Background: The Injury Severity Score (ISS) has been validated in numerous studies and has become one of the most common trauma scoring systems since its inception. The ISS equation was later modified to create the New Injury Severity Score (NISS). By using the three most severe injuries regardless of body region, the NISS seems well suited to describe patients of penetrating trauma, where injuries often cluster within a single body region. We hypothesized that NISS would better predict outcomes than ISS in penetrating trauma patients.
Methods: An analysis (June 2008 to March 2009) of all severely injured (length of hospital stay ≥ 48 hours, intensive care unit admission, interhospital transfer, or death) penetrating trauma patients revealed final study sample of 256 patients. ISS and NISS were compared as predictors for both mortality and complications through area under the receiver operating characteristic curve, Hanley-McNeil test, multiple-variable logistic regression, and Hosmer-Lemeshow goodness-of-fit test analysis.
Results: Of 256 study patients, 195 (76.2%) survived until discharge. The mean (ISS, 21.7 ± 21.1 vs. NISS, 27.4 ± 22.0; p < 0.001) and median (ISS, 14.0 vs. NISS, 21.0) ISS was lower than those of the NISS. Overall, 173 patients (67.6%) had discordant scores with 26% and 43% having scores greater than 25 (ISS and NISS, respectively, p < 0.01). The mortality area under the curve (AUC) for NISS was greater than the AUC for ISS in all penetrating patients (0.930 vs. 0.885, p = 0.008), those with penetrating torso injuries (NISS, 0.934 vs. ISS, 0.881, p < 0.001), and those with severe (score > 25) injuries (NISS, 0.845 vs. ISS, 0.761, p < 0.001). In patients surviving for more than 48 hours, the complications AUC for NISS was also greater than the AUC for ISS (NISS, 0.838 vs. ISS, 0.784; p = 0.023).
Conclusion: The NISS outperformed ISS as a predictor of both mortality and complications in civilian penetrating trauma patients. These results indicate that NISS is a superior scoring system for patients with penetrating injuries.
Level of evidence: Prognostic study, level III.