📄 public and global health

Early assessment of potential airline-mediated importation risk during the 2026 DRC-Uganda Bundibugyo virus disease outbreak

This paper presents a reproducible framework using effective-distance analysis and airline network data to project that major international hubs in Belgium, France, South Africa, Kenya, and the UAE represent the highest-risk gateways for the importation of Bundibugyo virus during the 2026 DRC-Uganda outbreak, while clarifying that importation risk does not equate to local transmission risk.

Original authors: Kinoshita, R., Suzuki, M., Yoneoka, D.

Published 2026-06-09

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Kinoshita, R., Suzuki, M., Yoneoka, D.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Imagine a contagious virus, like a very stubborn piece of dust, has started swirling in a specific room in the Democratic Republic of the Congo (DRC) and has drifted into a neighboring room in Uganda. The scientists in this paper are asking a simple but crucial question: "If someone with this dust gets on a plane, where is it most likely to land next?"

They aren't trying to predict if the dust will start a fire (a local outbreak) once it lands; they are just trying to figure out which airports are the most likely "landing pads" for the dust to arrive at.

Here is how they figured it out, using some creative mental shortcuts:

1. The "Effective Distance" Map (Not a Real Map)

Usually, when we think of distance, we look at a globe and measure miles. But for viruses traveling on planes, miles don't matter as much as connections.

The authors built a digital map of the world's airline routes. Instead of measuring physical miles, they measured "Effective Distance."

The Analogy: Imagine you are trying to get a message to a friend. If you have to walk 100 miles to a bus stop, then take a bus to a train station, then take a train, the "distance" isn't just the miles; it's the number of hops and how crowded the connections are.
The Result: A country might be physically far away (like the UK), but if it has a massive, direct flight hub connected to the outbreak zone, it is "close" in this new map. Conversely, a country might be physically close (like a neighbor) but have no direct flights, making it "far" in terms of travel risk.

2. The "Historical Crystal Ball"

Since this is a new outbreak in 2026, the scientists didn't have past data for this specific virus. So, they used a "crystal ball" based on history.

The Analogy: They looked at a similar dust storm that happened in West Africa back in 2013–2016. They asked, "How did that dust travel on planes back then?" They assumed the rules of travel haven't changed much since then.
The Calculation: They plugged the current flight schedules (from 2024) and the historical "travel behavior" of the virus into a computer model to predict where the virus would show up in the next 30 days.

3. The "High-Risk Landing Pads"

The model spit out a list of the top places where the virus is most likely to arrive via airplane.

The Top Contenders: The countries with the highest risk of receiving an imported case are Belgium, France, South Africa, Kenya, and the United Arab Emirates.
Why? These places act as giant "hubs" or "super-connectors." Just like a busy train station where many lines cross, these countries have airports that connect directly to the outbreak zone and then fan out to the rest of the world.
The Numbers: The model estimated that by June 9, 2026, there was a very high chance (over 75%) that the virus would reach major airports in Brussels, Paris, and Johannesburg.

4. What This Does (and Does Not) Mean

The authors are very careful to draw a line in the sand:

What it IS: A warning system. It tells countries like Belgium or France, "Hey, your airport is a likely destination for this virus. You should have your detectors ready and your staff trained."
What it IS NOT: A prediction of a local disaster. The paper explicitly states that just because the virus arrives doesn't mean it will spread.
- The Analogy: If a piece of dust lands on a table, it doesn't automatically fill the whole room. It only spreads if someone touches it and then touches their face. If the "table" (the local health system) is clean and the people are careful (isolation, testing), the dust stays on the table.

The Bottom Line

This paper is like a weather forecast for a virus traveling by plane. It says, "Based on flight paths and how viruses behaved in the past, here are the specific airports where we expect the virus to show up first."

It gives health officials a head start so they can be ready at the gate, but it reminds everyone that being ready doesn't mean the virus will definitely take over the city. It's about awareness, not alarm.

Technical Summary: Early Assessment of Potential Airline-Mediated Importation Risk During the 2026 DRC-Uganda Bundibugyo Virus Disease Outbreak

Problem Statement
Following the declaration of a Public Health Emergency of International Concern (PHEIC) on May 17, 2026, regarding a Bundibugyo virus disease (BVD) outbreak in the Democratic Republic of the Congo (DRC) and Uganda, there was an urgent need to assess the risk of long-distance international dissemination. While no airline-mediated importations had been documented as of May 28, 2026, the incubation period of the virus and the potential for travel during early illness phases presented a plausible risk for global spread. The study aimed to project potential airline-mediated importation risks to support international situational awareness, distinguishing between the risk of importation and the risk of subsequent local transmission.

Methodology
The authors employed a hazard-based modeling framework integrating contemporary airline transportation networks with an externally calibrated importation hazard derived from historical data.

Network Construction: A directed, weighted air-traffic network was constructed using origin–destination passenger-flow data from OAG for May 2024. This network comprised 11,264 airports (nodes), 569,741 routes (edges), and approximately 329 million passengers.
Effective Distance Calculation: The study utilized the "effective distance" metric to model global spread, defined by the shortest path on the airline network rather than geographic distance. The shortest paths were calculated using Dijkstra's algorithm. The primary analysis used the Brockmann–Helbing scale ( $\delta=1$ ), where each travel step contributes one unit of distance. A sensitivity analysis explored alternative $\delta$ values estimated from epidemic growth rates ( $\lambda$ ) and outbound passenger flow ratios.
Hazard Estimation: The arrival hazard ( $h_j$ ) at a destination airport $j$ was modeled as $h_j = \theta / d_j$ , where $d_j$ is the effective distance from the source (Kinshasa/N'djili International Airport, FIH) and $\theta$ is a hazard parameter. The primary value $\theta = 0.19$ was calibrated from the 2013–2016 West African Ebola virus disease (EVD) epidemic. Sensitivity analyses tested $\theta$ values of 0.06 and 0.32 based on the 95% confidence interval of the historical calibration.
Probability Calculation: The probability of at least one importation reaching an airport within $t$ days was calculated as $P_j(t) = 1 - \exp(-t \cdot h_j)$ . Country-level risk was defined as the maximum airport-level probability within that country.
Scope: The primary analysis used FIH as the origin. A sensitivity analysis repeated the process using Entebbe International Airport (EBB) in Uganda as the starting point.

Key Results

Network Dynamics: Effective distance pathways were driven by airline connectivity rather than geographic proximity. The estimated effective distance from FIH to global airports ranged from 3.02 to 29.5 (mean: 15.41).
Importation Probabilities: Within a 30-day window (by June 9, 2026), airport-level arrival probabilities varied significantly, ranging from 0.176 to 0.849 (mean: 0.321).
High-Risk Gateways: Higher probabilities were concentrated in major international hub countries with strong connectivity to FIH. The top-ranked countries by 30-day arrival probability were:
- Belgium (0.777)
- France (0.775)
- South Africa (0.753)
- Kenya (0.741)
- United Arab Emirates (0.721)
- Turkey, Uganda, India, Cote d'Ivoire, and Canada followed in the top 10.
Sensitivity and Robustness: Sensitivity analyses confirmed that while absolute probabilities shifted with different hazard parameters ( $\theta$ ) or starting airports (EBB), the geographical pattern of high-risk gateways remained consistent. The top countries consistently exceeded a 0.64 probability in the primary analysis, with ranges of 0.28 to 0.82 in sensitivity scenarios.

Significance and Claims
The paper claims to provide a reproducible framework for early international situational awareness during emerging outbreaks. By leveraging historical importation experience (the 2013–2016 EVD epidemic) and contemporary airline data, the model offers a method to translate outbreak data into actionable risk projections before actual exportation events occur.

The authors explicitly emphasize a critical distinction: importation risk does not imply local transmission risk. High-probability gateway countries are identified as priorities for awareness and preparedness (e.g., enhanced screening, contact tracing readiness) rather than as locations where local outbreaks are inevitable. The study acknowledges limitations, including the reliance on 2024 network data (which may not reflect real-time route changes or restrictions) and the exclusion of land-based spread. However, it concludes that such projections are valuable for supporting timely international response strategies as additional epidemiological information becomes available.

1. The "Effective Distance" Map (Not a Real Map)

2. The "Historical Crystal Ball"

3. The "High-Risk Landing Pads"

4. What This Does (and Does Not) Mean

The Bottom Line

More like this