67th Annual Meeting Abstracts

---

Does Hospital TURP Case Volume Predict Patient Outcomes?
Kristin Chrouser¹, *Sonja Gerald², *David C Chang^3
1Johns Hopkins Bloomberg School of Public Health, Baltimore, MD;²Howard University College of Medicine, Washington, DC;³Johns Hopkins Department of Surgery, Baltimore, MD

Introduction:
The surgical literature evaluates hospital volume-outcome relationships using death as an endpoint but limits application to urology, as mortality rates are low. Patient safety indicators (PSI) have recently been shown as a novel method for measuring urologic surgical outcomes. PSIs are validated measures of quality of care developed by the Agency for Healthcare Research and Quality. We hypothesize that there is a volume-outcome relationship in the care of transurethral resection of the prostate (TURP) cases using PSIs as the outcome measure.
Methods:
Retrospective analysis of the Nationwide Inpatient Sample was performed (2001-2005). TURPs were identified by ICD-9 procedure codes. Primary outcome variables were mortality and PSI events. Other variables included patient demographic variables and hospital characteristics.
Results:
A total of 104,494 TURP patients were identified. Mean age was 73 years, 81.6% white, 6.8% black, and 7.1% Hispanic. Mortality was 0.3% and overall PSI rate was 0.81%. Median annual hospital TURP volume was 21 (range 7-43). After adjusting for potentially confounding patient and hospital variables, TURP volume was found to be significantly associated with one or more PSI events. A ten-case increase in annual volume decreases the overall PSI event rate by 4.0%. The volume-outcome relationship was mostly influenced by PSIs regarding hip fracture, DVT, and sepsis.
Conclusion:
Hospital TURP volume was inversely correlated with patient outcomes as measured by PSI, many related to post operative care. Future work should focus on defining the minimum number of cases that a hospital should perform to maintain high quality urologic outcomes.

---