Predicting pediatric epilepsy surgery outcomes via a multidisciplinary epilepsy surgery conference survey.

Abstract

There is a paucity of data assessing the accuracy of clinician prognostication for pediatric epilepsy surgery outcomes during multidisciplinary epilepsy surgery conferences (ESCs). This cross-sectional study, completed at Children's Hospital of Colorado (CHCO) and Children's Health Dallas (CHD), compared seizure freedom predictions by the ESC panels (captured via anonymous surveys of ESC members) to estimated outcomes by the Seizure Freedom Score (SFS) and to observed post-operative seizure freedom rates. The sample comprised 20 patients from CHCO (12/2021 - 12/2022) and 10 patients from CHD (6/2022-12/2023) satisfying inclusion criteria of at least 9 months of follow-up and survey completion by ≥ 50% of the ESC panels. Six patients were excluded from ESC prediction analysis for lack of consensus in their predicted prognoses. Predicted seizure freedom probabilities were assigned in ranges. The ESC panels accurately predicted seizure freedom for 80% (n = 8/10) of children for whom they had assigned higher predictive probabilities. Most of these children had known lesions and concordant presurgical data. ESC physicians accurately predicted lack of seizure freedom for 75% (n = 3/4) of children for whom they assigned lower predictive probabilities for seizure freedom. However, for the intermediate probabilities of 50-69%, 27% (n = 3/11) of children achieved seizure freedom. The SFS (scores 3-4) accurately predicted seizure freedom for 69% (n = 9/13) of patients. The ESC panels were most accurate in predicting seizure freedom at the extremes of the predicted probability spectrum and performed similarly to the SFS.