OBJECTIVE: A previous European cost-utility study reported that use of buccal midazolam in the community setting for the treatment of prolonged seizures (ie, seizures lasting ≥5 minutes) in children was associated with an overall €12 507 399 reduction in annual costs charged to the Italian national health service compared with rectal diazepam. We re-evaluated these findings by applying a more conservative approach.

METHODS: The Italian Delphi panel reconvened to apply a more conservative assessment of available reports. A decision-tree model was used, allowing for different treatment pathways depending on whether or not a caregiver administers treatment, an ambulance is required for transport of the child to hospital, and an inpatient stay is required. Direct medical costs were derived from Italian healthcare system data. Estimates of the annual number of prolonged tonic-clonic seizures expected in the country were based on studies which assessed seizure duration using video-EEG recordings and medical records.

RESULTS: Although drug acquisition costs were greater for buccal midazolam than for rectal diazepam, the acquisition cost difference was outweighed by larger cost savings resulting mostly from a reduction in hospital admissions. Assuming that 1.2% of tonic and/or clonic seizures occurring in children and adolescents over a 12-month period are prolonged, the annual nationwide reduction in costs from preferring buccal midazolam to rectal diazepam was estimated at €3 577 587.9.

CONCLUSIONS: In this more conservative revised analysis, the high cost of buccal midazolam is still counteracted by greater cost savings compared with rectal diazepam, but cost reduction was less than previously estimated.

Original languageEnglish
Pages (from-to)24-28
Number of pages5
JournalActa Neurologica Scandinavica
Issue number1
Publication statusPublished - Jan 2018


Dive into the research topics of 'Midazolam vs diazepam in prolonged seizures in children: A pharmacoeconomic approach'. Together they form a unique fingerprint.

Cite this