P2A231: Initial End-tidal Carbon Dioxide: A Novel Evaluation for the Pediatric Trauma Patient in Shock

Background: Reliable early recognition of severely injured children in the prehospital setting remains challenging. ETCO2 has a specific range of standardized values and does not vary based on a patient’s age. We hypothesized that initial ETCO2 would predict mortality and the need for transfusion in traumatically injured children as well as perform comparably to other measures of shock such as shock index, pediatric age-adjusted and reverse shock index multiplied by Glasgow Coma Scale (rSIG).

Methods: A retrospective review was conducted from 2012-2022 of trauma patients ages 0-18 admitted to a quaternary children’s hospital. Patients were divided into two groups: ≥ 4 years of age < 4 years of age to allow for comparison with heart rate (HR) and systolic blood pressure (SBP). SIPA has only been validated in patients ≥ 4 years age. Variables collected included: ETCO2, SIPA, rSIG, HR, SBP, mortality, and need for blood transfusion. Receiver operating characteristic curves were created and used for comparison.

Results: A total of 308 patients were available for analysis: 200 patients ≥ 4 years of age and 108 patients < 4 years of age. For children ≥ 4 years of age, ROC curves utilizing rSIG , ETCO2, and SIPA as predictors of mortality yield respective AUCs of 0.725, 0.665, and 0.529. Pairwise comparison of the AUC corresponding to the ETCO2 ROC curve with AUCs of the rSIG and SIPA curves were not statistically significant (p = 0.319 [ETCO2 vs SI], p=0.505 [ETCO2 vs rSIG]). For children ≥ 4 years of age, ROC curves utilizing ETCO2, rSIG, and SIPA as predictors of transfusion resulted in AUROCs of 0.669, 0.625, and 0.614 respectively. Pairwise comparison of the AUC were not statistically significant (p = 0.686, [ETCO2 vs SI], p=0.550 [ETCO2 vs rSIG]). For < 4 years of age, ETCO2 outperformed heart rate (HR) and systolic blood pressure (SBP) and was predictive of mortality with AUROC of 0.812, 0.706, and 0.626, respectively. Pairwise comparison of the AUC corresponding to the ETCO2 ROC curve with AUCs of the SBP and HR curves was statistically significant (p = 0.022, [ETCO2 vs HR], p = 0.193 [ETCO2 vs SBP])/ Finally, no variable was helpful in determining transfusion need for children < 4 years of age (AUROC’s of 0.579 HR, 0.509 ETCO2, and 0.531 SBP, p = 0.736, [ETCO2 vs HR], p=0.858 [ETCO2 vs SBP].

Conclusion: ETCO2 is an adequate marker of shock in pediatric trauma patients. It is predictive of mortality in young children. ETCO2 has a standardized range that does not vary by age and can be measured non-invasively. ETCO2 as a marker for mortality and transfusion needs to be validated prospectively.