Predictors of Mortality Following Traumatic Cardiopulmonary Resuscitation in Pediatric Patients

Background: Pediatric trauma patients developing pre-hospital cardiac arrest have a dismal prognosis; few survive, often with severe neurologic deficits. Withholding or terminating cardiopulmonary resuscitation (CPR) for an injured child can be very difficult. Therefore, injured children may be subjected to protracted, futile CPR attempts which may consume vital resources beneficial to other patients with survivable injuries. This study was conducted to better define mortality in pediatric trauma patients receiving CPR and identify predictors of mortality that may guide decisions to withhold or terminate CPR in injured children. Methods: Pediatric (≤18 years) trauma patients who presented to Cook Children’s Medical Center from Jan. 2006 – Dec. 2017 and received CPR in the pre-hospital or emergency room setting were included: 88 patients met these criteria. Variables studied included the total time of CPR performance (≥15 vs. minutes), type of underlying cardiac rhythm (possibly perfusing vs. non-perfusing), the best recorded Glascow Coma Score (GCS; =3 vs [greater than] 3), and pupil reactivity (reactive vs. not reactive) present during the course of CPR. Fisher’s exact test was used to determine whether these variables were associated with survival. A p-value Results: Mortality in the 88 patients was 92%. Six of the 7 survivors had moderate (n=4) to severe (n=2) disability. Median total CPR time in the survivors was 5 minutes; only 2 surviving patients had CPR ≥15 minutes. The only survivor with no neurologic sequelae suffered a drowning event and had pre-hospital CPR for 5 minutes. Mortality following CPR performed only in the pre-hospital setting was associated with non-perfusing rhythm (p=0.001), GCS=3 (p=0.034), and CPR ≥15 minutes (p=0.022); pupil reactivity was not statistically associated with mortality. When CPR was performed in either the prehospital or emergency department setting, nonreactive pupils also were associated with mortality (p=0.003). There were no survivors who had CPR ≥15 minutes and a non-perfusing rhythm during the entire period of CPR. Conclusion: The results support withholding or terminating resuscitation in pediatric trauma patients who have received longer than 15 minutes of CPR during which time they had a non-perfusing rhythm, either PEA or asystole, and no evidence of neurologic function (GCS=3).