The first person we know of started running a fever on the 24th of April, in a place called Rwampara. She was a health worker, which is how these things usually begin: the people who lean closest to the sick are the first to fall. Within days the deaths in her province had a pattern to them, and the pattern had a name no one in the room could say yet, because every test they had came back negative: not Ebola Zaire, not dengue or malaria or cholera, not plague or mpox or COVID. Four health workers were dead inside four days before a laboratory could tell anyone what had killed them.

The lab caught up on the 14th of May, when scientists at Congo’s National Institute of Biomedical Research in Kinshasa ran thirteen blood samples out of Rwampara and watched eight of them light up for a filovirus that was not the famous one. Within a day it had a species name, Bundibugyo, and the World Health Organization had opened its formal outbreak notice on the disease.

If you have not heard of Bundibugyo, you are in good company. Until this spring it had surfaced exactly twice since it was first identified, in Uganda in 2007 and in the DRC in 2012, both times briefly, both times contained. It was the kind of pathogen that lived in review articles and biosafety inventories, a line item under “other filoviruses,” a footnote to its celebrity cousin. As of the 7th of September, that footnote had produced 6,778 confirmed cases and 3,269 deaths, a case fatality ratio of 48.3 percent in the DRC, spread across 61 health zones in six provinces. It is the second-largest Ebola outbreak on record, behind only the West African epidemic that began in 2014, and it has already killed more people than any Ebola outbreak except that one. It built that record in about four months.

Read the timeline forward. On the 2nd of June there were 378 cases and 63 deaths. By the 21st of August there were 5,458 cases and 2,606 deaths, roughly five thousand new cases in a hundred days, and the international response machinery had been running the whole time. The tools on the shelf were built for a different virus, and the surveillance never caught up to this one. Both failures were baked in before the first sample reached Kinshasa.

Start with the tools. There are licensed countermeasures for Ebola, and they work: Ervebo, the vaccine, and the monoclonal-antibody drugs that turned a near-certain death sentence into a survivable illness. Every one of them was built and licensed for Zaire ebolavirus. The world had a filovirus vaccine ready. It had the wrong filovirus.

That is not a rhetorical flourish. It is the WHO’s own position, stated in its own outbreak notice. Ervebo is being used in Congo right now, but only “within research protocols,” and as of the 6th of September 2,007 people had been vaccinated with it. In the same document the agency concedes that “available evidence remains insufficient to support the programmatic use of Ervebo for the prevention” of Bundibugyo disease. Follow the chain: a vaccine licensed against Zaire, no evidence that it holds against Bundibugyo, and so it goes into arms under a research protocol rather than a program. Two thousand people in the middle of a 48-percent-lethal epidemic are receiving a shot the agency deploying it cannot say works against the virus they are being vaccinated against. It may cross-protect. It may not. A review by Boston University virologist Nancy Sullivan in the New England Journal of Medicine this month makes the same admission from the other side of the table: there is no licensed vaccine or therapeutic that specifically targets Bundibugyo, and the hope that Zaire-targeted tools will hold the line is exactly that, a hope. A separate therapeutic trial, called PARTNERS, only began enrolling patients on the 2nd of July, six weeks and hundreds of deaths after the outbreak was confirmed.

Now the surveillance. The CDC’s own August field report grades Congo’s containment against its own targets, and the report cards are failing. Alerts investigated within 24 hours: 83 percent, against a target above 90. Contacts traced per case: 10.6, against a target of at least 20. Health zones with safe burial teams: 49 percent, against a target of 100. And the number that tells you where the transmission is happening: 59 percent of deaths occurred outside treatment units, against a target of 0. People are dying at home and in the community, uncounted until they are gone, which is precisely how a filovirus outruns the surveillance built to catch it.

The virus had help from the ground it landed on. The epicenter, the Mongbwalu area of Ituri province, is gold-mining country, a place of transient labor, displacement, and funeral rites that involve washing and handling the dead, every one of them a transmission opportunity a laboratory in Kinshasa cannot touch. The first public signal that something was wrong there was not a clinician’s report but a social-media post on the 5th of May about “unusual deaths of unknown cause.” By then the fever in Rwampara was eleven days old. Because Bundibugyo opens with the same acute fever, vomiting, and diarrhea as the malaria or typhoid that fill those clinics every ordinary week, the first patients were treated as something else while the specimens they needed crawled toward a lab that could tell the difference.

The spillover itself appears to be new, a fresh jump from an unknown zoonotic host rather than a reactivation of something old, which means the reservoir is still out there and no one can yet say what it is. The disease has also stepped outside Congo’s borders: twenty cases in neighboring Uganda, one imported case in France, two Congolese patients flown to Germany for treatment. The Uganda cases are the ones that should keep people up at night, because they are local spread across a border, not a patient carried out on a medevac flight.

Nineteen years ago a virus was named after a Ugandan district and then largely forgotten by a preparedness enterprise that funded the threats it could already picture and skipped the ones it could not. That is the warning Sullivan’s review draws out of the wreckage: prepare only for the familiar pathogen, and the rare one that reemerges finds you with the wrong vaccine on the shelf. The people who paid for that choice are not in the agencies that made it. They are the four health workers in Rwampara who died in the same week, before anyone with a sequencer could tell their families what to write on the death certificate.

Sources

  1. WHO: Disease Outbreak News, Ebola disease caused by Bundibugyo virus, DRC (as of 7 September 2026)
  2. CDC MMWR: Characteristics and Monitoring of the 2026 Bundibugyo Virus Outbreak, DRC, August 2026
  3. CDC MMWR: Notes from the Field, Outbreak of Ebola Disease Caused by Bundibugyo Virus, DRC and Uganda, May 2026
  4. PMC: Operational epidemiology of the early phase of the 2026 Bundibugyo virus disease outbreak, DRC and Uganda
  5. ScienceDaily: reporting on Nancy Sullivan’s NEJM review of the Bundibugyo response
  6. New England Journal of Medicine 2026;395(3):278, Nancy J. Sullivan, “Bundibugyo Virus Disease in 2026, Clinical and Public Health Responses”