Abstracts

Rationale: Despite the evidence that resective epilepsy surgery is more effective than medical therapy, the average time between seizure onset and surgery is 9 years for children and 19 years for adults. Promoting utilization of epilepsy surgery is a way to ensure that those with medically intractable epilepsy receive the best possible care promptly. Our goal was to develop a tool to assist clinicians identify patients who should be referred for an epilepsy surgery evaluation. Methods: The RAND/UCLA appropriateness method was followed. The first step involved performing systematic reviews of studies on the epidemiology, natural history, cost and utilization of surgery, and surgical outcomes of epilepsy, focusing on partial epilepsy. Clinical scenarios were then created based on: age, epilepsy duration, seizure type, frequency and severity of seizures, number of adequate antiepileptic drug (AED) trials, EEG and MRI findings. Twelve experts (adult and pediatric neurologists, epileptologists and epilepsy surgeons) rated the scenarios from 1-9 (1=most inappropriate; 9=most appropriate) for their appropriateness for an epilepsy surgery evaluation, first individually (based on the literature review) and then at a face-to-face meeting. All scenarios were re-rated after extensive discussion and all appropriate scenarios (rating ≥7) were re-rated for necessity. Necessity was defined as: appropriate (benefits exceed risks by sufficient margin to make it worth doing), it would be improper care not to offer referral, there is a reasonable chance referral will benefit patient, and magnitude of expected benefit from referral is not small. Results: For the literature review, 5061 abstracts were screened, with 763 articles selected for full review. 15 topics were summarized, 7 as systematic reviews. The first round of rating included 3072 indications for a surgical referral. This number was reduced to 2646 during the second round of appropriateness ratings. Of the 2646 scenarios, 20.6% were rated as appropriate (scale ≥7), 17.2% as uncertain (scale 4-6), and 61.5% as inappropriate (scale 1-3) for a surgical evaluation. The remaining 0.8% of cases could not be classified due to lack of consensus. Of the 544 appropriate cases, 56% were rated to be most necessary, 41.6% moderately necessary and 1.8% minimally necessary. No consensus was reached for the remaining 4 scenarios. None of the appropriate cases were rated as unnecessary. Failure of 1 AED only was always rated as inappropriate for a referral. Failure of 2 AEDs was usually rated as appropriate only if either MRI or both MRI and EEG were abnormal. EEG abnormalities without MRI abnormalities had minimal impact on appropriateness ratings. Additional examples of appropriate or inappropriate scenarios will be presented. Conclusions: Our study provides a comprehensive guide, based on the available evidence, for determining candidacy for an epilepsy surgery evaluation. The resulting decision support tool with all indications will be available on our study website.