Major Finding: Based on a decision-support tool, nearly 21% of 2,646 clinical scenarios created from different combinations of patient-level factors were considered appropriate for evaluation for epilepsy surgery; 17% were considered uncertain for appropriateness.
Data Source: An expert panel's use of a decision-support tool.
Disclosures: Dr. Jetté reported no conflicts of interest relevant to her presentation.
BOSTON — An online decision-support tool may help to close the protracted gap between seizure onset and referral for surgery in patients with medically intractable epilepsy, based on results obtained by an expert panel.
The user-friendly tool is designed for use by all clinicians who treat epilepsy patients but who may not be epilepsy specialists, said Dr. Nathalie Jett, who developed the tool with her colleagues at the University of Calgary (Alta.).
The tool rates the appropriateness and necessity of referring individual patients for a surgery evaluation based on factors such as age, epilepsy duration, seizure type, frequency and severity of seizures, the number of adequate epilepsy drug trials, and EEG and MRI findings, said Dr. Jetté of the department of neurology at the university.
“The goal is to help neurologists and other clinicians identify which patients should be referred for epilepsy surgery evaluation, and ultimately [to facilitate] earlier surgical treatment when appropriate,” Dr. Jetté said at the annual meeting of the American Epilepsy Society.
Despite surgical success rates as high as 90% and 60%, respectively, for patients with medically intractable temporal lobe epilepsy and other partial epilepsies, the average time between seizure onset and surgery for these patients is 9 years for children and 19 years for adults, according to Dr. Jetté, who attributed the underutilization of surgery to misconceptions about the associated risks. “Epilepsy surgery is often perceived as a last resort, rather than a reasonable option early on.”
To develop the rating tool, Dr. Jetté and her colleagues used the RAND/UCLA appropriateness method, in which they performed systematic literature reviews on the epidemiology and natural history of drug-resistant epilepsy, the cost and utilization of surgery, and outcomes of surgery for partial epilepsy.
Based on the literature review and on discussion during a face-to-face meeting, an expert panel comprising adult and pediatric neurologists, epileptologists, and epilepsy surgeons rated clinical scenarios (created from the possible combinations of the aforementioned patient factors) for their appropriateness for an epilepsy surgery evaluation, Dr. Jetté said.
“The scenarios were rated on a scale from 1 to 9, where 1 was the most inappropriate and 9 was the most appropriate. After extensive discussion, all of the scenarios were re-rated, and those that were appropriate for referral [rated a 7 or higher] were re-rated for necessity.”
For the rating purposes, surgical referral was considered a necessity if the presumed benefits of referral exceeded the risks by a sufficient margin, if failing to refer the patient would be improper care, if there was a reasonable chance the referral would benefit the patient, and if the magnitude of the expected benefit “was not small,” Dr. Jetté said.
Of 2,646 clinical scenarios, nearly 21% received a rating of at least 7 and as such were considered appropriate for a surgical evaluation. About 17% were considered uncertain for appropriateness because they were rated between 4 and 6, and nearly 62% were deemed inappropriate because they were rated between 1 and 3. Fewer than 1% of the scenarios could not be classified due to lack of consensus, she reported.
In practice, a patient who has failed one antiepileptic drug (AED) would be considered inappropriate for referral, whereas a patient who has failed two AEDs and has an abnormal MRI and EEG would typically be an appropriate candidate for surgical evaluation, Dr. Jetté explained.
With respect to necessity, “none of the appropriate cases were rated as unnecessary,” although four cases were not rated due to lack of consensus. Of the remaining appropriate cases, 56% were rated as most necessary, 42% as moderately necessary, and 2% as minimally necessary, she said.
The decision support tool, which is currently being tested and refined in Canadian clinics, is expected to be available online in mid-2010.