Investigators from the Neurological Emergencies Treatment Trials (NETT) network conducted a randomized, double-blind trial to determine whether intramuscular (IM) midazolam is as effective as intravenous (IV) lorazepam for the termination of status epilepticus in the prehospital setting. The noninferiority trial (named RAMPART) involved over 4,000 paramedics from 33 emergency medical services (EMS) agencies and 79 receiving hospitals throughout the United States. (For an explanation of noninferiority trials, see AAP Grand Rounds, September 2011;27:26). Eligible study participants were children with an estimated weight of at least 13 kg and adults who were experiencing a convulsive seizure (continuous or intermittent without regaining consciousness) for at least 5 minutes at the time of treatment by paramedics.
The study procedure involved initial stabilization according to local EMS protocols, followed by opening of an instrumented box that contained a study drug kit, prepared and assigned by a central pharmacy. Each patient was treated with an IM autoinjector that had been filled with active treatment or placebo, followed by IV access and delivery of placebo or active treatment. Children weighing 13 to 40 kg received active treatment with either 5 mg of IM midazolam or 2 mg of IV lorazepam, and participants who weighed >40 kg received active treatment with either 10 mg of IM midazolam or 4 mg of IV lorazepam. Opening of the instrumented box activated a voice recorder that instructed paramedics to make oral statements about drug administration and seizure activity, which were time stamped. Paramedics were directed to attempt IV (or intraosseous) access for at least 10 minutes. When seizure activity continued for 10 minutes after the last study medication was administrated, rescue therapy (per local EMS protocol) was recommended. The primary outcome of the study was seizure termination prior to arrival in the emergency department (ED) without the need for rescue therapy. Both intention-to-treat and per-protocol analyses were conducted.
A total of 2,145 patients were assessed for eligibility; 893 were enrolled, randomized, and assigned to a treatment group. Characteristics of study participants in the two treatment groups were similar at baseline. In the intention-to-treat analysis, 73.4% of subjects randomized to the IM midazolam group had termination of their seizure without rescue therapy, compared to 63.4% of those randomized to receive IV lorazepam. This difference of 10 percentage points (95% CI, 4.0–16.1) was significant for both noninferiority and superiority (P < .001). Similar results were obtained in the per-protocol analysis. The frequencies of intubation and recurrent seizures in the two study groups were similar. Subjects randomized to the IM midazolam group were more likely to be discharged from the ED than those who were randomized to receive IV lorazepam (OR=0.88; 95% CI, 0.79–0.98).
The authors conclude that IM midazolam is at least as safe and effective as IV lorazepam for...