Are serum anticonvulsant levels in people with epilepsy appropriately monitored?

The medical care of people with epilepsy has often been described as being poor, although objective markers for the quality of epilepsy care are lacking. This paper describes the results of using a simple quality marker, appropriate measuring of serum anticonvulsant levels, in assessing the quality of epilepsy care. The checking of serum phenytoin levels in certain clinical circumstances is advocated, whereas the checking of serum sodium valproate levels is not generally supported. A total of 1254 people with epilepsy in the community had their medical records examined for evidence of checking of anticonvulsant levels and 1204 of these individuals completed questionnaires about their epilepsy and its treatment. Of those on phenytoin, only 26% to 47% had phenytoin levels checked appropriately; 23% of patients on sodium valproate were inappropriately having their serum levels checked. The only clinical or organizational factor that predicted whether checking of serum phenytoin levels was performed was whether or not patients reported three common phenytoin side-effects but this still showed a small effect size (odds ratio 2.4).