Settling the Score: Injury Severity Score Fails to Capture Nuances in Pediatric Trauma. Issue 2 (9th February 2022)

• Record Type: Journal Article
• Authors: Hatchimonji, Justin S.; Luks, Valerie L.; Swendiman, Robert A.; Allukian, Myron; Nance, Michael L.; Nace, Gary W.
• Is Part Of: Pediatric emergency care. Volume 38:Issue 2(2022)
• Page Start: e828
• Page End: e832
• Publication Date: 2022-02-09
• Subjects: pediatric trauma -- risk score -- Injury Severity Score -- mortality prediction; Pediatric emergencies -- Periodicals; 618.92002505
• DOI: 10.1097/PEC.0000000000002363
• Languages: English
• ISSNs: 0749-5161

Abstract

Abstract : Objectives: Recent work has questioned the accuracy of the Injury Severity Score (ISS) and the Abbreviated Injury Scale (AIS) in the pediatric population. We sought to determine mortality rates in pediatric trauma patients at ISSs considered "severe" in adults and whether mortality would vary substantially between adults and children sustaining injuries with the same AIS. Methods: Univariate logistic regression was used to generate mortality rates associated with ISS scores, for children (<16 years of age) and adults, using the 2016 National Trauma Data Bank. Mortality rates at an ISS of 15 were calculated in both groups. We similarly calculated ISS scores associated with mortality rates of 10%, 25%, and 50%. Receiver operating characteristic curves were constructed to compare the discriminative ability of ISS to predict mortality after blunt and penetrating injuries in adults and children. Mortality rates associated with 1 or more AIS 3 injuries per body region were defined. Results: There were 855, 454 cases, 86, 414 (10.1%) of which were children. The ISS associated with 10%, 25%, and 50% mortality were 35, 44, and 53, respectively, in children; they were 27, 38, and 48 in adults. At an ISS of 15, pediatric mortality was 1.0%; in adults, it was 3.1%. A 3.1% mortality rate was not observed in children until an ISS of 25. On receiver operating characteristic analysis, the ISS performed better in children compared with adults (area under the curve, 0.965 vs 0.860 [ P < 0.001]). Adults consistently suffered from higher mortality rates than did children with the same number of severe injuries to a body region, and mortality varied widely between specific selected AIS 3 injuries. Conclusions: Although the ISS predicts mortality well, children have lower mortality than do adults for the same ISS, and therefore, the accepted definition of severe injury is not equivalent between these 2 cohorts. Mortality risk is highly dependent on the specific nature of the injury, with large variability in outcomes despite identical AIS scores.