1546 PRACTICE PATTERNS FOR INPATIENT URINARY STONE SURGERY IN U.S.

Abstract

INTRODUCTION AND OBJECTIVES: The increasing prevalence of urinary stone disease has been demonstrated in numerous studies. Understanding differences in urinary stone surgery practice patterns may optimize healthcare, since different surgical techniques have different outcomes. We sought to describe the U.S. inpatient surgical practice patterns for urinary stones and determine what factors predict particular urinary stone surgery. METHODS: We used the NIS (Nationwide Inpatient Sample) database from 1998–2007 to identify 380,531 patients who underwent one of three urinary stone surgeries: percutaneous nephrolithotomy (PNL) ICD-9 55.04, 55.03, 55.21; shock wave lithotripsy (SWL) ICD-9 98.51; ureteroscopy (URS) ICD-9 56.31. Descriptive statistics were used to illustrate surgical frequencies and their potential covariates: demographics, co-morbidities, academic/community hospital, rural/urban location, U.S. geographic region (Northeast, Midwest, South, West), patient income, payer mix, and year. Linear regression illustrated surgical trends over time. Using above covariates, we compared on multi-variate analysis the odds of undergoing SWL to PNL as well as SWL to URS, as these are commonly comparable modalities. RESULTS: URS was the most common procedure (58.9%) with SWL slightly less common (22.4%) and PNL least common (16.9%) among inpatients undergoing urinary stone surgery. Over time, there was a significant increase in the utilization of PNL (p 0.001), SWL remained stable (p 0.213) and ureteroscopy declined (p 0.001). On multivariate analysis SWL was statistically more likely performed than PNL in younger subjects, male subjects, community hospitals, or rural hospitals. On multivariate analysis SWL was statistically more likely performed than URS in older subjects, male subjects, rural hospitals, or if location was South or West. CONCLUSIONS: Trends in inpatient renal stone surgery continue to evolve during the time period of our study with increasing adoption of PNL and decreasing use of ureteroscopy. Age and gender as well as type of hospital and its rural location predict SWL use over PNL, despite the lower efficacy associated with SWL technology. Assuming equal stone size (data not available), further explanation is necessary to reconcile these differences. Similar findings were found in the SWL versus URS analysis, though SWL appeared to have regionalization. Further investigation is necessary to understand causes for the differences inpatient practice patterns and whether outpatient practice patterns share common findings.