Abstract 18242: Association Between Electrographic Seizure Suppression and Neurological Outcomes After Resuscitation From Cardiac Arrest
Introduction: Seizures after cardiac arrest are associated with worse outcomes, but there is no evidence that treatment improves outcomes. We leveraged existing practice variation to compare the effectiveness of aggressive electroencephalography (EEG) and antiepileptic drug (AED) use to infrequent spot EEG and AED use.
Methods: We included a cohort of comatose post-arrest patients at two academic medical centers from 2011-2014. The same critical care group staffs both centers, but Center 1 uses continuous EEG (cEEG) monitoring and aggressively treats malignant EEG patterns while Center 2 uses infrequent spot EEGs. We classified each patient’s daily EEG from admission until death, awakening or day 5 as “malignant” (status epilepticus, polyspikes or periodic discharges), “not malignant,” or “not performed.” We abstracted covariates and outcomes from our prospective registry: age, illness severity (Pittsburgh Cardiac Arrest Category), shockable rhythm, arrest location, survival to hospital discharge and functionally favorable survival (Cerebral Performance Category 1-2). We used generalized estimating equations to test for an interaction between center and EEG pattern predicting outcomes after adjusting for clinical covariates.
Results: We included 824 subjects (Center 1: 513, Center 2: 311). Center 1 subjects were younger, more likely to arrest out-of-hospital and had higher illness severity (all P<0.01). Overall, 366 (71%) Center 1 subjects were EEG-monitored (median 2d (IQR 1-4)), 196 (38%) had malignant patterns and 227 (44%) received an AED (median 2 per patient). In comparison, 155 (50%) Center 2 subjects were monitored a median of 1d (IQR 1-1), 62 (20%) had malignant patterns, and 89 (29%) received an AED (median 1 per patient) (P<0.01 for all between-center comparisons). Center 1 subjects with malignant patterns were more likely to have myoclonic status epilepticus (P<0.01) and overall had worse survival and outcomes (P=0.02). By contrast, subjects who were not EEG-monitored at Center 2 had worse outcomes (P<0.01). There was no evidence of a center effect on outcome.
Conclusion: Post-arrest cEEG monitoring and AED usage is not associated with improved outcomes at discharge. This appears to be influenced by selection bias at Center 2.
Author Disclosures: J. Faro: None. P.J. Coppler: None. C. Dezfulian: Research Grant; Significant; NIH grant K08NS069817. M. Baldwin: None. J.C. Rittenberger: None. C.W. Callaway: None. J. Elmer: Research Grant; Significant; NIH grant 5K12HL109068.
- © 2016 by American Heart Association, Inc.