A new prediction model for operative time of flexible ureteroscopy with lithotripsy for the treatment of renal stones

Abstract

This study aimed to develop a prediction model for the operative time of flexible ureteroscopy (fURS) for renal stones. We retrospectively evaluated patients with renal stones who had been treated successfully and had stone-free status determined by non-contrast computed tomography (NCCT) 3 months after fURS and holmium laser lithotripsy between December 2009 and September 2014 at a single institute. Correlations between possible factors and the operative time were analyzed using Spearman’s correlation coefficients and a multivariate linear regression model. The P value < 0.1 was used for entry of variables into the model and for keeping the variables in the model. Internal validation was performed using 10,000 bootstrap resamples. Flexible URS was performed in 472 patients, and 316 patients were considered to have stone-free status and were enrolled in this study. Spearman’s correlation coefficients showed a significant positive relationship between the operation time and stone volume (ρ = 0.417, p < 0.001), and between the operation time and maximum Hounsfield units (ρ = 0.323, p < 0.001). A multivariate assessment with forced entry and stepwise selection revealed six factors to predict the operative time of fURS: preoperative stenting, stone volume, maximum Hounsfield unit, surgeon experience, sex, and sheath diameter. Based on this finding, we developed a model to predict operative time of fURS. The coefficient of determination (R ²) in this model was 0.319; the mean R ² value for the prediction model was 0.320 ± 0.049. To our knowledge, this is the first report of a model for predicting the operative time of fURS treatment of renal stones. The model may be used to reliably predict operative time preoperatively based on patient characteristics and the surgeons’ experience, plan staged URS, and avoid surgical complications.