Logistic regression model for predicting stone-free rate after minimally invasive percutaneous nephrolithotomy

To investigate the prognostic factors associated with the treatment efficacy of minimally invasive percutaneous nephrolithotomy (MPCNL) and develop a preoperative logistic regression model for predicting the stone-free rate after the initial procedure.

We retrospectively analyzed the records of 865 patients who had undergone MPCNL in our department from January 2006 to September 2009. Patient age, sex, body mass index, degree of hydronephrosis, and stone side, number, size, and location were the investigated variables. According to the treatment outcome, the patients were divided into 2 groups, those who became stone free and those who did not. Student's t test, chi-square test, and multiple logistic regression analysis were performed to determine the statistically significant variables and to develop a predictive mathematical model.

The stone-free rate after primary MPCNL was 80.1% (693 of 865). On univariate analysis, the stone number, size, and location and degree of hydronephrosis were identified as significant factors between the 2 groups. On multivariate analysis, they were also independent predictors of the surgical outcome. Next, a logistic regression model was developed using these variables to estimate the stone-free rate after MPCNL.

The results of our study have demonstrated that an increased stone number and size, location in a calix, staghorn calculus, and moderate to severe hydronephrosis were associated with decreased stone-free rates after MPCNL. We have developed a mathematical model for predicting the stone-free rate that will be helpful for patient counseling and surgeon decision-making regarding the management of upper urinary tract calculi.