3.62，95％CI，1.31~10.02，P=0.01）。2000年之前亚组的8个研究Meta分析结果提示，Gamma钉固定治疗股骨粗隆间骨折术后发生股骨干骨折的危险是动力髋内固定的4倍，两者有统计学差异（RR=4.03，95％CI，1.20~13.54，P=0.02）。2000年之后亚组包括5个研究，Meta分析结果提示， Gamma钉固定与动力髋固定相比，两者术后股骨干骨折发生率差异无统计学意义（ RR=2.80，95％CI，0.43~18.23，P=0.28）。结论随着Gamma钉设计改进及Gamma钉置入技术经验的丰富，Gamma钉固定股骨粗隆间骨折继发股骨干骨折的风险逐渐降低，因此，对先前的随机对照试验及Meta分析的结论需要保持谨慎。
