Predictors of clinical outcome after minimally invasive percutaneous nephrolithotomy for renal calculus.

To investigate all the predictors of operative duration, hospital stay and stone-free rate post-minimally invasive percutaneous nephrolithotomy (MPCNL) and to establish a logistic regression formula to predict the probability of stone-free post-MPCNL. From August 2009 to August 2012, 396 patients were enrolled in the present study. The patients' characteristics, history, laboratory examination and imaging information were used as independent variables, and operative duration, hospital stay, residuals (≥4 mm) as outcomes. Univariate and multiple regression analysis were performed to determine independent variables. According to the result of multiple logistic regression analysis, a logistic regression formula was established to predict the stone-free probability post-MPCNL. The predictors of operative duration were type of stones and stone burden. Factors affecting hospital stay included leucocytes, stone burden, type of stones and hemoglobin. Residuals were significantly related with stone burden and type of stones. With the above results, we developed a logistic regression formula, which can effectively predict the probability of stone-free with 91.7% of sensitivity, 70.4% of specificity and 85.9% of overall accuracy.