Neurologic prognosis and withdrawal of life support after resuscitation from cardiac arrest

Objective: To study the impact of neurologic prognostication on the decision to withdraw life-sustaining therapies (LST) in comatose patients resuscitated after cardiac arrest. Methods: The authors prospectively studied a consecutive series of post-resuscitation comatose patients referred for neurologic prognostication at a single center for 4 years. For most patients, neurologic prognostication was not sought due to early death or rapid return to consciousness. Prognostication was based on Glasgow Coma Score (GCS) and Brainstem Reflex Score (BRS), with EEG and cortical evoked potentials (CEP), which were graded as benign, uncertain, and malignant. The outcomes were as follows: survivors (Group S), brain or cardiac death (Group D), and death from withdrawal of life sustaining therapy (Group W). In Group W, the time interval to withdrawal of LST was analyzed by EEG and CEP grades. Results: Of 58 patients studied, 10 were in Group S, 8 in Group D, and 40 in Group W. Initial median GCS and BRS was similar for all groups with significant improvement noted in Group S, but not in Group D or Group W. In Group W, CEP grade correlated with the median duration of continued therapy before a decision to withdraw LST: 7 days for benign CEP, 2 days for uncertain CEP, and 1 day for malignant CEP, p = 0.0004. Conclusion: In patients with poor neurologic recovery early after resuscitation from cardiac arrest, physicians appear to use the cortical evoked potential grade to estimate prognosis. Cortical evoked potential grade correlated with the waiting time until life sustaining therapies were withdrawn after no improvement in neurologic examination was seen.