Management of Ebola: Acting Before Arrival and the Closing Window for Containment
The 2026 Ebola outbreak in eastern Democratic Republic of the Congo has produced over a thousand suspected and confirmed cases across three Congolese provinces, a sustained transmission chain in Kampala, an evacuee in a Berlin isolation unit, and suspected cases testing negative in Milan and Bengaluru. This is the visible part. There is also a large invisible part: people who are infected but whose infections have not yet been identified, either because they have not yet developed symptoms, or because surveillance has not recognized them, or because they have not been traced as contacts of a known case. The invisible part needs our full attention. What we fail to identify, we lose the opportunity to act upon. What we fail to act on becomes the next wave of cases.
Western nations have been reassured, as they were in 2014 and again in early 2020, that advanced healthcare systems can manage imported cases. However, this reassurance does not address the crux of the problem. By the time a case reaches a hospital, the moment to use the levers of control—the brake, the accelerator, the steering wheel—has passed. The outcome of the evolution of the outbreak has already been made upstream: whether the outbreak was detected early at the community level, whether populations in the affected zones were screened before they traveled, whether the transit hubs they reached were buffered from the rest of the world, and whether the travel measures applied to those hubs were real or symbolic. These are not medical decisions. These decisions are not made by hospitals, by physicians, or by vaccine manufacturers. They are made by governments, days and weeks before the patient arrives at the emergency room door. These decisions are being made, or not being made, now. Offloading responsibility to hospitals is akin to insisting that medics alone should prevent war casualties.
Structural aspects of the situation at hand
The disease travels the global transport network to anywhere and from there everywhere. In such a situation, chasing it with contact tracing cannot succeed. The reason is twofold: the scale and connectivity of the world’s transportation network on the one hand, and the inherent nature of contact tracing on the other.
The world’s transport network
The outbreak is caused by the Bundibugyo virus, for which no licensed vaccine or approved treatment exists, meaning containment must rely primarily on non-pharmaceutical interventions and basic clinical care rather than pharmaceutical protection. The primary affected region is a mining and trading hub with intense informal mobility across Uganda and South Sudan, while Uganda’s capital, Kampala, already has its own transmission chain and is connected by daily flights to Dubai, Doha, Addis Ababa, Nairobi, Mumbai, Bengaluru, London, Brussels, Istanbul, and onward to virtually every major global city. With international air travel now roughly fifty percent higher than in 2014 and East African hub connectivity growing even faster, the effective response window for policymakers has dramatically narrowed.
Historical evidence also shows that airport symptom screening detects almost none of the infected travelers who matter: Thomas Eric Duncan traveled from Liberia to Dallas undetected in 2014, and COVID-19 spread worldwide through screened airports in 2020. The significance of any single detected case is that it implies that if there is one detected case, there are many that are not yet detected, and the next week there are two, then four, then eight… Transmission expands exponentially rather than linearly. The cities most exposed are not necessarily those dominating Western pandemic narratives—but those most densely connected to Uganda’s capital Kampala and to one another, many of which are largely absent from international media coverage. The structural problem is simple, and it has not changed.
The nature of contact tracing
The capacity of any health system to chase down contacts is finite. One case can generate a thousand contacts. A hundred cases generate a hundred thousand. A thousand cases generate a million contacts. No system can track that many. The arithmetic does not care how dedicated the contact tracers are. In October 2014, Fanta Kone, a two-year-old girl traveled twelve hundred kilometers by public bus from Guinea through Bamako—the capital of Mali—to the city of Kayes. She was bleeding from her nose during the journey. By the time Malian authorities identified her as having Ebola, six of the ten passengers on her bus could not be found. They were never found, despite tracing over 100 people who had been in contact with the child. In another incident just a month later, a man from Guinea arrived at a clinic in Bamako with Ebola. He died there. Identifying his contacts required tracing 332 people. He had already infected a nurse, who infected others, including a doctor. Several of them died before the chain was closed. The two cases; the child in the bus and the man in the clinic, required tracing hundreds of contacts each. The current outbreak has over a thousand cases, thus far the third largest Ebola outbreak in history. And next week the number doubles. Even with perfect organization, the numbers alone guarantee that tracing will fall short. This is why running behind an outbreak does not work. It is a question of arithmetic, not of effort or competence. There is no version of this calculation in which chasing wins, but fortunately it is not the only tool we have, as we will see next.
Tools at our disposal
The aforementioned facts do not determine an outcome. They describe the terrain on which a choice is being made. Much of science takes the world as given and calculates what follows from it. That is not the task here. The task here is to identify the desired outcome—an outbreak contained at its source—and to identify the required actions. The actions themselves exist and are well known. They were tested in Liberia in late 2014 and they worked. Community-level early detection thrives over individual contact tracing, once the chain is too large to chase. Population-zone segmentation with buffer regions between exposed and unexposed populations shields non-infected communities. Real travel measures at hubs that have been reached, help when applied to everyone, not just to those without the resources or status to be exempted.
Alternatives to this failed in 2014, and they failed for COVID-19 in early 2020. Failed procedures include travel measures that exempt citizens, residents, and transit passengers, precautions relying on symptom screening at the destination, and ones that constrain the source country to handle exit screening without support. When a system depends on finding every contact, missing even a small fraction is enough to keep the outbreak going. On the other hand, successful measures to prevent a widespread disaster mean suspending entry for all but a narrow set of necessary exceptions, who go through mandatory quarantine.The measures include source-country exit screening with adequate isolation capacity. They consist of the buffer-zone architecture that succeeded in West Africa, applied at an international scale. These are choices that governments can make. The reason to make them now rather than later is that the cost of waiting is paid in cases who expose their contacts, leading to an explosive multiplication of economical, political, and human harm.
Finally, the management of this outbreak should take into consideration the customs and culture of the affected populations, in order for these populations to support the implementation of these measures. For instance, in Congo, burial rites are of high importance, where family members gather around and often come into contact with the deceased. As such, the body of the deceased is important for these populations. Not managing the expectations of these populations with regard to containment of the cadaver can lead to societal unrest and disruption to the containment measures. This can lead to outbursts of violence as have already been witnessed.
The cost of inaction
It is tempting to hope the outbreak will burn itself out. Fires do burn out—when they run out of fuel. But a fire with fuel ahead of it does not die down. It accelerates. Anyone who has watched a wildfire knows the shape of it: a line of flame creeping along a ridge looks containable. But in the absence of firebreaks it makes a run, doubling and redoubling faster than crews can reposition, and a fire that covered a hillside at midday has taken the whole valley by dark. An epidemic that doubles every week is this kind of fire, not the dying kind. In the early ebola outbreaks the case count doubled weekly, and it is doubling in the DRC now. Doubling is deceptive precisely because it feels slow at the start and becomes unstoppable at the end while never changing its rate. Two cases become four, four become eight—small numbers, easy to wave away. But the same doubling that turns two into four turns two thousand into four thousand, and then four thousand into eight thousand. The response that would have worked at two cases cannot be built fast enough at eight thousand. The quiet early weeks are not a reprieve. They are the fire before flashover—the only time when it can still be put out, and the time when it looks least like it needs to be.
There was a point where one might think that the outbreak would burn out on its own, that is before there are hundreds of cases. The question is “does this fire burn at all” and we might say some outbreaks burn out themselves. We are well past that point today. Policy now must be based on prevention if we act early. The right comparison is not between acting now and a world where the problem simply disappears, but between acting now and being forced to react later under far worse conditions. The matter hinges on implementing a containment strategy that minimizes harm on a global scale.
In practice, this means three kinds of action can be instigated: first, building early community-level detection so that “unknown” infections shrink and triggers for response are clear. Second, segmenting populations with buffer zones—akin to firebreaks—so that outbreaks are contained where they start, instead of bleeding silently into every network. Third, applying simple, even-handed travel measures at key hubs so that rapid connectivity does not turn a local crisis into a global one. Taken together, these steps give governments a real, if brief, window in which to choose a controlled, time‑limited disruption now instead of being forced into a long, uncontrolled disruption later by the arithmetic of doubling time.
Positive signals to build on
Two recent policy actions are important seeds of effective action. The United States’ establishment of a receiving facility in Kenya for exposed U.S. individuals recognizes that transport inherently increases risk of spread. Focusing only on U.S. individuals should be generalized to what is needed: quarantine and monitoring facilities near the point of origin to prevent international dispersal before cases are identified. Any global dispersal is dispersal everywhere. Uganda’s border restrictions with the Democratic Republic of Congo point to the critical importance of border control. A few cases in Uganda may be overcome. However, uncontrolled importing of unidentified cases from DRC will overwhelm such efforts. But standard border closures are unlikely to be sufficient in regions where substantial movement occurs through unofficial crossings on motorbikes on dirt paths, making rapid deployment of all means surveillance and control measures essential. New technologies such as drones may have high leverage for effective control under these conditions.
In a nutshell, if we wait, the growth of cases will force our hand. We urge governments to take the helm now, while they still can.
References
3) https://apnews.com/article/ebola-congo-uganda-border-virus-b96734598ea95b1cdb71986c8b1adf43
5) Associated Press, Ebola treatment tent set ablaze again in Congo, with 18 suspected cases leaving, Canadian Broadcasting Corporation, May 23, 2026 (https://www.cbc.ca/news/world/ebola-tent-fire-congo-9.7210001)
Further reading:
6) Philippe, C., Bar-Yam, Y., Bilodeau, S., Gershenson, C., Raina, S. K., Chiou, S.-T., Nyborg, G. A., & Schneider, M. F. (2023). Mass testing to end the COVID-19 public health threat. The Lancet Regional Health – Europe, 25, 100574. https://doi.org/10.1016/j.lanepe.2022.100574
7) Siegenfeld, A. F., & Bar-Yam, Y. (2020). The impact of travel and timing in eliminating COVID-19. Communications Physics, 3(1), Article 1. https://doi.org/10/ghh8hg
8) Wong, V., Cooney, D., & Bar-Yam, Y. (2016). Beyond Contact Tracing: Community-Based Early Detection for Ebola Response. PLoS Currents. https://doi.org/10/ggwd3w
