Midazolam for Acute Seizures in Children: What a Review Found

What was studied

This systematic review and meta-analysis combined 23 randomized controlled trials involving 2,238 children with acute seizures. The trials compared midazolam given into the nose (intranasal) or between the cheek and gum (buccal) with benzodiazepine medicines given into a vein, into a muscle, rectally, or by mouth.

The researchers searched four databases through January 15, 2026. They combined results using a random-effects model, which allows for differences between studies. Outcomes included seizure stopping within 5 or 10 minutes, seizure recurrence within 60 minutes, treatment timing, treatment failure, and breathing-related side effects.

What they found

Across the comparison treatments, there was no clear difference in seizure stopping within 10 minutes (13 studies; 1,368 children; risk ratio [RR] 1.04, 95% confidence interval [CI] 0.99–1.09) or within 5 minutes (7 studies; 552 children; RR 1.05, 95% CI 0.92–1.19).

Compared specifically with rectal diazepam, nasal or buccal midazolam improved seizure stopping within 10 minutes (6 studies; 904 children; RR 1.18, 95% CI 1.02–1.37). Results were similar to diazepam and lorazepam given into a vein.

Nasal or buccal midazolam reduced seizure recurrence within 60 minutes (7 studies; 791 children; RR 0.62, 95% CI 0.43–0.91). It was also given faster (7 studies; 546 children; standardized mean difference [SMD] −1.94, 95% CI −3.16 to −0.72) and shortened the overall time to seizure stopping (6 studies; 379 children; SMD −0.91, 95% CI −1.64 to −0.18). These standardized timing results cannot be read directly as minutes saved.

There were no statistically significant differences in treatment failure, slowed or reduced breathing, or low oxygen levels.

Limits of the evidence

The abstract combines nasal and buccal midazolam results, so it does not establish which route works better. It does not give detailed age ranges, seizure types, doses, treatment settings, or study-quality assessments. It also does not provide absolute success or side-effect rates, define treatment failure, or report how many minutes were saved. Similar overall results do not prove the treatments are exactly equivalent. Finding no statistically significant difference in breathing-related harms does not rule out uncommon serious harms.

For families and caregivers

These results support nasal and buccal midazolam as effective needle-free rescue options for stopping acute seizures in children. They were given faster and shortened the overall time to seizure control compared with the other treatments studied. The clearest advantage in stopping seizures within 10 minutes was over rectal diazepam; there was no clear overall advantage across all comparison treatments. The findings do not establish which rescue medicine or route is best for every child.

What to watch next

Evidence separating nasal from buccal treatment, reporting actual minutes saved and absolute outcome rates, and tracking uncommon serious side effects would strengthen these findings.

Terms in this summary

Benzodiazepines

A group of medicines, including midazolam, diazepam, and lorazepam, used to stop acute seizures.

Intranasal

Given into the nose.

Buccal

Given between the cheek and gum.

Meta-analysis

A method that combines numerical results from several studies.

Randomized controlled trial

A study that assigns participants by chance to different treatments and compares their outcomes.

Risk ratio (RR)

A comparison of how often an outcome happens in two groups. A value of 1 means no difference; whether higher or lower is better depends on the outcome.

95% confidence interval (CI)

A range showing uncertainty around an estimate. For a risk ratio, a range that includes 1 does not show a clear difference between groups.

Standardized mean difference (SMD)

A way to compare average results across studies using a common scale rather than units such as minutes. Here, negative values favor nasal or buccal midazolam.