Human Travel, Not Geography, Predicted Spread in DRC Ebola Study

Human travel, not geography or town size, predicted where Bundibugyo Ebola spread next in the Democratic Republic of the Congo, according to an outbreak reconstruction published this week in Nature — findings that arrive as the six-month-old outbreak has passed 10,000 confirmed infections and 5,198 deaths.

The study, led by Congolese institutions with partners at the University of Oxford, Northeastern University and others, used anonymized mobile-phone data and test records to reconstruct the critical first weeks of the outbreak. People moving out of Mongbwalu, Bunia and Rwampara between April 24 and May 24 predicted which health zones would report infections by June 1 with near-perfect accuracy — outperforming models built on population size alone.

“Human mobility data predicted almost perfectly where this virus would appear next,” said Moritz Kraemer, DPhil, of the University of Oxford, and Samuel Scarpino, PhD, of Northeastern University, the study’s co-senior authors. “We need to position laboratory and care capability permanently in these high-risk emergence settings for improved responses in the future.”

Virus spread tied to travel patterns

  • Where people travel — not how many live in a town — best predicted which areas the virus reached next.
  • Cases doubled approximately every 10 days between early May and June 1st, 2026.
  • There are no licenced vaccines or approved treatments specifically for Bundibugyo.

A virus moving faster than the surveillance

The DRC declared the outbreak on 15 May after laboratory tests confirmed the first case, a person from Ituri province in the country’s northeast, was infected with the Bundibugyo strain. That same person had shown symptoms as early as 25 April — suggesting the virus circulated undetected for weeks.

The World Health Organization declared the outbreak a public health emergency of international concern on 17 May, citing infections among health-care workers, the geographical spread of cases and clusters of undocumented deaths. Africa CDC designated it a Public Health Emergency of Continental Security two days later.

Human Travel, Not Geography, Predicted Spread in DRC Ebola Study
Photo: nature.com

The speed was stark. Between May 10 and June 1, the number of health zones and areas reporting infections skyrocketed by 12 and 49, respectively. The estimated early case doubling time was 10 days.

What the phone data showed

The risk of Bundibugyo invasion fell with increasing distance from Bunia and Mongbwalu in Ituri and Rwampara in Uganda, up to 100 kilometers. But it rose slightly from 150 to 200 kilometers, likely due to human mobility along established travel corridors and then local growth. Pierre Akilimali, an epidemiologist at the National Institute of Public Health in Kinshasa, said that people travelled to larger towns and cities along established routes.

Destination population size, by contrast, was not predictive of disease spread. Mobility alone strongly delineated health zones that did and did not report infections.

The testing record also reveals how surveillance lagged behind the virus. Test positivity was highest for samples collected from people with symptom onset before May 13 (54%) and then stabilized at about 36%, likely due to the scale-up of case investigation and testing after the outbreak declaration.

The authors attributed some invalid results to insufficient blood taken, a lack of biological material from swabs of deceased patients, and difficulties transporting samples to the laboratory during the early phase of the outbreak, when local laboratories were just being established.

Human Travel, Not Geography, Predicted Spread in DRC Ebola Study
Photo: miragenews.com
Measure Figure Source
Confirmed infections (six months in) More than 10,000 Nature
Deaths 5,198 Nature
Early case doubling time 10 days Source 1
Health zones added, May 10–June 1 12 Source 1
Health areas added, May 10–June 1 49 Source 1
Test positivity, onset before May 13 54% Source 1

Who got sick, and who answered the data

Among 555 known test outcomes with documented age, sex and symptom onset, the median age of infected patients was 34.5 years, with women accounting for 59% of confirmed cases.

More than 50 scientists from over 25 institutions took part, including the Institut National de Recherche Biomédicale (INRB), the Institut National de Santé Publique (INSP), the University of Kinshasa, Africa CDC, WHO, the Pandemic Sciences Institute at the University of Oxford, Northeastern University, Imperial College London, Institut Pasteur and Flowminder.

Initially funded by the Gates Foundation, the group continues to support the Bundibugyo virus response and is now funded by the Coalition for Epidemic Preparedness Innovation (CEPI) as the Institut National de Recherche Biomedicale-led B-REACT network.

Placide Mbala-Kingebeni (INRB & Directorate of Research, Clinical Trials and Innovation, Africa CDC), senior author, said: “Twenty years ago, samples left this country and answers came back months later. This time, the sequencing, the diagnostics and the epidemiological analysis were led from Kinshasa and Bunia while the outbreak was still growing.”

Why this matters beyond Ituri

The findings carry a direct operational message for outbreak responders anywhere: mobility and travel-time data available within days of an alert can show where a virus is likely to go next, allowing surveillance, diagnostics and care to be positioned ahead of the epidemic wave.

The researchers concluded that the current lack of medical countermeasures and the rapid spatial expansion of the outbreak mean that swift and continued action is needed to control the outbreak.

For the DRC’s health system, the study is also a proof of national capability. “This study shows what our national public health institutions can deliver when the world invests in them — the evidence that guides the response was generated here, in the DRC, by the people closest to the outbreak,” said Placide Mbala-Kingebeni.

Recognizing virus symptoms is key

This article reports on epidemiological research, not a treatment, so there is no medication to avoid. The relevant clinical caution is symptom recognition: Bundibugyo virus disease presents with fever, and the sources note clusters of undocumented deaths in this outbreak.

The question still open

The study reconstructs the outbreak’s first weeks but cannot answer why the virus went undetected for weeks before May 15, or when Bundibugyo-specific vaccines and treatments will become available. Future work will focus on earlier detection of outbreaks through decentralised laboratory systems along human movement corridors.

References

  • Basu, M. “Record-breaking Ebola outbreak: phone data reconstruct first weeks of spread.” Nature, Nov. 12, 2026.
Photo of author

Health Desk

Health Desk at Archyde.

Dan Titus Week 5 Fantasy Notes: Curry Injury, Waiver Adds, Buy-Low Targets