Session VIII - Polytrauma

The New Injury Severity Score is a Better Predictor of Extended Hospitalization and ICU Admission than the ISS for Polytrauma Patients with Multiple Orthopaedic Injuries

Zsolt J. Balogh, MD; Endre Varga, MD, PhD; János Tomka, MD; Gábor Süveges, MD; László Tóth, MD; János A. Simonka, MD, PhD, University of Szeged, Szeged, Hungary

Purpose: Measures of tissue injury, in particular the Injury Severity Score (ISS), are robust predictors of mortality and multiple organ failure after injury. The ISS, however, has several inadequacies. It fails to account for multiple injuries to the same body region, limiting its usefulness in assessing the severity of penetrating trauma. It also underestimates multiple orthopaedic trauma in polytrauma scenarios and has been found to be a poor predictor of length of hospital stay (LOS). The modified ISS (New ISS, the NISS) is based on the three most severe injuries, regardless of body region. This modification of the ISS made calculation easier and improved the prediction of mortality and multiple organ failure, especially after penetrating trauma. Our purpose was to determine how orthopaedic injuries affect the discrepancies between the two scores and to what extent they are responsible for the extended (10 days) LOS and ICU admission. It was hypothesized that orthopaedic trauma is a major cause of extended LOS.

Methods: A total of 3100 consecutive patients with age>14 years and survival >24 hours admitted to an urban trauma center were entered into the study. Gender, age, co-morbid conditions, all Abbreviated Injury Scores for each of the six body regions, and hospital and ICU LOS were collected prospectively. The ISS and the NISS were prospectively calculated and were compared as univariate predictors of extended hospital LOS and ICU admission. Multivariate analysis was used to determine whether substitution of NISS for ISS resulted in a superior predictive model. The subgroup of patients with discrepant scores (NISS>ISS) was further evaluated to determine the impact of multiple orthopaedic trauma on patients with extended LOS and ICU admission.

Results: A total of 230 (7.5 %) patients had higher NISS than ISS, spent more days in the hospital (21.56 vs. 8.2; P <0.0001) and in the ICU (3.19 vs. 0.12; P <0.0001), than those with identical scores. The NISS was found to be more predictive of long (10 days) LOS (P <0.0001) and ICU admission (P <0.0001). The multivariate predictive model including NISS showed a better goodness of fit compared with the same model that included ISS. Among patients with discrepant scores, 61 % had multiple orthopaedic trauma responsible for the NISS>ISS; among patients with identical scores, only 0.8 % had multiple musculoskeletal injuries (P <0.0001). Sixty-six percent (25 % of the entire cohort) of the patients with more than one orthopaedic injury experienced extended hospital LOS and 37 % (1.8 % of the entire cohort) of them were admitted to the ICU (P <0.0001 for both extended LOS and ICU admission).

Conclusion: Orthopaedic injuries do have a significant effect on trauma outcomes such as LOS and ICU admission. Disregarding the anatomical regions makes calculation of the NISS easier and gives extra accuracy to the new anatomic measure. The traditional ISS should be replaced by the NISS in trauma-outcome research.