Background: The Military Injury Severity Score (mISS) was developed to better predict mortality in complex combat injuries but has yet to be validated.
Methods: US combat trauma data from Afghanistan and Iraq from January 1, 2003, to December 31, 2014, from the US Department of Defense Trauma Registry (DoDTR) were analyzed. Military ISS, a variation of the ISS, was calculated and compared with standard ISS scores.Receiver operating characteristic curve, area under the curve, and Hosmer-Lemeshow statistics were used to discriminate and calibrate between mISS and ISS. Wilcoxon-Mann-Whitney, t test and χ tests were used, and sensitivity and specificity calculated. Logistic regression was used to calculate the likelihood of mortality associated with levels of mISS and ISS overall.
Results: Thirty thousand three hundred sixty-four patients were analyzed. Most were male (96.8%). Median age was 24 years (interquartile range [IQR], 21-29 years). Battle injuries comprised 65.3%. Penetrating (39.5%) and blunt (54.2%) injury types and explosion (51%) and gunshot wound (15%) mechanisms predominated. Overall mortality was 6.0%.Median mISS and ISS were similar in survivors (5 [IQR, 2-10] vs. 5 [IQR, 2-10]) but different in nonsurvivors, 30 (IQR, 16-75) versus 24 (IQR, 9-23), respectively (p < 0.0001). Military ISS and ISS were discordant in 17.6% (n = 5,352), accounting for 56.2% (n = 1,016) of deaths. Among cases with discordant severity scores, the median difference between mISS and ISS was 9 (IQR, 7-16); range, 1 to 59. Military ISS and ISS shared 78% variability (R = 0.78).Area under the curve was higher in mISS than in ISS overall (0.82 vs. 0.79), for battle injury (0.79 vs. 0.76), non-battle injury (0.87 vs. 0.86), penetrating (0.81 vs. 0.77), blunt (0.77 vs. 0.75), explosion (0.81 vs. 0.78), and gunshot (0.79 vs. 0.73), all p < 0.0001. Higher mISS and ISS were associated with higher mortality. Compared with ISS, mISS had higher sensitivity (81.2 vs. 63.9) and slightly lower specificity (80.2 vs. 85.7).
Conclusion: Military ISS predicts combat mortality better than does ISS.
Level of evidence: Prognostic and epidemiologic study, level III.