Prevention of simple febrile convulsions

I should like to comment on Ngwane and Bower's paper.1 First of all, their statement that phenobarbitone takes 72 hours to achieve an adequate blood level is an often repeated inaccuracy. Phenobarbitone, like any other drug, will only achieve so-called therapeutic blood levels if an adequate dose is given. We have achieved this level within 90 minutes, orally and intramuscularly, after a dose of 15 mg/kg.2 I was intrigued that about 75% of their 265 children did not meet their criteria for simple febrile convulsions. The significance of their definition is not clear, and it is unfortunate that so few were left on whom to perform the trial. It would appear that the untreated group, even allowing for their age at onset, had very severe febrile seizures if 6 out of 21 children had prolonged fits lasting 25 minutes or longer within the follow-up period. In our study of 83 unselected children with febrile convulsions, to whom no prophylaxis was given but only strict instructions for the prevention of prolonged seizures, only 2 children went on to have prolonged fits in at least a oneyear follow-up.3 Is the untreated group therefore a satisfactory one for comparison, particularly as it was not designed as a control group? Although not commented on in the text, their results also show that phenobarbitone is not statistically better than no treatment in preventing further fits. Blood levels in the 4 treatment failures at the time of the fit would have been interesting. Sodium valproate did appear to be more effective than no treatment. However for those therapeutic nihilists, like myself, who feel that prolonged seizures can generally be prevented by other means, Ngwane and Bower performed one valuable service; they suggested that sodium valproate was just as harmful to children as phenobarbitone. Again however the numbers were small, and it would have been interesting to know the exact nature of the side effects.