Health officials combating the Ebola crisis in eastern Democratic Republic of Congo are utilizing de-identified mobile phone data to track population movements and pinpoint potential areas for virus transmission, as disclosed by the World Health Organization (WHO) and collaborating researchers.
This innovative strategy, unprecedented in previous Ebola responses, is being employed as authorities grapple with what the WHO has labeled as the most rapidly expanding Ebola outbreak on record. The outbreak has been reported in six provinces, with over 6,250 individuals infected and 3,039 fatalities since the outbreak’s declaration on May 15, according to the most recent government data.
This particular outbreak, the 17th in Congo, is caused by the Bundibugyo strain of Ebola, for which no approved vaccines or treatments exist.
In light of the virus’s progression, health officials are seeking methods to predict potential outbreak locations and allocate limited surveillance and response resources effectively.
Olivier Le Polain, head of epidemiology and analytics for response at the WHO Health Emergencies Programme, emphasized the critical role of population movements in outbreaks, particularly in the current epidemic.
While traditional risk assessments typically focus on outbreak hotspots and assume the virus spreads to nearby areas, analyzing mobility data enables responders to identify risks in locations that may be geographically distant but interconnected through travel patterns.
The data analysis, conducted by Swedish non-profit Flowminder, leverages de-identified records generated when mobile phone users connect to telecommunications networks. As individuals move around the country, their phones connect to various network antennas, creating records that indicate where these connections occurred. Researchers utilize this information to estimate travel patterns between different regions.
Vodacom, Congo’s largest telecommunications operator, supplies the data free of charge. Flowminder ensures that the data received is aggregated and anonymized, with no access to individual subscriber information.
Le Polain pointed out the significance of understanding population movements in areas like Ituri, where mining activities and trade contribute to substantial population mobility, affecting the virus’s spread.
The analysis initially focused on Bunia, Mongbwalu, and Rwampara, the suspected origin areas of the outbreak, tracking travel patterns from April 3 to April 23. The data revealed significant flows to neighboring regions in Ituri and North Kivu provinces, with all top destinations receiving travelers from the outbreak areas reporting confirmed Ebola cases by the end of June.
As the outbreak expanded beyond its initial epicenter, Flowminder extended the analysis to additional affected regions, examining travel patterns from these hotspots. Kisangani, a major transport hub on the Congo River connecting eastern and western Congo, emerged as a key location of interest.
Recent studies highlighted connections between Kisangani and Kinshasa, as well as settlements along the Congo River and areas near borders with Central African Republic and South Sudan. This underscores the potential risk of transmission to Congo’s capital and largest city.
The WHO continues to caution about the outbreak spreading to Kinshasa and neighboring nations. A modeling exercise presented in late June estimated a 70% probability of a case being detected in Kinshasa by the end of September.
While the analysis offers valuable insights, its reliance on data from Congolese mobile networks poses limitations, providing only a partial view of movements across international borders. Efforts are underway to collaborate with other telecom operators to enhance coverage and obtain more comprehensive data for improved response coordination.