Impact of ASA score misclassification on NSQIP predicted mortality: a retrospective analysis

Alex Helkin, Sumeet V. Jain, Angelika Gruessner, Maureen Fleming, Leslie Kohman, Michael Costanza, Robert N. Cooney

The ASA physical classification score has a major impact on the observed/expected (O/E) mortality ratio in the NSQIP General Vascular Mortality Model. The difference in predicted mortality is greatest between ASAs 3 and 4. We hypothesized under-classified ASA scores significantly affect the O/E mortality. We conducted a retrospective review of NSQIP essential surgery cases from January 2014 to December 2014 (n = 1264) with mortality sub-analysis (n = 33) at our institution. We recorded transfer and emergency status and independently calculated the ASA score for mortalities using published definitions. A random sample of 50 survivors and 10 emergency survivors were reviewed and ASA recalculated. We performed statistical modeling to simulate the effects of ASA misclassifications. Statistical analysis was performed using JMP 10 and SAS 9.4. ASA was under-classified in 18.2% of mortalities, most commonly ASAs 3 and 4. Sixteen percent of ASA 3 survivors were misclassified, including 60% in the emergency subgroup (p < 0.05 vs. elective cases). Patients transferred from other institutions were more likely to be emergency cases than non-transferred patients (43.5 vs. 7.84%, p < 0.05). Transferred patients had a higher proportion of ASAs 3-5 vs. ASAs 1-2 compared with non-transfers (84.38 vs. 49.76%, p < 0.05) Simulation data showed ASA misclassification underestimated predicted mortality by 2.5 deaths on average. ASA misclassification significantly impacts O/E mortality. With accurate ASA classification, observed mortality would not have exceeded expected mortality in our institution. Education regarding the impact of ASA scoring is critical to ensure accurate O/E mortality data at hospitals using NSQIP to assess surgical quality.