SAN DIEGO – Following in the wake of a previous prospective comparison of ketamine and haloperidol for control of severe prehospital agitation, a new prospective study by the same group and presented at the annual meeting of the American College of Emergency Physicians found ketamine superior to midazolam for the same indication.
“The difference between the two drugs was larger when agitation was more profound,” reported Jon B. Cole, MD, an emergency medicine physician at Hennepin Healthcare and associate professor of emergency medicine at the University of Minnesota, Minneapolis.
Relative to ketamine (Ketalar), midazolam (Versed) was associated with fewer side effects, “so for less agitated patients, midazolam may still be the preferred therapy,” he added in his late-breaker presentation.
In recent years, ketamine has become a popular drug in emergency medical services for control of agitation, according to Dr. Cole, but he said that this approach has been adopted with relatively limited support from objective evidence. In his literature search, only 2 of 11 original studies that addressed ketamine for agitation involved prospective comparative studies.
One of those prior prospective studies was one Dr. Cole led and published 2 years ago (Clin Toxicol. 2016;54:556-62). In that study, comparing 5 mg/kg of intramuscular (IM) ketamine to 10 mg IM haloperidol, ketamine had a faster onset (median 5 vs. 17 minutes) but produced more side effects, including higher rates of intubation (39% vs. 4%).
In the new prospective study, ketamine and midazolam were compared over separate consecutive 6-month periods in which ketamine and then midazolam were employed as the dominant strategy for agitation control. This was the same open-label, nonrandomized design used for the comparison of ketamine and haloperidol, but with one difference. When the Altered Mental Status Score was +2 or +3, considered severe agitation, patients received 3 mg/kg of IM ketamine or 5 mg of IM midazolam. When the AMS score was +4, called profound agitation, the doses were 5 mg/kg and 15 mg, respectively.
The primary result was that adequate sedation overall was achieved in a median 4.3 minutes on ketamine but 8.8 minutes on midazolam, producing a more than 3-minute advantage for ketamine, which translated into an odds ratio (OR) of 1.8 for adequate agitation control favoring ketamine, according to Dr. Cole. In those with severe agitation, the median advantage was less than 2 minutes, but the OR of 1.6 remained significant. In those with profound agitation, the mean difference climbed above 5 minutes with an OR of 2.5.
but Dr. Cole did report that midazolam was better tolerated. On the lower doses of the two drugs, for example, adverse events that were more common on ketamine than midazolam included vomiting (7% vs. 1%) and hypersalivation (20% vs. 0%).
In addition to the nonrandomized design, one limitation was unequal numbers of patients in the comparative groups. Only 113 were treated with midazolam while 202 patients were treated with ketamine. The reason was that the study of these drugs, which had been in widespread use at Dr. Cole’s institution, was conducted without asking patients to agree to participate. When negative stories in local papers framed this as a study conducted without consent, the institution asked the investigators to halt the study, and they complied.
“Subsequently, we have been assured after multiple evaluations that the rights and safety of our patient population were never violated,” Dr. Cole said, but he acknowledged that closing the study was an appropriate step in order to preserve patient confidence. “What we learned from this experience is that we may need to reconsider how we do waive of consent research at our institution,” he added.
In the meantime, the comparisons of ketamine with haloperidol and midazolam have provided objective evidence of their relative efficacy and safety in the management of prehospital agitation.
Dr. Cole reported no conflicts of interest.
SOURCE: Ann Emerg Med. 2018 Oct. doi. org/10.1016/j.annemergmed.2018.08.007.