Abstracts

Rationale:
Electrographic seizures (ES) are common in critically ill children and a high seizure burden has been shown to be associated with neurological decline. Early treatment of ES decreases seizure burden and has been associated with greater likelihood of seizure termination. The process in which ES are treated can include multiple steps. These steps include identification of ES, the appropriate medication order, preparation and accessibility of the medication desired, and finally prompt administration. Delays in any of these steps result in treatment delays, and potential for higher seizure burden. We aimed to identify potential barriers to treatment focusing on interventions to improve them.
Method:
We assessed the timing of seizure identification, medication order type (STAT versus routine), and medication administration for patients admitted to the neonatal or pediatric ICU with EEG only or subtle clinical seizures identified by continuous EEG monitoring not recognized by the bedside provider. Patients with a pre-existing diagnosis of epilepsy were excluded. Results10 patients were identified from November 2019 to May 2020 with a median age of 39 days (IQR 204). Patients were admitted to the neonatal ICU (n=6), cardiac ICU (n=2), and medical/surgical ICU (n=2). Indications for EEG included detection of subclinical seizures in hypoxic ischemic encephalopathy, presumed CNS infection, cardiac arrest, and cerebrovascular disorder. Median time from 1) seizure onset to first medication ordered was 96 minutes (IQR 124 minutes) 2) medication order to administration was 43 minutes (IQR 38 minutes) and 3) seizure onset to medication administration was 166 minutes (IQR 149 minutes).  When looking at time of day, the median time from seizure onset to medication order was 74 minutes (IQR 49 minutes) from 8am-5pm compared to 120 minutes (IQR 135 minutes) from 5pm-8am. The median time from medication order to administration was 47 minutes (IQR 30 minutes) from 8am-5pm compared to 39 minutes (IQR 33 minutes) from 5pm-8am.  Medication orders were placed as STAT orders seventy percent of the time. Median time from medication order to administration was 31 minutes for STAT orders compared to 55 minutes for routine orders. For first line treatment, median time from seizure onset to medication administration was 28 minutes for lorazepam (n=1), 180 minutes for levetiracetam (n=5), and 107 minutes for phenobarbital (n=4). Phenobarbital and lorazepam were obtained from an automated medication dispensing system whereas levetiracetam was delivered from pharmacy.  
Conclusion:
Medication selection, order status, and delivery processes impact time to treatment.  Once ordered, medications are administered faster overnight which may reflect differences in resources depending on time of day. Standardizing medication ordering and delivery processes could improve treatment times which may improve seizure burden and subsequent outcome. 
Funding:
:N/A