Efficacy of nonvenous medications for acute convulsive seizures: A network meta-analysis

Abstract

Objective: This is a network meta-analysis of nonvenous drugs used in randomized controlled trials (RCTs) for treatment of acute convulsive seizures and convulsive status epilepticus.

Methods: Literature was searched according to Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines for RCTs examining treatment of acute convulsive seizures or status epilepticus with at least one of the study arms being a nonvenous medication. After demographic and outcome data extraction, a Bayesian network meta-analysis was performed and efficacy results were summarized using treatment effects and their credible intervals (CrI). We also calculated the probability of each route-drug combination being the most clinically effective for a given outcome, and provided their Bayesian hierarchical ranking.

Results: This meta-analysis of 16 studies found that intramuscular midazolam (IM-MDZ) is superior to other nonvenous medications regarding time to seizure termination after administration (2.145 minutes, 95% CrI 1.308-3.489), time to seizure cessation after arrival in the hospital (3.841 minutes, 95% CrI 2.697-5.416), and time to initiate treatment (0.779 minutes, 95% CrI 0.495-1.221). Additionally, intranasal midazolam (IN-MDZ) was adjudged most efficacious for seizure cessation within 10 minutes of administration (90.4% of participants, 95% CrI 79.4%-96.9%), and persistent seizure cessation for ≥1 hour (78.5% of participants, 95% CrI 59.5%-92.1%). Paucity of RCTs produced evidence gaps resulting in small networks, routes/drugs included in some networks but not others, and some trials not being connected to any network.

Conclusions: Despite the evidence gaps, IM-MDZ and IN-MDZ exhibit the best efficacy data for the nonvenous treatment of acute convulsive seizures or status epilepticus.

Figure 1. Bayesian network representations for individual outcomes

The size of the bubble represents the total number of participants randomized to receive a particular route–drug combination. The number by the side of a line joining 2 bubbles indicates the number of trials for that particular comparison (A.a: seizure cessation within 10 minutes of drug administration, A.b: seizure cessation within 5 minutes of drug administration, A.c: persistent seizure cessation for at least 1 hour after drug administration, B.a: time from drug administration to termination of seizure [minutes], B.b: time to cessation of seizures after arrival in hospital [minutes], B.c: time to initiate treatment [minutes]) (reference numbers in square brackets). BC = buccal; DZP = diazepam; IM = intramuscular; IN = intranasal; LOR = lorazepam; MDZ = midazolam; PR = rectal; SL = sublingual.

Figure 2. Forest plot and funnel diagrams for BC-MDZ and PR-DZP head-to-head comparison

Forest plot (A) for conventional head-to-head comparison between buccal (BC) midazolam (MDZ) and rectal (PR) diazepam (DZP) for seizure cessation within 10 minutes of drug administration shows minimal heterogeneity and similar relative effect compared to network meta-analysis. Funnel plots without (B) and with (C) a pseudostudy (open circle, created using trim-and-fill method to adjust for funnel plot asymmetry) shows no relevant influence on relative treatment effect of the publication bias. CI = confidence interval; OR = odds ratio.