An algorithm for preoperative prediction of reoperation risk after internal fixation of femoral neck fractures

An algorithm was designed for preoperative prediction of the risk for reoperation, and the mortality risk, after internal fixation of femoral neck fractures. Out of 51 reviewed studies of femoral neck fractures, eight met specified inclusion criteria such as low dropout rates, a minimum of ten surgeons performing the surgery, a minimum of 2 years follow-up, and a standard age, sex, and Garden class distribution. Five of these studies were used for the construction of the algorithm, and the remaining three for testing the specificity and sensitivity of the algorithm. A separate analysis of the influence of age on the reoperation rate was also performed. In the analysis of 399 reviewed cases of femoral neck fractures, the specificity for the algorithm in predicting the risk for reoperation was 96%, and the sensitivity was 51%. The positive predictive value for the algorithm in predicting the risk for reoperation was 77%, which was three times higher compared to the commonly used predictors age and Garden class (positive predictive value 25%). For prediction of the mortality risk the positive predictive value for-the algorithm was 57%.