Uganda (2007–2008)

The first recognized outbreak of BVD occurred in Bundibugyo District, a predominantly rural district in western Uganda bordering the Democratic Republic of the Congo and situated near the Semliki National Park and wildlife reserve. Although the initial introduction was presumed to have involved wildlife exposure, the source of zoonotic transmission was not established. Transmission remained highly localized: more than 97% of probable and confirmed cases occurred in four of the district’s 10 subcounties, with no sustained transmission documented outside Bundibugyo District.¹

By the end of the outbreak in February 2008, 93 putative and 56 laboratory-confirmed cases had been reported, with 37 deaths, corresponding to an overall reported case fatality ratio (CFR) of 25%. CFR estimates among laboratory-confirmed cases were approximately 40%. Among confirmed cases, the median age was 37.4 years and 54% were male.^2,3

Isiro, Democratic Republic of the Congo (2012)

The second documented BVD outbreak occurred in 2012 in the Isiro area of Haut-Uélé, north-eastern Democratic Republic of the Congo. In contrast to the predominantly rural setting of the 2007 outbreak, Isiro was a regional urban and mining centre with an established health-care referral network. Nevertheless, transmission remained geographically localized, principally in Isiro with a secondary focus in Viadana, approximately 75 km to the south-west.⁴

The outbreak lasted approximately four months and was declared over on 26 November 2012. A total of 62 cases were reported, including five suspected, 21 probable and 36 confirmed cases, with 34 deaths, corresponding to a crude CFR of 54.8%. Among laboratory-confirmed cases, the CFR was approximately 34%. Among 52 cases included in a detailed analysis, the majority were female: 85% of community cases and 61% of cases treated at the Ebola treatment centre were female.^5,6

Ituri and eastern Democratic Republic of the Congo (2026)

The ongoing outbreak is unprecedented in scale among Ebola outbreaks reported in the Democratic Republic of the Congo. As of 16 August 2026, 5021 confirmed cases and 2378 confirmed deaths had been reported. The confirmed-case count alone had already exceeded the combined published case totals of the 16 previous Ebola outbreaks documented in the country (approximately 4800 cases in aggregate), although historical case definitions and classifications were not fully standardized.^7,8,9 Approximately 85% of reported cases remain concentrated in Ituri Province, but transmission has expanded to 56 health zones across six provinces: Ituri, North Kivu, South Kivu, Haut-Uélé, Tshopo and Bas-Uélé. Cross-border spread also occurred, with 15 imported cases and five secondary cases reported in Uganda and one imported case detected in France. Reported case counts underestimate the true burden, although the extent of under-ascertainment remains uncertain.^7,8

Adults aged 18–49 years have accounted for the largest proportion of reported cases, with broadly similar numbers among men and women overall. At least 155 health-care workers had been infected by mid-August, indicating continued occupational and health-care-associated exposure.⁸

Difference in Outbreak Setting

The three outbreaks occurred in very different settings. In 2007, transmission remained concentrated in a rural, mountainous border district. Bundibugyo had experienced displacement related to armed group activity in earlier years, but insecurity was not described as a major constraint on the Ebola response itself. ^1,10 In 2012, Isiro was an urban and mining centre in a region with a history of violence, displacement and poor transport links, but similarly, insecurity was not considered as a main barrier to the response.

The 2026 outbreak is currently unfolding in a different environment. The area where the outbreak is thought to have started, Mongbwalu, is a busy gold-mining area with frequent movement of workers and traders, while Ituri’s capital city of Bunia is a major commercial centre and health-care referral hub. Roads, rivers and mining routes connect affected communities across eastern Democratic Republic of the Congo and with neighboring countries. At the same time, the region is struggling with a humanitarian crisis, with insecurity and displacement throughout the region facilitating disease transmission while simultaneously stymying response activities.^11,8

The setting matters when considering delay. A few weeks of unrecognized transmission in a relatively contained setting do not carry the same risk as the same delay in a highly mobile, multi-province network where response teams may also struggle to reach affected communities.

Delays in Recognition and Confirmation

All three outbreaks have involved a period of transmission before formal recognition, but where in the process that delay occurred and the extent of that delay vary between outbreaks.

In the 2007 Uganda outbreak, an initial alert following cases of febrile diarrheal illness was raised on 2 August, but the investigation was inconclusive. A second alert was raised on 5 November after 20 deaths had occurred in Bundibugyo District. BDBV was confirmed 24 days later, on 29 November, as a previously unknown ebolavirus species. Overall, approximately 119 days separated the first alert and confirmation, reflecting both the difficulty in recognizing the event after the initial signal and the time needed to identify a new virus.^12,1

By the time of the 2012 outbreak in Isiro, the pathogen was known and laboratory confirmation was faster. Yet, recognizing the outbreak remained difficult, as BVD can resemble common febrile illnesses, including malaria. A cluster of unexplained deaths eventually triggered an alert on 2 August 2012. Retrospective investigation linked the earliest laboratory-confirmed case to symptom onset on 28 June, indicating at least 35 days of unrecognized transmission before the alert. Laboratory confirmation followed 14 days later, on 16 August, and the outbreak was declared the following day. Compared with 2007, confirmation was faster; the main delay had shifted to recognizing and escalating the event.^4,6

In 2026, the delay before recognition is hard to determine. A phylogenetic analysis of 139 BDBV genomes estimated the time to most recent common ancestor to be in early to mid-March 2026, suggesting that the virus may have been circulating since early to mid-March, several weeks before the alert on 5 May. Anthropological research [unpublished] has suggested an even earlier start. Once the alert was raised, confirmation took 10 days. Initial GeneXpert testing for Zaire ebolavirus was negative, and broader pan-Orthoebolavirus testing and sequencing at INRB Kinshasa were needed for confirmation. This again points to two issues: recognizing and reporting unusual transmission early and ensuring that negative species-specific tests trigger broader testing quickly.^11,13

The figure shows a clear shift. Confirmation became faster over time, but recognition remained a recurring problem. In the ongoing 2026 outbreak, that delay occurred in a setting with far greater mobility and with insecurity directly affecting response operations, making the consequences potentially much greater.

Figure 2. Time Delays Between Early Evidence, Alert, and Laboratory Confirmation across the Three Documented BVD Outbreaks

Close

Time Delays Between Early Evidence, Alert, and Laboratory Confirmation across the Three Documented BVD Outbreaks.

Implications for Surveillance and Response

Faster laboratory testing and confirmation alone are not sufficient. Surveillance systems must also prioritize the rapid detection and reporting of early evidence of transmission. In all three outbreaks, the time before recognition delayed the response that followed. Clusters of unexplained deaths and illness among health-care workers, observed in these outbreaks, warrant investigation as potential early signals rather than retrospective explanation. When such signals recur, they should trigger faster identification, allowing response activities to begin as soon as possible.

The 2026 outbreak also illustrates why diagnostic algorithms need to look beyond Zaire ebolavirus when initial testing is negative. Unlike in 2007, when Bundibugyo virus was an unrecognized pathogen, BVD is now a known and characterized disease. Faster identification and laboratory confirmation of BVD will trigger a more rapid response.

The current outbreak is therefore not simply a larger version of the two earlier BVD outbreaks. It is the same virus spreading in a different environment. The earlier outbreaks show that delays in recognition are not new, but the 2026 shows how much more consequential such delays can become when transmission occurs across highly connected populations during an active humanitarian emergency.