Background
Paediatric generalised convulsive status epilepticus (GCSE) are some thing we see a fair amount of in Virchester. Many are a result of febrile seizures, but as we are a tertiary centre we see the full range of causes presenting to PED. Seizures that fail to stop spontaneously within five minutes are much less likely. to terminate without treatment and may lead to progressive neuronal injury, worsening outcomes, and, in some cases, death. The longer a child seizes, the more resistant the seizure becomes to medication, a phenomenon linked to dynamic changes in synaptic receptor physiology and neurotransmitter balance. In Virchester we have
Current international guidelines recommend benzodiazepines (BDZs) as the first-line treatment for GCSE in both adults and children. In the UK, midazolam has become the workhorse benzodiazepine because of its flexibility — it can be given intravenously, intranasally, buccally, or intramuscularly. These agents act as positive allosteric modulators of the GABAA_A receptor, enhancing inhibitory signalling in the central nervous system.
However, benzodiazepines are not perfect. Around 30–40% of children with GCSE are refractory to the initial dose, and pharmacoresistance (love this word – Ed) increases rapidly with time. If seizures persist, current practice is to escalate through stepwise protocols, adding second-line agents such as levetiracetam, phenytoin, or valproate. But beyond that, options become limited, and management transitions into the territory of refractory and super-refractory status epilepticus, where evidence is sparse and treatments often involve induction of anaesthesia.
Ketamine has been attracting increasing attention as a possible adjunct or alternative. It works via an entirely different mechanism, acting as a non-competitive NMDA receptor antagonist and thereby targeting the excitatory side of the balance, rather than inhibitory pathways. Preclinical studies in animal models have consistently demonstrated that ketamine can reduce seizure severity and even has neuroprotective effects in prolonged seizures. Clinically, we already use ketamine widely in emergency medicine for procedural sedation, analgesia, and as an induction agent for rapid sequence intubation. Its safety profile is well established in those contexts, and it is increasingly used in refractory and super-refractory status epilepticus, where small case series suggest success rates of around 70%.
What has been missing until now is randomised controlled trial data testing ketamine’s role in the early management of GCSE, as opposed to its use as a rescue therapy when everything else has failed. That’s where the Ket-Mid study, published in Paediatric Neurology in 2025, comes in. This is the first RCT investigating whether adding ketamine to midazolam as first-line therapy improves seizure control in children presenting with GCSE.
The Skeptics’ Guide to Emergency Medicine (SGEM) covered this paper in episode #482 and provided an excellent podcast review. Their discussion informed much of this post, and we’d highly recommend subscribing if you want deeper dives into the evidence base.
Abstract
Background: Approximately one third of children with generalized convulsive status epilepticus (GCSE) are not controlled by initial benzodiazepine therapy. We investigated the efficacy of adding ketamine to midazolam for first-line treatment of pediatric GCSE.
Methods: This randomized controlled trial included 144 children with GCSE aged between six months and 16 years, who were equally randomized to receive ketamine plus midazolam (Ket-Mid group) or placebo plus midazolam (Pla-Mid group). Primary outcome was cessation of clinical seizures at five-minute study timepoint. Secondary outcomes were the need for a second midazolam bolus; cessation of clinical seizures at 15-, 35-, and 55-minute timepoints; 24-hour seizure control; and adverse effects.
Results: Cessation of clinical seizures at five-minute occurred in 76% of children in the Ket-Mid group compared with 21% in the Pla-Mid group (risk ratio [RR] 3.7; 95% confidence interval [CI] 2.3-5.9; P < 0.001). Compared with the Pla-Mid group, the Ket-Mid group had higher percentages of seizure cessation at 15-minute (76.4% vs 23.6%; RR, 3.2; 95% CI, 2.1-5.0), 35-minute (83.3% vs 45.8%; RR, 1.8; 95% CI, 1.4-2.4), and 55-minute (88.9% vs 72.2%; RR, 1.2; 95% CI, 1.04-1.45) study timepoints as well as lower percentages of repeating midazolam (23.6% vs 79.2%; RR, 0.3; 95% CI, 0.19-0.46) and endotracheal intubation (4.2% vs 20.8%; RR, 0.2; 95% CI, 0.06-0.66). Both groups showed no significant differences in other outcome measures.
Conclusions: Ketamine-midazolam combination may be more effective than midazolam alone for the initial treatment of pediatric GCSE, but this should be confirmed in future research.
What Kind of Study Is This?
The Ket-Mid trial was a randomised, double-blind, placebo-controlled superiority trial conducted at Sohag University Hospital in Egypt between March 2023 and August 2024. Children aged between six months and 16 years presenting to the paediatric emergency department with GCSE were randomised in a 1:1 ratio to receive either intravenous ketamine (2 mg/kg) plus midazolam or placebo plus midazolam.
Randomisation was concealed, and all clinicians, families, investigators, and outcome assessors were blinded to allocation. The trial was prospectively registered on ClinicalTrials.gov, and the methodology adhered closely to CONSORT guidelines.
This was a single-centre study in a middle-income country, an important contextual factor when thinking about how these findings might (or might not) translate into other healthcare systems, particularly those with different prehospital arrangements and earlier access to treatment.
Tell Me About the Patients
A total of 144 children were enrolled, with 72 randomised to each arm. Baseline characteristics were well balanced between groups.
The median age was just over 31 months, with roughly equal numbers of boys and girls. The most common causes of GCSE were febrile seizures (43%) and known epilepsy (42%), followed by central nervous system infections (11%) and other or unknown causes (4%).
One striking feature of this cohort is the delay to treatment. The median seizure duration prior to hospital arrival was 34 minutes, and nearly 60% of the children had been seizing for more than half an hour before receiving any medication. None of the patients had received prehospital benzodiazepines or any other antiseizure medications before arrival — a stark contrast to many UK systems, where early treatment by paramedics or families is common.
The trial used strict inclusion and exclusion criteria. Children were excluded if they had prior treatment with antiseizure medication, significant comorbidities such as traumatic brain injury or organ failure, or if IV access couldn’t be established within five minutes. While these criteria improve internal validity, they may limit generalisability and could introduce selection bias, especially since some exclusions (for example, pheochromocytoma or hyperthyroidism) would be challenging to identify rapidly in emergency settings.
What Were the Outcomes?
The primary outcome was cessation of clinically visible seizure activity five minutes after administration of the allocated treatment.
Secondary outcomes included seizure cessation at 15, 35, and 55 minutes; need for a second dose of midazolam; seizure control at 24 hours; adverse events such as hypotension, arrhythmias, and emergence phenomena; need for intubation; and mortality.
A potential limitation from some commentators is that EEG monitoring was not used. Seizure cessation was assessed clinically, meaning that ongoing subclinical seizure activity could not be detected, but that refelcts clinical prractice. We don’t have EEG monitoring in the ED and therefore irt'[s the same outcome that I would used here. The authors also did not report whether patients returned to their neurological baseline, which makes interpretation of ultimate clinical recovery more difficult.
What Are the Main Results?
At five minutes, 76% of children receiving ketamine plus midazolam had stopped seizing compared with 21% of those receiving midazolam alone. This equates to a risk ratio of 3.7 (95% CI 2.3–5.9, p<0.001). So statiustically and arguably also a clinically important difference. The benefit persisted at later time points. At 15 minutes, seizure cessation was seen in 76% of the ketamine group versus 24% in the control group; at 35 minutes, 83% versus 46%; and at 55 minutes, 89% versus 72%. So same story later on and persists.
The ketamine group required fewer repeat doses of midazolam (24% versus 79%) and significantly fewer were intubated (4% versus 21%). Mortality was low overall (1.4%), and adverse events were rare and similar between groups.
Subgroup analysis showed that ketamine’s effect was greatest in children who had been seizing for more than 30 minutes prior to arrival, which made up around 60% of the cohort.
Can we believe the results?
Randomisation and blinding were robust, and the methodology was pretty sound. The effect size for the primary outcome was large and statistically significant, which adds weight to the findings. However, the relevance to my practice is arguably limited. Ketamine is a controlled drug in the UKJ, so I don’t think we will see it handed out to families anytime soon, similarly it’s a legal problem for the ambulance service too where specific training is required to use ketamine (it’s not impossible, but it’s tricky).
A lot of patients were excluded including many of the type that we see in PED, so this may represent a highly selected subgroup of children with GCSE. The authors excluded children with significant comorbidities, complex metabolic conditions, or delayed IV access, which are often the children we struggle with most in clinical practice. SGEM noted that some of these exclusions would be difficult to achieve in an acute setting.
Similarly, the median seizure duration before arrival was 34 minutes, and none of the children had received prehospital benzodiazepines. In UK and other high-resource settings, treatment is typically initiated far earlier, either by parents, paramedics, or community clinicians. So, to reiterate, this is simply a different group of patients to those I see in the ED. This difference matters because the pathophysiology of GCSE evolves over time. In early status, benzodiazepines are highly effective. As seizures progress, GABAA_A receptor trafficking reduces benzodiazepine sensitivity, while NMDA receptor upregulation increases excitatory signalling. Ketamine’s mechanism of action makes it particularly appealing in prolonged or benzodiazepine-refractory status, but its benefit may be less pronounced in systems where most patients are treated early.
We also need to be mindful that this was a single-centre trial conducted in a middle-income country with different healthcare infrastructure, medication availability, and treatment timelines compared to many high-resource systems. While the underlying biology is the same, differences in seizure duration, treatment pathways, and patient populations mean we should be cautious about directly applying these findings to UK practice.
Should We Change Practice Based on This Study?
No, well sort of no…..
In UK practice, the priority remains early administration of benzodiazepines. Giving effective doses rapidly intranasally, buccally, or intravenously is still the cornerstone of management. For those who remain refractory, escalation to second-line agents such as levetiracetam or phenytoin is supported by better evidence than we currently have for ketamine as part of first-line therapy.
That said, ketamine may have a role in benzodiazepine-refractory status and could be considered earlier than we currently do, especially in prolonged seizures . To change guidelines, though, we need larger, multicentre trials in diverse settings.
I supose the only thing that it does reassure me about is the choice of drugs for RSI in status. At the VAAC (Virchester Air Ambulance Charity) we have debated what induction agent to use for status patients. For those who use ketamine for most (if not all) patients, this study supports (but does not prove) their approach.
Final thoughts
The Ket-Mid study is an interesting study, but it is single centre and the patients are different to my own. I’m considering this trial to be hypothesis-generating rather than practice-changing.
For a deeper dive, listen to the SGEM podcast episode #482, which provides excellent discussion and analysis.
vb
S
References
- Othman AA, et al. Combined ketamine and midazolam vs. midazolam alone for initial treatment of pediatric generalized convulsive status epilepticus (Ket-Mid study): A randomized controlled trial. Pediatric Neurology. June 2025
- SGEM#482: Seize the Day with Ketamine and Midazolam for Pediatric Status Epilepticus
