Accuracy of expert predictions of seizure freedom after epilepsy surgery

To evaluate the accuracy of expert estimations of achieving seizure freedom after epilepsy surgery in the context of presurgical patient counseling.

The retrospective study comprised a random sample of 200 patients who underwent any type of resective epilepsy surgery at the University of Bonn Epilepsy Center and the routine 1-year postoperative control visit in the years from 2008-2016. The prediction by a team of epileptologists and the actual seizure outcome were extracted from the pre- and postsurgical medical files, respectively. A deviation of >10% was a priori defined as a relevant discrepancy.

Estimated chances of achieving seizure freedom ranged from 30 to 90% (mean: 67%). The actual seizure freedom rate was 66% (Engel Ia/ ILAE 1a). Nine of 12 estimation categories showed a tolerable deviation of ≤10%, none of these with a worse than expected outcome. Two estimation categories (40-50%, and 80%) showed a worse actual seizure outcome with deviations of -40% (n = 3); and -17% (n = 30), respectively. All in all, for 83% of the patients a correct prediction was provided.

For the vast majority of surgical patients, the expert prediction of postsurgical seizure freedom at the 1-year follow-up was accurate despite the heterogeneity of patients and surgical procedures.