1539 THE S.T.O.N.E. SCORE: A NOVEL INSTRUMENT TO PREDICT STONE FREE RATES IN URETEROSCOPY FROM PRE-OPERATIVE FEATURES

INTRODUCTION AND OBJECTIVES: Success of treatment for renal and ureteral stones depends on qualitative data such as stone size and location. Surgical decision making and data set comparisons would be significantly enhanced by a consistent, reproducible system that quantitates the pertinent characteristics of renal and ureteral stones. We have developed and propose a standardized lithometric scoring system (S.T.O.N.E Score) to quantify the anatomical characteristics of stones on computerized tomography. METHODS: The STONE score consists of 5 features known to effect the stone free rates of nephrolithiasis with ureteroscopy. The features examined include (S)ize, (T)opography (location), (O)bstruction, (N)umber of stones, and (E)valuation of Hounsfield Units. Each component is scored on a 1-3 point scale as shown in Table 1. We applied the STONE score to 186 consecutive ureteroscopy at Denver Health Medical Center. A logistic model was formed to our data for stone free rates. Stone free was considered the absence of stones or residual stone fragments less than 2 mm on visual inspection or by KUB. RESULTS: Stone free rates were found to be related to STONE Sum. As STONE Sum increased, the stone free rates decreased with a logical regression trend. The logistic model found was Stone Free 1 1/[1 Exp(5.5246 0.3669 * Score)] with an ROC 0.7. The comparison between our data and the model is shown in Table 2. In the STONE model, a score less than 9 has a stone free rate of more than 90% while a stone score greater than 14 has a rate less than 60%. Sum scores of 5, 6 and 14 did not fit with the model, likely because of too few cases. None of the patients in our series had a sum of 15. CONCLUSIONS: Our model correlated with our stone free rate data; however, Our model was limited by our small sample size. A model with individual weights to each feature from a multi-institutional study would more accurately predict stone free rates. This is a pragmatic model to implement in a clinical setting and is the first model to predict stone free rates from ureteroscopy.