Abstracts

Rationale: Epilepsy affects 50 million people worldwide, and is disproportionately prevalent in regions where specialist care and diagnostic tests are scarce. To improve diagnostic accuracy, we created a simple questionnaire with the aim to discriminate between focal versus generalized epilepsy, with the future goal to guide medication choices.Methods: Through literature review, 30 key semiology features were identified and validated by retrospective chart review of 75 children with epilepsy in a North-American tertiary care hospital to identify 15 key questions (Figure 1). The questionnaire was prospectively validated in Tanzania (translated into Kiswahili) and Zambia (translated into Bemba and Nyanja). Children aged 6 months to 18 years with suspected or active epilepsy were identified in the general pediatrics clinic at Muhimbili National Hospital in Dar Es Salaam, Tanzania, and the University Teaching Hospital Pediatric Neurology Clinic in Lusaka, Zambia. A non-physician administered the questionnaire to the patient’s caregiver. A pediatric neurologist performed a separate history and physical. Routine EEG was performed and interpreted locally for clinical purposes. The obtained EEG was uploaded to a secure web-based EEG platform, where it was interpreted remotely in Boston, by clinical neurophysiologists blinded to clinical data. The performance of the questionnaire was then compared to the patient’s final electroclinical diagnosis.Results: A total of 59 children were included, 28 from Tanzania and 31 from Zambia. 5 patients from Zambia were excluded due to incomplete data collection. We found focal or multifocal epilepsy in 31 (57%) patients and 12 (22%) with generalized epilepsy. The remaining 11 (21%) did not meet criteria for epilepsy. Data was analyzed using a Rasch model, testing the questionnaire for reliability and discrimination validation of classifying focal versus generalized epilepsy against the electroclinical diagnosis (Figure 2). The mean distribution for individuals with focal epilepsy was +1.648 logits and +0.277 logits for those with generalized epilepsy. The difference of 1.371 logits reflects a large effect size. These differences were also significant with an ANOVA test (F(1, 41)= 11.968, p=0.001).Conclusions: Our questionnaire provides a simple method to improve diagnostic accuracy, which may help assist in bridging the diagnostic gap in pediatric epilepsy in resource limited settings. This tool was specifically designed to be easily implemented by any healthcare provider. Future studies using this tool in different regions and languages will increase the power and utility of this questionnaire in different settings. While impact of its use on current standards of practice needs to be assessed, it is our hope it could ultimately assist in helping guide medication choices and improve treatment outcomes, in regions where there are limited resources available. Sources of funding: World Federation of Neurology, Boston Children’s Hospital Global Health Program