Prospective Evaluation of Association Between American College of Surgeons NSQIP Risk Estimates and Multidisciplinary Operability Decisions in Complex Surgical Patients.

Abstract

Complex surgical patients often present with frailty, comorbidities, and limited physiological reserve, making operability decisions highly challenging. We hypothesized that parameters derived from the American College of Surgeons NSQIP (ACS-NSQIP) Surgical Risk Calculator could reflect multidisciplinary assessments of operability and support preoperative decision-making in this population. A single-center prospective observational cohort study included 134 consecutive complex surgical patients evaluated by a multidisciplinary committee between November 2022 and April 2025. All patients underwent standardized preoperative assessment, including ACS-NSQIP estimation, but operability decisions were made independently of the ACS-NSQIP results. The primary endpoint was the ability of the NSQIP-derived risk deviation to discriminate between operable and nonoperable patients. This variable was analyzed using receiver operating characteristic curves, and findings were used to develop a Nonoperability Risk Score. Of the 134 patients, 98 (73.1%) were evaluated as operable and 35 (26.1%) as nonoperable. The "risk deviation" variable showed moderate to good discriminatory power (area under the curve range 0.678 to 0.740, all p < 0.01) for outcomes such as serious complications, readmission, or death. The Nonoperability Risk Score, derived from 6 ACS-NSQIP outputs, correlated strongly with the committee's operability decisions (area under the curve 0.805, p < 0.001) and demonstrated a significant trend: higher scores were associated with increased likelihood of nonoperability. Among the 80 patients who underwent operation, complication rates did not correlate with the Nonoperability Risk Score. ACS-NSQIP, particularly the derived Nonoperability Risk Score, is a useful tool to support multidisciplinary operability decisions in complex surgical patients. Their use may help standardize preoperative discussions, though further multicenter validation is needed to confirm generalizability.