Abstracts

Rationale: Adverse events in the Epilepsy Monitoring Unit (EMU) occur in 7% of admissions. Surveys assessed the range of safety measures, and accreditation boards recommended best practices to mitigate morbidity and mortality. However, prospective assessments of interventions are lacking. We propose that patient position after a seizure is a surrogate marker of safety because prone position may place patients at risk of apnea and active/upright positions facilitate falls. The role of active intervention (bystander or nurse interaction with patient), push-button activation, and remote surveillance on the rate of body position change was assessed as a surrogate marker for patient safety. The effect of active interventions was also evaluated against the postictal nadir of heart rate (PINHR), a marker for postictal parasympathetic response.

Methods: 102 seizures from 42 patients admitted to the EMU between 02/2018–08/2019 were analyzed. Video and annotations were reviewed in 30-second epochs from (–)5 minutes to (+)10 minutes after seizure onset and scored for body position, bystander actions, and monitor technician awareness. The primary outcome was the effect of active intervention, push-button response, or monitor watcher response on the duration of position change timed from seizure onset to first change and PINHR. Cox proportional regression hazard analysis was used with seizure type (major motor vs. other) as a covariate. PINHR vs intervention group was tested with t-test.

Results: Changes in body position occurred in 13% (N=13) of seizures within 30 seconds after onset, and in 21% (N=21) of seizures within 5 minutes of onset. Active intervention doubled the rate of position change (OR 2.3, p=0.006, 95% CI 1.28 - 4.16). Similar results were seen with the presence of a push button event (1.8, 0.01, 95% CI 1.13 - 3.02) or a monitor watcher event (1.8, 0.02, 95% CI 1.09 - 3.02). In all, interventions of any type halved the time of the surrogate measure regardless of seizure type. The mean PINHR of seizures that were accompanied by active intervention (76.4 ± 12.8 BPM) was higher than seizures which did not receive intervention (69.7 ± 12 BPM, p=0.002).

Conclusions: Active interventions and seizure witnesses who activate clinical alarms potentially improve patient safety as determined through patient positioning. The results remained highly significant regardless of the presence of obvious motor symptoms. We conclude that active intervention shortens the position change time following a seizure and may indicate that interventions lessen patient risk. Active intervention showed a decreased propensity for physiologic deterioration (post-ictal bradycardia). More prospective studies assessing the physiologic impact of common EMU safety practices are needed to validate their use and help standardize and justify potentially expensive EMU protocols.

Funding: Please list any funding that was received in support of this abstract.: None.