Letter: Disodium cromoglycate and DNA-ase in treatment of bronchial asthma.

the convulsion rate. No such evidence has been published, and indeed in Thorn's paper the recurrence rate was 12 00% in those with mean serum phenobarbitone levels over 61 umol (14 mg)/l and 12 80o in those with lesser concentrations. Further, we do not like the concept that some children may be protected by phenobarbitone and others not, without evidence that we were treating two types of patient. Even if there is a reduction in febrile convulsion recurrence rates on phenobarbitone the cost must be counted. There is no evidence that the risk of later recurrent afebrile seizures ("epilepsy") is reduced,6 while the incidence of adverse effects is unacceptably high, even at lower blood concentrations.3 Although a recent publication7 indicates that sodium valproate may reduce the recurrence rate of febrile convulsions without the side effects often associated with phenobarbitone, we must reiterate our point that the really serious complication, febrile status epilepticus, usually occurs with the first convulsion. The resultant brain damage must be prevented by a research-based countrywide emergency treatment service. In this field there is much anecdotal but little scientific evidence. Recommendations include paraldehyde, parenteral phenobarbitone, fever reduction, and diazepam intravenously, rectally, or intramuscularly with hyalase, chlormethiazole, and lignocaine. If other authors do have clear evidence of the relative value of one or other therapy or management of febrile status, then we and many of your readers would like to know where this evidence is to be found.