Abstract: PD57-09
Sources of Funding: University of Chicago Institute of Translational Medicine, Core Subsidy Grant

Introduction

The American College of Surgeons&[prime] National Surgical Quality Improvement Program (NSQIP) Risk Calculator is commonly used in the preoperative setting. The risk profile generated is often used during the informed consent process and may in the near future be used as a quality measure linked with reimbursement. We aimed to evaluate the accuracy of this NSQIP risk calculator in patients undergoing radical cystectomy with urinary diversion.

Methods

We retrospectively reviewed our institutional database to identify patients undergoing radical cystectomy with urinary diversion between 2010 and 2015. We used the proprietary NSQIP online calculator, which incorporated the procedure-specific CPT code, to obtain a 30-day postoperative risk profile for each of eleven outcomes, which were then compared to actual outcomes for each patient. Brier scores (BS) were calculated as a measure of NSQIP calculator accuracy. Consistent with prior studies, we selected a threshold of BS <0.01 (90% accuracy) as an acceptable calculator.

Results

We included 567 patients who underwent radical cystectomy, of whom 364 (64%) received an ileal conduit (IC) and 203 (36%) received orthotopic neobladder diversion (ONB). Mean age was 68.24 years (±10.40) and 435 (76%) were male. BS exceeded the threshold of 0.01 (indicating poor predictive value) for serious complications, any complications, surgical site infection, urinary tract infection, deep venous thrombosis, renal failure, readmission, return to operating room, and discharge to rehabilitation facility, regardless of diversion type (see Figure 1) - risk was underestimated for each of these complications. Length of stay was underestimated by 17% and risk of serious complication was underestimated by 31%. The calculator did adequately predict the risk of death for patients receiving ONB and overall.

Conclusions

The universal NSQIP surgical risk calculator inaccurately predicts most postoperative complications in patients undergoing radical cystectomy with urinary diversion. This highlights the need for a procedure-specific risk calculator in order to better counsel patients in the preoperative setting and generate realistic quality measures._x000D_ _x000D_

Funding

University of Chicago Institute of Translational Medicine, Core Subsidy Grant