Only a minority of patients showed a significant drop in IQ or developmental scores, more often in those studies that reported on surgery in younger children. Some authors have commented on those children who experienced a drop in standardized developmental scores (e.g., Ref. [24]) as they nevertheless often do show developmental progress albeit at a slower pace than their healthy peers (see limitations below). On the positive side is that about twice as many patients improved postoperatively than showed cognitive deterioration. Many studies reported a beneficial IQ effect of either surgery or seizure freedom associated with surgery. Nevertheless, the potential benefit of significant seizure reduction has not been sufficiently considered. The impact of AED withdrawal can be demonstrated in some studies but is not universally seen. This is likely due to the coarseness of the AED analysis conducted, e.g., reduction in number of drugs used and the contribution of individual drugs have not been considered. There is nevertheless compelling evidence of AED impacting on cognition [18].

A direct comparison of studies with long- versus short-term follow-up did not support our prediction of improved outcomes with longer follow-up. Although the expected trends were seen, those did not reach significance, perhaps due to the large variability in study characteristics. When taking into account the variability in participant numbers by computing weighted indices, modest support for this association was indeed observed, in particularly when statistically controlling for the large differences in reported seizure freedom between studies (from 35 to 100 %).

A key caveat is that the majority of studies did not specify a minimum follow-up time, which meant that most studies had a mixture of very short and very long postsurgical periods, with the inevitably possibility for bias (see limitations below). Two studies included a minimum follow-up period of over 5 years, but came to different conclusions regarding IQ change [12, 30]. The study of Skirrow and colleagues examined outcome in patients who had undergone temporal lobe surgery in childhood after a mean follow-up of 9 years (range 5–15 years). They report improved IQ in the surgical group, a change not observed in a non-surgery epilepsy control group. Greater IQ improvements were found among patients with lower IQs before surgery. Discontinuation of AEDs was a positive predictor of IQ change. An analysis of interim follow-up data points (available in a proportion of patients only) suggested that the IQ increase was only observed after 6 or more years post-surgery. In contrast, a 10-year follow-up of a heterogeneous cohort of 17 patients (including temporal and extratemporal resections, callosotomy, multiple subpial transections) by Viggedal and colleagues [30] did not report any significant group-level changes in IQ, although the absolute degree of change was similar to that of Skirrow et al. [12]. Notwithstanding the differing study populations (temporal lobectomy vs mixed sample), the cohorts also differed much with respect to seizure freedom (86 % vs 35 %) and proportion of AED discontinuation (57 % vs 18 %). The discrepancy in study conclusions is not incompatible with the impact of those postoperative factors as demonstrated above.