Abstracts

Rationale: It is considered a standard of care that once a child with epilepsy has been seizure-free on anti-epileptic drugs (AED) for 2 years they are assessed for potential weaning of their medication. Despite a paucity of evidence to support it, many pediatric neurologists will perform an EEG prior to making this decision, typically to assist in their decision making. Our objective was to design a decision tree to assist with the decision of whether performing an EEG prior to making the decision to wean has an impact on the outcome. Methods: This was a case-specific decision tree evaluating cost effectiveness with three decision strategies considered - EEG, No EEG & Wean and No EEG & Do not wean. As our reference case, we chose a child with focal epilepsy presumed genetic. Our primary outcome was relapse within 5 years. We considered both the timing of the relapse and the impact of relapse (considering whether hospital or ICU admission was required). Cost and health state scores were incorporated into the model to allow a cost-effectiveness analysis to be performed. Outcome data for each epilepsy type was obtained from published pediatric literature where available. Additional data sources and assumptions of the model are explicitly presented. Results: In the reference case, the strategy No EEG & Wean was preferred. It strongly dominated other strategies, with higher quality-adjusted life years (QALY) and improved cost savings. Sensitivity analysis was conducted with little change to recommended strategy except at extreme medication costs, when performing an EEG was recommended. Decision analysis was performed for all other epilepsy classifications and EEG was never the preferred strategy at baseline variable estimates. Conclusions: Our analysis suggests the current widespread practice of performing an EEG prior to weaning AED in children is not cost effective and warrants review.