Aims To systematically observe the use of four second-line antiepileptic drugs (AEDs) in children and the retention, perceived effectiveness, tolerability and safety over 12 months follow-up.

Methods All children starting clobazam (CLB), gabapentin (GAB), lamotrigine (LAM) or topiramate (TOP) in a defined period and hospital clinic population were ascertained retrospectively from the hospital pharmacy and paediatric neurology secretaries' databases and clinic letters. A standard data sheet was used to record background data and outcomes at more than 2, 6 and 12 months follow-up. This was a registered clinical audit and service evaluation, and was undertaken as part of medical student's BMedSci courses. Simple statistical analysis was undertaken in SPSS on an intention to treat basis.

Results 147 children aged 0.1–17 years have been studied to date: CLB (49), GAB (21), LAM (41), TOP (36), and notes on a further 100 are being interrogated. Epilepsies were evenly split between focal and generalised; apart from GAB (75% on GAB had focal epilepsies). Most were symptomatic or probably symptomatic with a median age of onset 2 years. The AEDs were used after a mean of two other AEDs had been withdrawn (range 1–10).

Preliminary analysis showed no major differences in perceived effectiveness between the AEDs, although there was a fall-off of efficacy for all from 2 months to 12 months follow-up: CLB 46–30%, GAB 53–43%, LAM 38–29%, TOP 33–22%. Probably, AED-related adverse events were found in 50% at some stage, although no serious adverse events were recorded. Overall, 54% were still taking the AED after 12 months.

Conclusions This is a simple way of auditing the effects of AED treatment in an unselected population with difficult to treat childhood epilepsies and complements data from randomised controlled trials which are usually in more selected populations with shorter follow-up. More data will be available by March 2010, and clinically significant differences in perceived effectiveness might then be confirmed or excluded, and subset analyses undertaken. Comparison with other second-line AEDs levetiracetam, zonisamide, lacosamide and others used in this population will be helpful.