This retrospective, cross-sectional study evaluated seizure-detection rates using ambulatory EEG (aEEG) in 382 adult patients from an underserved epilepsy population at the University of Illinois Hospital, Chicago, and examined clinical and socioeconomic factors — including mild cognitive impairment (MCI) — linked to discrepancies between self-reported and objectively detected seizure control.
Seizures were captured on aEEG in 46/382 patients (12%), with 80% of those being subclinical. On multivariate analysis, number of anti-seizure medications (OR 2.10, 95% CI 1.44–3.07) and presence of MCI (OR 6.13, 95% CI 3.04–12.38) independently predicted seizures on aEEG; ~70% of patients with subclinical seizures who reported rare seizures at home had MCI.
- Retrospective, cross-sectional design limits causal inference between MCI and subclinical seizures. - Single-center study at a safety-net hospital may limit generalizability to other underserved or well-resourced settings. - Reliance on self-reported seizure frequency introduces recall bias, especially in patients with MCI.
In patients with epilepsy — especially those with MCI or on multiple anti-seizure medications — self-reported seizure control is unreliable; ambulatory EEG should be considered to detect subclinical seizures that patients cannot perceive or accurately report. This is particularly critical in underserved populations where diagnostic gaps are wider.
Explore related topics
