New research supported by the Oxford Martin School shows how short journeys and travel for healthcare can substantially affect early estimates of an Ebola outbreak. Published in Nature Health, the study forms part of a wider University of Oxford effort to support the international response to the ongoing Bundibugyo Ebola crisis.

When an epidemic is unfolding, reliable case numbers are often hardest to obtain at precisely the point when decisions are most urgent. One way researchers can fill that gap is by looking at infected people detected after travelling outside the outbreak area. New research shows that interpreting those cases depends critically on understanding how people cross borders, and why they make the journey.

Professor Christl Donnelly and Dr Younjung Kim, from the Oxford Martin Programme on Digital Pandemic Preparedness, together with Dr Cathal Mills, examined this problem during the ongoing outbreak of Bundibugyo virus disease in the Democratic Republic of Congo (DRC).

Professor Donnelly and Dr Kim had confronted a similar modelling challenge during a hantavirus outbreak associated with the cruise ship MV Hondius. Estimating the number of clinically significant infections in Argentina required them to consider the length of each cruise, because this affected the chance that an infection would be detected while a visitor was still travelling.

When Dr Kim read early estimates of the Bundibugyo outbreak based on cases confirmed across the border in Uganda, he recognised that the same issue could matter here. Many journeys across the DRC-Uganda border are brief, meaning someone infected in the DRC could cross into Uganda and return home before their infection was detected there.

Border monitoring data used by the researchers showed that 90.6% of trips to Uganda from Ituri and Nord Kivu in the DRC lasted a week or less, while 64.7% lasted less than a day.

Conversations with colleagues in Uganda revealed another important factor: cases confirmed in Uganda had travelled from the DRC specifically to seek medical care. For the researchers, the reason behind these journeys was crucial, because the conventional approach assumes that infection does not influence whether someone travels.

Dr Kim said: "Accounting for both the short journeys and why these cases travelled allowed us to better reflect the local context in our estimates."

Incorporating these factors substantially changed estimates of the epidemic's likely size and the uncertainty around them. The study does not provide a definitive count of infections in the DRC, but shows how local patterns of movement can affect estimates during the uncertain early stages of an outbreak.

"In my experience, the response works best with open communication between epidemiologists and public health policymakers."

Professor Christl Donnelly

For Professor Donnelly, the work underlines the importance of connecting analysis with the needs of those making decisions during an epidemic: "In epidemics we want there to be as few cases as possible, while learning as much as possible from the cases that do arise. In my experience, the response works best with open communication between epidemiologists and public health policymakers. Understanding policy priorities focusses analyses on the questions that matter most."

Professor Donnelly and Dr Kim are continuing discussions with colleagues in Uganda about how modelling evidence can best inform policymaking.

Their research forms part of a wider University of Oxford response to the Bundibugyo Ebola outbreak, spanning vaccine and treatment trials, clinical research, genomics and epidemiology. In July, the Oxford Vaccine Group began the world's first Phase I clinical trial of a vaccine designed specifically to protect against Bundibugyo ebolavirus.