BACKGROUND:The methods to treat intertrochanteric fracture are controversial. A large number of clinical studies concern the therapeutic effects of several popular methods to repair intertrochanteric fracture, but these results lack of independence, and may have bias that cannot be measured in the variable and observational studies. Thus, relevant studies have been limited. OBJECTIVE:To compare the repair effects of proximal femoral nail anti-rotation, dynamic hip screw and total hip arthroplasty on intertrochanteric fracture in aged patients by meta-analysis. METHODS: Four electronic databases were searched, including Cochrane library, PubMed, Web of Science and Chinese BioMedical Literature Database, to colect al randomized controled trials concerning the treatment of intertrochanteric fractures in the elderly with proximal femoral nail anti-rotation, dynamic hip screw and total hip arthroplasty. Using meta-analysis of evidence-based medicine, X-ray exposure time, postoperative complication rate, postoperative bed time, postoperative wound infection rate, Harris hip scores, mean operation time, intraoperative blood loss, length of hospital stays and length of the incision were compared and evaluated. Standard and methodology quality of the trials were criticaly assessed and relative data were extracted. This study used the Review Manager 5.0 software provided by Cochrane colaboration network. RESULTS AND CONCLUSION:Twelve randomized controled trials with 1 454 patients were included. Significant differences in mean operation time, intraoperative blood loss, X-ray exposure time, postoperative complication rate, and postoperative bed time were detected between the proximal femoral nail anti-rotation and dynamic hip screw groups (P < 0.05). Significant differences in length of hospital stays, postoperative complication rate, and postoperative bed time were observed between the proximal femoral nail anti-rotation and total hip arthroplasty groups (P < 0.05). Significant differences in postoperative complication rate, postoperative bed time, and Harris hip scores were detectable between the dynamic hip screw and total hip arthroplasty groups (P < 0.05). These data confirm that proximal femoral nail anti-rotation was apparently better than dynamic hip screw and total hip arthroplasty in operation time, intraoperative blood loss, length of hospital stays, postoperative complication rate, and postoperative bed time. Dynamic hip screw was better than proximal femoral nail anti-rotation in X-ray exposure time. Total hip arthroplasty was better than dynamic hip screw and proximal femoral nail anti-rotation in length of hospital stays, postoperative complication rate and postoperative bed time.