RESUMO
BACKGROUND@#The purpose of this study was to analyze cases of AO31-A2 intertrochanteric fractures (ITFs) and to identify the relationship between the loss of the posteromedial support and implant failure.@*METHODS@#Three hundred ninety-four patients who underwent operative treatment for ITF from January 2003 to December 2017 were enrolled. Focusing on posteromedial support, the A2 ITFs were divided into two groups, namely, those with (Group A, n = 153) or without (Group B, n = 241) posteromedial support post-operatively, and the failure rates were compared. Based on the final outcomes (failed or not), we allocated all of the patients into two groups: failed (Group C, n = 66) and normal (Group D, n = 328). We separately analyzed each dataset to identify the factors that exhibited statistically significant differences between the groups. In addition, a logistic regression was conducted to identify whether the loss of posteromedial support of A2 ITFs was an independent risk factor for fixation failure. The basic factors were age, sex, American Society of Anesthesiologists (ASA) score, side of affected limb, fixation method (intramedullary or extramedullary), time from injury to operation, blood loss, operative time and length of stay.@*RESULTS@#The failure rate of group B (58, 24.07%) was significantly higher than that of group A (8, 5.23%) (χ = 23.814, P < 0.001). Regarding Groups C and D, the comparisons of the fixation method (P = 0.005), operative time (P = 0.001), blood loss (P = 0.002) and length of stay (P = 0.033) showed that the differences were significant. The logistic regression revealed that the loss of posteromedial support was an independent risk factor for implant failure (OR = 5.986, 95% CI: 2.667-13.432) (P < 0.001).@*CONCLUSIONS@#For AO31-A2 ITFs, the loss of posteromedial support was an independent risk factor for fixation failure. Therefore, posteromedial wall reconstruction might be necessary for the effective treatment of A2 fractures that lose posteromedial support.