To report findings from >2000 point-of-care smartphone-based electroencephalography (EEG) recordings in Kenya in routine clinical settings.
Access to EEG services remains limited in many low- and middle-income countries (LMICs) because of cost, infrastructure needs, and shortage of trained staff. BrainCapture’s BC-1 is a smartphone-based EEG system designed to enable non-specialist workers to record EEGs with remote expert interpretation.
Standardized 30-minute, 27-lead, asleep and awake scalp EEGs were performed on patients of any age clinically seen across 20 hospitals in Kenya. Primary indications included convulsions (72%), altered consciousness/headache (12%), or other suspected seizure-related events (16%). Each recording was collected by a trained allied healthcare worker and subsequently evaluated by clinical neurophysiologists for (1) quality of recording, (2) presence of epileptiform discharges, and (3) other abnormal EEG findings. Demographic variables included age group (≤3 years, 4-11 years, 12-18 years, 19-50 years, and ≥51 years) and sex.
A total of 2162 EEG recordings were performed. 1199 (58%) of the patients were male. 2076 (96%) of the recordings met interpretation standards, with a total of 618 (30%) of the recordings reported as abnormal, 458 (22%) displaying epileptiform activity and 160 (8%) indicating other abnormal activity. Among children ≤3 years old, 28% had abnormal EEGs and 23% showed epileptiform activity; ages 4–11 years (38% abnormal and 32% epileptiform); ages 12–18 years (34% abnormal and 28% epileptiform); adults 19-50 years (23% abnormal and 14% epileptiform); and adults ≥51 years (30% abnormal and 15% epileptiform).
A substantial number of EEGs showed clinically relevant abnormal activity. These findings demonstrate the feasibility and implementation of EEG acquisition in Kenya using a point-of-care remote interpretation system. Expansion of the use of the BrainCapture system may improve diagnostic confirmation, direct treatment, and inform care in additional LMICs and other resource limited settings.