The opinions of families of 76 children with epilepsy (>3 months seizure-free) and their 4 physician epilepsy specialists regarding acceptable risks of seizure recurrence (RSR) after AED withdrawal were investigated by questionnaire at the IWK Children’s Hospital, Halifax, Nova Scotia, Canada. Families’ responses were very variable: a RSR of 25% was unacceptable to 42% of families, whereas a >75% risk was considered acceptable by 20%. Families responses were dependent on previous seizure frequency, multiple seizure types, school grades repeated, and the habits of playing the lottery. The degree of risk acceptable to a particular family was not predicted by their physicians. Physician’s opinions of acceptable RSR were more consistent, a median of 40%, but they varied from 0 to 90% for individual children. Families gave the primary responsibility for discontinuing AEDs to the physician in 80% of cases, but 46% reserved a secondary role in the decision for the parents. The majority of families (89%) denied the physician an exclusive role in the decision making. 
COMMENT. This center’s practice of discontinuing AEDs after a seizure-free 2 year period, with an expected seizure recurrence of 30-40%, results in an unacceptable degree of risk for more than half the families in this study. Withdrawal of AEDs after a 1 year seizure-free period provided a similar high risk of recurrence of 30-40%, in a recent study from the same center. 
Physicians should be aware of the parents’ opinions and attitudes in regard to risk of seizure recurrence, when advising on the time to discontinue treatment of a child with epilepsy, but they cannot allow a “parent’s strategy of playing lotteries” to intervene in this important decision. Each child with epilepsy is an individual, and the time for anticonvulsant withdrawal should be determined on an individual basis, having regard to varying predictive factors, not based on a generalized fixed period of 1 or 2 years. An expected seizure recurrence of 30-40% should be unacceptable in current practice, and the consequences of a single seizure relapse, especially in adolescents and young adults, should demand a more conservative and individual approach to this important decision. See Progress in Pediatric Neurology I, PNB Publishers, 1991, pp 100-104, for further references and commentaries on anticonvulsant withdrawal practices.