The anticonvulsant (AED) history for 216 children and adolescents with epil
epsy was reviewed to determine the incidence and types of significant side
effects (SSE) which warranted ceasing the drug (not due to a lack of respon
se or a high dose). All parents of patients with epilepsy seen by the autho
r over a 2 year period (March 1996 - March 1998) were questioned about SSE
to previous AEDs, and the child's current therapy was also monitored prospe
ctively to determine SSE. There were 107 girls and 109 boys ranging in age
from 3 months - 18 years. Eighty-three patients had been exposed to a singl
e AED while 133 had multiple AED exposures: mean 3.6 drugs; range 2-10 drug
s, They were exposed to a total of 568 AEDs with SSE occurring in 15% of dr
ug contacts: 7% due to behavioural changes such as irritability, aggression
or hyperactivity: 8% were due to other factors such as a rash, headache, g
astrointestinal disturbance or drowsiness. Fifty-seven children (26%) had e
xperienced at least one SSE with 19 (9%) having SSE to more than one AED (r
ange 2-4). Global developmental delay or an intellectual disability (ID) we
re present in 67 patients, and 27 (40%) of these experienced SSE compared w
ith 30 (20%) of the group with normal cognition. This difference was princi
pally due to the higher incidence of behavioural SSE in the ID group 28% ve
rsus 6% for the normal cognition group. Allowing for the higher number of A
EDs used in the ID group (implying that their epilepsy was more difficult t
o control), behavioural SSE were still significantly more likely to occur i
n this group, i.e, 1:9.6 drug exposures compared with 1:31.8 exposures for
the normal cognition group (P<0.001). Monotherapy trials underestimate the
true incidence of SSE in clinical practice as 26% of children had experienc
ed at least one SSE and 9% had SSE to more than one AED. Those with ID were
three times more likely to have behavioural SSE than children with normal
cognition. (C) 2000 Harcourt Publishers Ltd.