Authors: Rey A. Rossetti AO, et al. Late Awakening in Survivors of Postanoxic Coma: Early Neurophysiologic Predictors and Association With ICU and Long-Term Neurologic Recovery. Critical Care Medicine. 2019 Jan 1;47(1):85-92.
Reviewed by: Wazim Mohamed, MD, Assistant Professor of Neurology, Division of Neurotrauma and Critical Care Neurology, Wayne State University
Over the last decade, survival from coma after cardiac arrest (CA) has approached 50% and up to 1/3 of survivors may experience late awakening, defined as time to awakening >48 hours after temperature management (TTM) and initial sedation interruption. Characteristics of arrest, TTM, use of sedatives and degree of organ dysfunction influence the time to awakening. The authors of this study investigated early neurologic and neurophysiological markers among patients who awoke from post-anoxic coma. They sought to establish a relationship between the degree of hypoxic-ischemic injury (as marked by neurologic and neurophysiologic parameters) and late awakening. They also quantified the degree of ICU delirium and examined the effects of late awakening on 3-month neurological outcome.
This was a single-center observational study, over 7 years, that included consecutive adult patients that were successfully resuscitated from CA. All patients were treated with TTM and sedation as per institutional protocol. Sedation/analgesia consisted of IV midazolam as first choice, propofol as second choice or in combination with midazolam. Following rewarming to 36°C for at least three hours, sedatives and analgesics were stopped. This was considered the start point to determine time to awakening. Clinical exam included Glasgow Coma Scale motor response (GCS-M) and brainstem reflexes (pupillary, corneal and oculocephalic) performed daily during the first 72-hours post CA. Brainstem reflexes were considered absent if one of these 3 reflexes was absent. EEG was performed on day 1 and day 2 (post TTM and off sedation) and categorized based on patterns (continuous, discontinuous or background suppression). Neuron-specific enolase (NSE) was sent within 72 hours.
Late awakening was defined as greater than 48 hours from initial sedation interruption, with patient following commands. Delirium was diagnosed based on RASS> 2 with the need for repeated antipsychotic use. Neurological recovery was assessed at 3 months via telephone using Cerebral Performance Category (CPC). CPC was dichotomized as favorable (CPC 1-2) and unfavorable (CPC 3-5). Decision to withdraw life-sustaining therapy was made at 72 hours based on prespecified algorithm. Patients having two of the following were considered to have a poor prognosis: absent pupillary and corneal reflexes, presence of myoclonus, nonreactive EEG background or bilaterally absent somatosensory evoked potentials (SSEP). Baseline patient characteristics (age, gender, CA etiology, initial rhythm, CA duration, Sequential Organ Failure Assessment [SOFA] score at day 1) were analyzed for univariate association between early versus late awakening. A multivariate logistic regression analysis was performed including all variables with a p value less than 0.20 on univariate analysis.
Four hundred and two patients were included in this study; 228 patients awoke and were included in the analysis. TTM was instituted at 33°C in 68% of patients and 36°C in the remaining 32%. Late awakening was seen in 34% of patients. Among all awakeners, 83% had a favorable 3-month outcome, 17% unfavorable outcome of which 7% died. On univariate analysis, late awakeners had significantly prolonged duration of CA (median 15 mins [IQR 10-27] vs 15 mins [IQR 10-20];p=0.02) and higher SOFA scores (p<0.01). Patients with late awakening also had significantly higher rate of absent GCS-M (38% vs. 11%; p<0.01), absent brainstem reflexes (23% vs. 4%; p<0.01), discontinuous EEG (48% vs. 21% at day 1; 21% vs. 6 % at day 2 [p<0.01 in both instances]) and higher peak serum NSE (p<0.01). Goal temperature for TTM was not associated with late awakening. Late awakening was associated with higher proportion of patient receiving midazolam, a longer duration of sedation and a higher cumulative dose of midazolam (all p<0.05).
On multivariate analysis to identify factors associated with late awakening, duration of cardiac arrest, (p=0.04), day 1 SOFA score (p=0.04), use of midazolam alone (p=0.05), day 2 discontinuous EEG (p=0.01), absence of GCS-M (p<0.01) and absent brainstem reflexes at day 3 (p=0.04) were significant.
Late awakening was also significantly associated with longer ICU stays and greater risk of delirium. On 3-month follow up, late awakeners had significantly higher likelihood of unfavorable outcomes (27% vs. 12%; p<0.01). However, even among the late awakeners, 73% of patients had a favorable outcome (CPC 1-2) at 3 months.
The authors conclude that late awakening after CA is common and is associated with severity of the post CA syndrome, increased utilization of midazolam, early neurologic and EEG signs of hypoxic ischemic brain injury, however a large percentage of late awakeners still had favorable outcome at 3 months. The late awakeners had unfavorable neurophysiological parameters at 72 hours. This brings into question the 72-hour prognostication time point, which may erroneously suggest patients have a poor prognosis when they in fact may go onto favorable recovery. The median time for late awakening was 5 (IQR 3-23) days. This study also elucidates the role of midazolam, length of sedation and cumulative dose of sedatives in late awakening post CA thereby promoting short acting agents or ‘moderate dose sedation’ strategies. The study limitations include its single-center nature. Additionally, as with most CA studies, there is potential bias based on which patients undergo withdrawal of life sustaining therapy. These findings provide additional support for delaying outcome prediction and using multiple prognostication modalities to assess recovery for post-anoxic coma.