Abstracts

Rationale: To determine if we can predict outcomes based on EEG patterns in patients with anoxic brain injury who are on therapeutic hypothermia. Targeted hypothermia after cardiac arrest is now standard of care at most institutions across the country and Europe. The current AAN guidelines regarding prognostication of neurological recovery after cardiac arrest at times do not apply to the hypothermic patient populations in entirety. Methods: A retrospective chart analysis of all patients 18 years and older, who were part of the hypothermia after cardiac arrest study was performed. Continuous EEG monitoring was performed and we gathered data on their EEG, neurological examination findings and clinical performance score [CPC] at discharge. Results: We tested the association between presence or absence of brain stem reflexes and discharge CPC score. There was not a statistically significant association; however there was a trend towards significance: patients who had preserved brain stem reflexes had a 1.72 times higher odds of having a good CPC outcome. In patients with EEG patterns considered to be normal to benign on day 1, the crude odds ratio for a good outcome was 31.38 [ 95% CI (0.89-3.36) with p<0.0001]. Similarly, for day 2 the crude odds ratio was 42.85 (95% C.I. [14.49, 126.76]) and for day 3, the crude odds ratio was 24.36 (95% C.I. [8.00, 74.20]) Conclusions: There are multiple factors that affect the neurological examination and hence relying on these for prognostication might have a high false positive rate. The EEG characteristics can help predict CPC score and outcomes at discharge with good reliability.