Assessment of the NSQIP Surgical Risk Calculator in Predicting Microvascular Head and Neck Reconstruction Outcomes

Yue Ma, Benjamin M. Laitman, Vir Patel, Marita Teng, Eric Genden, Samuel DeMaria, Brett A. Miles

Research output: Contribution to journalArticlepeer-review

8 Scopus citations

Abstract

Objective: This study evaluated the accuracy of the Surgical Risk Calculator (SRC) of the ACS NSQIP (American College of Surgeons National Surgical Quality Improvement Program) in predicting head and neck microvascular reconstruction outcomes. Study Design: Retrospective analysis. Setting: Tertiary medical center. Subjects and Methods: A total of 561 free flaps were included in the analysis. The SRC-predicted 30-day rates of postoperative complications, hospital length of stay (LOS), and rehabilitation discharge were compared with the actual rates and events. The SRC’s predictive value was examined with Brier scores and receiver operating characteristic area under the curve. Results: A total of 425 myocutaneous, 134 osseous (84 fibula, 47 scapula, and 3 iliac crest), and 2 omental free flaps were included in this study. All perioperative complications evaluated had area under the curve values ≤0.75, ranging from 0.480 to 0.728. All but 2 postoperative complications had Brier scores >0.01. SRC-predicted LOS was 9.4 ± 2.38 days (mean ± SD), which did not strongly correlate with the actual LOS of 11.98 ± 9.30 days (r = 0.174, P <.0001). Conclusion: The SRC is a poor predictor for surgical outcome among patients undergoing microvascular head and neck reconstruction.

Original languageEnglish
Pages (from-to)100-106
Number of pages7
JournalOtolaryngology - Head and Neck Surgery
Volume160
Issue number1
DOIs
StatePublished - 1 Jan 2019

Keywords

  • NSQIP
  • free flap
  • free tissue transfer
  • surgical risk calculator

Fingerprint

Dive into the research topics of 'Assessment of the NSQIP Surgical Risk Calculator in Predicting Microvascular Head and Neck Reconstruction Outcomes'. Together they form a unique fingerprint.

Cite this