Abstracts

Rationale: Prognostication in children after cardiac arrest is critical to guiding management decisions, yet accurate prediction of neurological outcome poses a challenge to clinicians. Several prognostic indicators have been described in adults1; however, few studies substantiate their value in children. For instance, neurologic exam findings are not considered predictive of outcome in adults within the first 24 hours of cardiac arrest1. However, it is not clear if early neurologic examination provides useful prognostic information in children. Moreover, while EEG is a valuable, noninvasive tool for assessing cortical function, there is uncertainty as to its most efficient utilization for adequate detection of seizures and prognostication after anoxic injury. Some have suggested that early EEG may not be es­sential for detection of most seizures2, yet others claim that the first hour of EEG recording may be the most prognostically significant3. This study seeks to further characterize the association between various clinical and EEG findings and neurological outcome. Methods: This tertiary, pediatric, single-center, retrospective chart review analyzes 41 patients who had cardiac arrests from March, 2011 to January, 2015. Various clinical variables were collected (table 1). Pediatric cerebral performance category scale (PCPC) was determined by evaluation of prior medical record and neurologic status at discharge. Neurological outcome was dichotomously categorized as favorable (increase in PCPC 1) or poor (increase in PCPC > 1). EEG findings were gathered from a hospital EEG database. Results: Of all clinical variables analyzed, only the following were predictors of poor outcome (table 1): duration of CPR (p = 0.0129), out-of-hospital arrest (p = 0.0049), arterial pH (0.0142), and lactate (0.0037), lack of pupillary response (p < 0.0001), absent motor response to pain (p < 0.0001), and absent brainstem reflexes (p = 0.0001). 12 patients had absent pupillary responses within 12 hours of cardiac arrest and all had poor outcomes. Only EEG background suppression (p = 0.0046) was associated with poor outcome. 9 patients had seizures, of whom 7 had poor outcomes. The majority of seizures began within the first 2 days following cardiac arrest; however, one patient had a first recorded seizure on day 5 and another first exhibited seizures on day 7 (figure 1). 2 patients experienced myoclonic status epilepticus and both had poor outcomes. Conclusions: Absent pupillary response, absent motor response to pain, and absent brainstem reflexes were all invariably associated with poor outcome. Notably, absent pupillary response within 12 hours of arrest was invariably associated with poor outcome, suggesting that early neurologic examination may be prognostically significant in children. Background suppression on EEG was significantly associated with poor outcome. Most seizures began within 48 hours of cardiac arrest, but seizure onset varied from Funding: none