A Japanese database study compared lacosamide and levetiracetam in patients who also received a benzodiazepine, separating the adjunctive comparison from first-line treatment.
The investigators used a nationwide Japanese inpatient database to identify status-epilepticus admissions during a period ending in March 2023. Eligible patients received intravenous diazepam or lorazepam on admission day and either intravenous lacosamide or levetiracetam that day. There were 4,605 eligible patients: 227 in the lacosamide group and 4,378 in the levetiracetam group.
The comparison was therefore between adjunctive regimens in a benzodiazepine-treated population. It was not a trial of lacosamide instead of diazepam or lorazepam, and it did not establish the relative effectiveness of the two benzodiazepines.
After propensity-score overlap weighting, in-hospital mortality was 4.0% with lacosamide and 4.6% with levetiracetam. The adjusted risk-difference interval crossed zero, and the study did not detect a significant mortality difference. Length of stay and total hospitalization cost also did not differ significantly.
A discharge Glasgow Coma Scale score of nine or lower was less frequent in the lacosamide group: 0.6% versus 2.4%, with an adjusted risk difference of −2.1 percentage points and a 95% confidence interval of −3.3 to −0.9. This finding concerned a particular consciousness endpoint, not a comprehensive long-term neurological assessment.
Diagnosis was based on codes and medication records rather than uniform neurological confirmation. The database did not provide sufficiently precise timing to establish every administration sequence, although the authors considered the pattern consistent with usual initial benzodiazepine treatment.
Weighting addresses measured differences between groups but cannot eliminate unmeasured treatment-selection effects. The study also could not identify refractory subtypes reliably from the available coding. Its contribution is a large observational comparison of early adjunctive care. It does not prove that a particular patient’s outcome was caused by the chosen adjunct or define whether initial benzodiazepine treatment was adequate.
Original analysis of the supplied journal article, identified as PMC12903811.