To examine neurophysiologic predictors and outcomes of patients with late awakening following cardiac arrest.
Observational cohort study.
Academic ICU.
Adult comatose cardiac arrest patients treated with targeted temperature management and sedation.
None.
Time to awakening was calculated starting from initial sedation stop following targeted temperature management and rewarming (median 34 hr from ICU admission). Two-hundred twenty-eight of 402 patients (57%) awoke: late awakening (> 48 hr from sedation stop; median time to awakening 5 days [range, 3-23 d]) was observed in 78 subjects (34%). When considering single neurophysiologic tests, late awakening was associated with a higher proportion of discontinuous electroencephalography (21% vs 6% of early awakeners), absent motor and brainstem responses (38% vs 11%; 23 vs 4%, respectively), and serum neuron specific enolase greater than 33 ng/mL (23% vs 8%; all p < 0.01): no patient had greater than 2 unfavorable tests. By multivariable analysis-adjusting for cardiac arrest duration, Sequential Organ Failure Assessment score, and type of sedation-discontinuous electroencephalography and absent neurologic responses were independently associated with late awakening. Late awakening was more frequent with midazolam (58% vs 45%) and was associated with higher rates of delirium (62% vs 39%) and unfavorable 3-months outcome (27% vs 12%; all p = 0.005).
Late awakening is frequent after cardiac arrest, despite early unfavorable neurophysiologic signs and is associated with greater neurologic complications. Limiting benzodiazepines during targeted temperature management may accelerate awakening. Postcardiac arrest patients with late awakening had a high rate of favorable outcome, thereby supporting prognostication strategies relying on multiple rather than single tests and that allow sufficient time for outcome prediction.