Abstracts

Rationale: Existing research on academic outcome following pediatric epilepsy surgery has yielded variable results; a 2015 study with a large sample size (Puka et al.) reported significant declines in reading decoding, math calculation and spelling one year after resective surgery. Methodological differences across studies and failure to account for factors that can impact test-retest change in the absence of any surgical effect (e.g., test reliability, practice effects, measurement error) likely account for inconsistent findings. Reliable change indices (RCIs), which control for these important artifacts, were recently published for children with epilepsy. The current study used these RCIs to examine clinically meaningful change in academic achievement scores following surgical resections in children with pharmacoresistant seizures.Methods: Sixty-five children (mean age=11 years, range 6-15) completed portions of the Woodcock-Johnson III Tests of Achievement (WJ-III) before and approximately 6.5 months after epilepsy surgery. Most (n= 35) underwent temporal lobe resections; children with frontal (n=15), parietal (n=3) and multilobar resections (n=12) were also included. Surgical side (57% left) and sex (54% female) was balanced in this sample. We calculated change scores for select WJ-III composites and subtests and categorized them as “improve,” “decline” or “no change” based on published RCIs (90% confidence interval) for children with epilepsy. Subsequent t-tests and chi-squares were conducted to identify factors related to postoperative academic decline.Results: No meaningful change was observed on most of the academic measures (Broad Math, Letter Word Identification, Reading Fluency, Calculation, Applied Problems, Math Fluency, Spelling, Writing Fluency, Word Attack, Punctuation & Capitalization). More children than expected demonstrated clinically meaningful change in both directions (improvements and declines) on Passage Comprehension, Writing Samples, Broad Writing, and Broad Reading (p range = 0.04-0.003). Declines in Academic Fluency also occurred at a greater rate than expected (p range = 0.04). The proportion of patients who demonstrated clinically meaningful change on each of the WJ-III indices and subtests is depicted in Figures 1 and 2, respectively. Academic decline was not related to demographic or seizure variables.Conclusions: This is the first known study using RCIs developed on a large sample of children with epilepsy to assess clinically meaningful change in academic scores after epilepsy surgery. It is encouraging that many of the children in this sample showed little change in academic performance. More children than expected demonstrated meaningful score declines or improvements in reading comprehension with word retrieval demands (Passage Comprehension) and writing (Writing Samples). Academic fluency scores declined more frequently than expected. Changes in other cognitive functions (e.g., word access, organization and processing speed) could explain the greater than expected academic score improvements and declines.