A Cochrane review [Abstract] 1 included 12 studies. Individual patient data were available for 595 participants representing 50% of the total of 1192 participants recruited into 12 trials that met the inclusion criteria.
TThe main overall results were time to withdrawal of allocated treatment: 1.04 (95% CI 0.78 to 1.39; 3 studies, n=546); time to 12-month remission: 1.01 (95% CI 0.78 to 1.31); time to 6-month remission: 1.11 (95% CI 0.81 to 1.37); and time to first seizure: 0.85 (95% CI 0.70 to 1.04). The results suggest no overall statistically significant difference between the drugs for these outcomes. There is some evidence of an advantage for phenytoin for individuals with generalised onset seizures for our primary outcome (time to withdrawal of allocated treatment): pooled HR 0.42 (95% CI 0.18 to 0.96; 3 studies, n=118); and a statistical interaction between treatment effect and epilepsy type (partial versus generalised) for this outcome (P = 0.02), however misclassification of seizure type for up to 48 individuals (32% of those with generalised epilepsy) may have confounded the results of this review. There was no evidence that phenytoin is more likely to be associated with serious side effects than carbamazepine; 26 individuals withdrew from 290 randomised (9%) to carbamazepine due to adverse effects compared to 12 out of 299 (4%) randomised to phenytoin from 4 studies conducted in the USA and Europe (RR 1.42, 95% CI 1.13 to 1.80, 3 studies, n=546).
Comment: The quality of evidence is downgraded by indirectness (data was missing from 50% of eligible patients).
Primary/Secondary Keywords