Citation

Rutledge R, Fakhry SM, Baker C, Oller D. J. Trauma 1993; 35(4): 497-506; discussion 506-7.

Affiliation

North Carolina Trauma Registry, North Carolina Memorial Hospital, Wake Medical Center/Area Health Education Center, Chapel Hill.

Copyright

(Copyright © 1993, Lippincott Williams and Wilkins)

DOI

unavailable

PMID

8411270

Abstract

The purpose of this study was to develop a simplified method of stratifying patient risk of death based on ICD-9 codes. METHODS: Data were obtained from a statewide trauma registry. A mortality risk ratio (MRR) was derived from a "training" subset by calculating a mortality rate for each ICD-9 code of interest. The independent variables of interest included TS, ISS, and MRRs (for the 1st & 2nd Dx, 1st op, & E code). RESULTS: (n = 37,100). When the 1st Dx and ISS were used as candidate variables in stepwise multivariate modeling, the MRR for the 1st Dx was the first variable to be entered into the model (1st Dx partial R2 = 0.37, ISS partial R2 = 0.02). CONCLUSION: This study shows that the 1st Dx is a better predictor of outcome than ISS. Since ICD-9 codes are more easily obtained and are better predictors of outcome, this study suggests that they may supersede the use of the ISS in injury severity scoring.