Since the method of femoral fixation was decided by the attending surgeon this study is susceptible to selection bias, we tried to compensate this in the adjusted analysis using the CCI to compensate for different patient baseline characteristics.