Ketamine plus midazolam compared to midazolam infusion for the management of refractory status epilepticus

Data for the use of ketamine (Ket) in treatment of refractory and super-refractory status epilepticus (RSE, SRSE) is lacking despite its widespread growing use. We examined the efficacy of ketamine plus midazolam (MDZ) infusions for treating RSE versus midazolam alone. We hypothesized that ketamine initiation would result in earlier seizure termination.

Data was obtained from electronic health records (EHR) of adult patients who received intravenous anesthetic agents for RSE in our neurointensive care unit. Two cohorts were identified. The MDZ cohort received midazolam as the only intravenous anesthetic agent for RSE. The Ket+MDZ cohort received midazolam infusion followed by ketamine infusion. The primary outcomes were time from midazolam infusion start to SE end in both cohorts, and time from ketamine infusion start (Ket Start) to SE end in the Ket+MDZ cohort versus midazolam infusion start (MDZ start) to SE end in the MDZ cohort.

73 patients were included (MDZ cohort n=17, Ket + MDZ cohort n=56). Cohorts did not differ significantly in age, sex, race, RSE etiology, or GCS on admission. Mean APACHE II score was higher in the Ket +MDZ cohort (26 ± 7.32 SD) versus the MDZ cohort (22 ± 5.89 SD)(P=.015). In survival analyses, cohorts did not differ significantly in time from midazolam start to SE end (HR=0.965, 95 % CI=0.556-1.673, P=.899; median [IQR]: MDZ: 25 h [4.5-58]; Ket+MDZ: 21.5 h [IQR 13.5-49]). Time from Ket start (Ket+MDZ group) versus time from MDZ start (MDZ group) to SE end was significantly shorter in the Ket+MDZ cohort (HR=1.895, 95 % CI=1.083-3.314, P=.025). The pattern of results was similar when including APACHE II and MDZ maximum dosage as covariates.

Time to SE end was significantly shorter after addition of ketamine infusion to midazolam infusion, versus after initiation of midazolam infusion monotherapy. Patients with higher disease severity favored Ket+MDZ. Randomized controlled trials are warranted in determining optimal anesthetics in RSE and SRSE.