Abstracts

Rationale:
The occurrence of sub-clinical seizures has been established in people with a variety of cerebral insults.  Continuous EEG (cEEG) is the gold standard for detection of such seizures.  There is higher risk of seizures in the presence of altered mental status (AMS) or coma.  However, there is no widely accepted standard for patient selection for cEEG.  Application of cEEG is likely widely discrepant across providers and medical centers.  For example, it has been reported that many cEEG are initiated after a clinical seizure has been witnessed. Such seizures occur in a minority of patients in whom seizures are detected by cEEG, suggesting this approach may lead to under-utilization of cEEG.  This study evaluated the utilization of cEEG at a large, academic medical center with continuous availability of cEEG.
Method:
This retrospective cohort single-center study was performed with approval by the IRB.  People at risk for seizure were defined as having intracranial hemorrhage (ICH) and stratified by AMS.  All patients with ICH who were admitted between July 2018 and December 2019 were identified, excluding people with code status “allow a natural death”.  AMS was determined using the Confusion Assessment Method (CAM) score, which is applied to all inpatients every nursing shift, and in which a positive score indicates AMS.  AMS was categorized as definitive (3 consecutive positive CAMs), probable (2 positive CAMs in 48 hours), possible (any non-negative CAM) and definitely not altered (no non-negative CAMs during admission).  The primary outcome variable was utilization of cEEG, calculated for total group as well as AMS strata.  Additional variables included age, level of care (ICU, step-down, other), primary service (neurology, neurosurgery, other), and mechanical ventilation.  Chi square tests were used to identify stratum specific differences in the cEEG utilization. Results514 patients met inclusion criteria.  Overall, 38 (7.4%) people with ICH had cEEG.  There was a trend towards increased cEEG on the neurology service (p=0.059).  AMS categories were as follows: 71 (13.8%) had definite AMS, 25 (4.9%) probable, 80 (15.5%) possible and 338 (65.8%) were definitely not altered.  People with definite AMS were significantly more likely to have cEEG (17 people (24%) p< 0.01).  Among people with ICH and within the subgroup with definite AMS, mechanical ventilation, level of care and age were not associated with significantly different likelihood of cEEG utilization.
Conclusion:
Definite AMS was associated with higher likelihood of cEEG application, and there was a trend towards increased cEEG on the neurology service.  However, overall utilization rate of cEEG of 7.4% among patients with ICH was lower than expected.  There may be patients with ICH with undetected sub-clinical seizures, which may impact outcome from ICH.  Further research should focus on improving patient selection factors for cEEG.
Funding:
:None