Abstracts

Rationale: Most people with poorly controlled epilepsy live in low- and middle-income countries. In the lowest income countries, access to epilepsy care is often challenging. Computational and practical studies have been undertaken to examine the use of Unmanned Aerial Vehicles (UAVs) to deliver medical supplies, like blood and vaccines; however, no such studies focusing on UAV delivery of treatments for epilepsy have been completed in a lower-income country.Previously, we identified a cohort of >300 patients with poorly managed epilepsy in Conakry, the capital of the Republic of Guinea. Most patients did not own a car or have regular access to public transport, leading to a confluence of poor urban planning, congested roadways, and high clinical need in Conakry.  Methods: We developed a theoretical model for UAVs to deliver anti-epileptic drugs (AEDs) via a hub-and-spoke system in Conakry. Ignace Deen Hospital (IDH), a large, academic referral hospital near the Western edge of the city, was chosen as the hub. A total of 27 mosques, pharmacies, and gas stations were chosen as the spokes because they are attended most hours of the day, are areas of socialization and commerce, and are distributed relatively evenly in the city.UAV specifications were based on previously published studies: ascent and descent times were set to 10 seconds, cruising altitude to 60m, top speed to 27.8m/s, on-scene delivery time to 3 minutes, and post-delivery maintenance time to 30 minutes. Commercially available UAV prices vary from 150-1,500 USD and can carry up to 5kg. Results: We applied the model to regular and emergency delivery of AEDs. We determined the number of spokes that could be served by 1, 2, 3, and 4 UAVs in an 8-, 12-, 24-, and 40-hour period. A single UAV could serve all spoke locations once in 20.41 hours, all mosques in 5.63 hours, all gas stations in 6.60 hours, and all pharmacies in 8.18 hours. Service capability increased with fleet size. A fleet of 4 UAVs, could service every spoke location >1 time within an 8-hour period, >2 times within a 12-hour period, >4 times within a 24-hour period, and >7 times within a 40-hour period. Additionally, in an 8-hour period, a fleet of 4 drones could service all pharmacies >3 times, all mosques >5 times, or all gas stations >4 times. In response to a case of status epilepticus, a single drone delivering a benzodiazepine could reach 24 of 27 spoke locations within 15 minutes of launch from the central hub, 20 of 27 within 10 minutes, and 8 of 27 within 5 minutes. Conclusions: We present a theoretical model to implement UAV delivery of AEDs in Conakry. The model is based on real locations in a city with extremely high need for epilepsy care. A relatively small fleet of UAVs (n=4) would be able to adequately service a well-dispersed series of 27 spokes throughout the city. In the case of status epilepticus, a UAV would be able to reach most spoke destinations within 15 minutes of launch from the hub hospital. Funding: NIH (R21NS098886)