Computed tomography-based novel prediction model for the stone-free rate of ureteroscopic lithotripsy

Abstract

The purpose of this study was to evaluate whether computed tomography (CT) parameters can predict the success of ureteroscopic lithotripsy (URSL) and establish a model for predicting the success rates of a single URSL procedure for the treatment of a single ureteral stone. We retrospectively reviewed the records of 237 patients who underwent URSL for ureteral stones diagnosed by CT between January 2009 and June 2012. Stone-free status was defined as the absence of stones or residual stone fragments < 2 mm by ureteroscopy and plain abdominal radiography. We analyzed the correlations between the outcome of URSL and the patients' sex, age, height, body weight, body mass index, and history of ureteral stone. Stone factors such as the diameter (D), stone height (H), volumetric stone burden (VSB; D (2) x H x 5 mm x pi x 1/6), estimated stone location (ESL; number of axial cut images between the stone and uretero-vesical junction), tissue rim sign (RS; 0-3), perinephric edema (0-3), hydronephrosis (0-3), and Hounsfield unit (HU) were also analyzed. We then developed a model to predict the probability of successful URSL by applying a logistic model to our data. The success rate of URSL was 85.7 % (203/237). Univariate analysis found that stone diameter, length, VSB, ESL, HU and RS significantly affected the stone-free rate. Multivariate analysis indicated that stone diameter, ESL and RS independently influenced the stone-free rate. The logistic model indicated that success rates = 1/[1 + exp{-6.146 + 0.071(D) + 0.153(ESL) + 1.534(RS)}] with an area under the receiver operating characteristic curve of 0.825. Stone diameter, ESL, and RS were independent predictors of the outcome of a single URSL for a single ureteral stone.