Evaluation of the Performance of ACS NSQIP Surgical Risk Calculator in Emergency General Surgery Patients

Am Surg. 2020 Feb 1;86(2):83-89.

Abstract

The ACS NSQIP Surgical Risk Calculator is designed to estimate the chance of an unfavorable outcome after surgery. Our goal was to evaluate the accuracy of the calculator in our emergency general surgery population. Surgical outcomes were compared to predicted risk. The risk was calculated with surgeon adjustment scores (SASs) of 1 (no adjustment), 2 (risk somewhat higher), and 3 (risk significantly higher than estimate). Two hundred and twenty-seven patients met the inclusion criteria. An SAS of 1 or 2 accurately predicted risk of mortality (5.7% and 8.5% predicted versus 7.9% actual), whereas a risk adjustment of 3 indicated significant overestimation of mortality rate (14.8% predicted). There was good overall prediction performance for most variables with no clear preference for SAS 1, 2, or 3. Poor correlation was seen with SSI, urinary tract infection, and length of stay variables. The ACS NSQIP Surgical Risk Calculator yields valid predictions in the emergency general surgery population, and the data support its use to inform conversations about outcome expectations.

Publication types

  • Evaluation Study

MeSH terms

  • Data Accuracy
  • Emergencies*
  • Humans
  • Length of Stay
  • Logistic Models
  • Middle Aged
  • Postoperative Complications / classification
  • Postoperative Complications / mortality
  • Predictive Value of Tests
  • Retrospective Studies
  • Risk Assessment / methods*
  • Risk Factors
  • Surgical Procedures, Operative / mortality*
  • Treatment Outcome
  • Urinary Tract Infections / mortality
  • Wounds and Injuries / classification
  • Wounds and Injuries / mortality
  • Wounds and Injuries / surgery