TY - JOUR PY - 2006// TI - Different AIS triplets: Different mortality predictions in identical ISS and NISS JO - Journal of trauma A1 - Aharonson-Daniel, Limor A1 - Giveon, Adi A1 - Stein, Michael A1 - Peleg, Kobi SP - 711 EP - 717 VL - 61 IS - 3 N2 - BACKGROUND: Previous studies demonstrated different mortality predictions for identical Injury Severity Scores (ISS) from different Abbreviated Injury Scale (AIS) triplets. This study elaborates in both scope and volume producing results of a larger magnitude, applicable to specific injury subgroups of blunt or penetrating, traumatic brain injury, various age groups, and replicated on NISS. METHODS: All patients hospitalized after trauma at 10 hospitals, with ISS/NISS (new ISS) generated by two AIS triplets, excluding patients with isolated minor or moderate injuries to a single body region were studied. Patients were separated into two groups based on the different triplets. Inpatient-mortality rates were calculated for each triplet group. Odds ratios were calculated to estimate the risk of dying in one triplet group as compared with the other. The chi test determined whether the difference in mortality rate between the two groups was significantly different. Differences were further explored for various subgroups. RESULTS: There were 35,827 patients who had ISS/NISS scores generated by two different AIS triplets. Significant differences in death rates were noted between triplet groups forming identical ISS/NISS. Odds ratio for being in the second group (always containing the higher AIS score) ranged from 2.3 to 7.4. CONCLUSIONS: ISS and NISS that are formed by different AIS triplets have significantly different inpatient-mortality rates. The triplet with the higher AIS score has higher inpatient-mortality rates, overall and in several sub-populations of varying vulnerability. The comparison of populations and the interpretation of ISS/NISS based outcome data should take this important information into account and the components of AIS triplets creating each ISS and NISS should be reported.

Language: en

LA - en SN - 0022-5282 UR - http://dx.doi.org/10.1097/01.ta.0000235294.32326.e6 ID - ref1 ER -