We Didn't Fail to Stop Ebola Bundibugyo — We Chose Not to Make the Vaccine for 19 Years

Looking at the short-term picture — the next one to three months — the honest assessment is discouraging. The Bundibugyo Ebola outbreak in the DRC has almost certainly not yet reached its peak case count. Given the diagnostic blind spot created by GeneXpert's inability to detect BDBV, official case numbers are likely a significant underestimate of actual transmission; field epidemiologists working in environments where multiple febrile illness causes compete for clinical attention routinely identify substantial underreporting in the early outbreak phase. Eastern DRC's combination of active conflict zones, severely limited road infrastructure, and population displacement makes contact tracing and isolation extraordinarily difficult even when the surveillance architecture is functioning properly. I estimate roughly a 60% probability that cumulative confirmed cases double or triple from current levels within three months of the PHEIC declaration. WHO will issue emergency funding appeals, but based on the documented timeline from the 2018–2020 Kivu Ebola response — where the average gap between initial funding appeal and actual field disbursement was eleven weeks — that money will not reach operational teams fast enough to close the current containment gap. In practical terms, MSF and DRC local health workers will continue to bear disproportionate operational burden while the international funding machinery catches up to the urgency on the ground.

The border spread scenario is the single highest-acuity near-term development to monitor. Bundibugyo was first discovered in Uganda in 2007, and Eastern DRC shares a deeply permeable, high-traffic border with Uganda, South Sudan, Rwanda, and Burundi. According to UNHCR estimates, tens of thousands of people cross the DRC-Uganda border daily through both formal checkpoints and informal routes that are essentially impossible to monitor under normal conditions, let alone during an active outbreak response. Suspected cases have already been reported in Uganda, and if a confirmed cluster emerges there, the response complexity escalates exponentially — every additional affected country means navigating a distinct national health system, separate regulatory framework, different political decision-making chain, and varying baseline health infrastructure. I estimate the probability of confirmed cross-border spread to at least one neighboring country by August 2026 at roughly 35–40%. Factoring in South Sudan alongside Uganda, the probability of at least one country with confirmed cases by year-end exceeds 50%. Those are not comfortable numbers, and I want to be clear that 35–40% is not a low-probability tail risk — it is a near-even coin flip under current conditions. Eastern DRC's overlap with active armed conflict zones, including M23 operations, additionally limits health response options in ways that have already materialized as treatment center attacks in previous outbreaks.

In the medium term — six months to two years out — the most consequential question is not the DRC outbreak's trajectory but whether this crisis catalyzes genuine structural change in global health R&D investment patterns, or whether it follows the well-worn path of every previous Ebola outbreak and fades from international agendas once the acute crisis pressure eases. CEPI's formal elevation of the pan-Ebola vaccine strategy is directionally correct, with stated targets for entering Phase 1 clinical trials by 2027. I think that timeline is optimistic to the point of being misleading. A pan-Ebola platform must demonstrate cross-protective immunity against at minimum Zaire, Sudan, and Bundibugyo ebolavirus. Bundibugyo's glycoprotein surface structure differs from Zaire's by approximately 30–35% at the amino acid sequence level — a meaningful distance that means the rVSV-ZEBOV platform underlying the currently approved Zaire vaccine will not transfer without substantial engineering modification. My realistic estimate for Phase 2 clinical data is no earlier than 2028–2029, with regulatory approval most likely in the early 2030s. In the meantime, there will be at least one, probably multiple, additional Bundibugyo outbreaks managed with zero species-specific vaccine or therapeutic tools. The science is finally moving; the timeline simply does not match the urgency of the problem.

The mid-term funding picture presents a structural challenge that may prove more consequential than the vaccine development timeline. USAID's withdrawal is not merely a financial event — it represents the structural exit of the world's historically largest bilateral global health aid donor from the surveillance architecture it helped build over decades. The question of whether Europe and Japan can compensate is, in my assessment, essentially already answered in the negative. The EU is channeling unprecedented resources into defense spending following the restructuring of European security dynamics, and internal political pressure around migration is crowding out discretionary health aid budgets. Japan faces accelerating demographic aging costs that create persistent fiscal pressure on overseas development assistance expansion. The World Bank's Pandemic Fund — announced with considerable political momentum following COVID-19 — held less than $2 billion in total capital as of 2025, an amount structurally insufficient to sustain infectious disease surveillance networks across 54 African countries with their combined 1.4 billion people and dozens of high-priority zoonotic spillover zones. No combination of alternative donors currently available can fully replace the U.S. contribution, which historically covered roughly 25–30% of global infectious disease surveillance infrastructure funding. This means that for at minimum two years, and realistically longer, the global health early warning system will operate with a structural deficit that affects not just Ebola but every high-consequence pathogen monitored through shared surveillance networks: Marburg, Nipah, Lassa fever, Crimean-Congo hemorrhagic fever, and others all depend on the same degraded infrastructure.

The long-term view — two to five years out — offers a picture that is conditional on how this outbreak plays out on the world stage and in high-income policy environments. If Bundibugyo spread remains geographically contained to the DRC and perhaps one or two neighboring countries with sub-catastrophic total mortality, I estimate the probability of sustained, structural reform in neglected tropical disease R&D investment at roughly 15–20%. That range reflects the historical base rate of post-Ebola structural reform: the 2015 agenda produced real but incomplete institutional change; post-Kivu reform produced even less. The reform impulse consistently correlates with the degree to which high-income policymakers experience personal proximity to the threat — which is exactly the dynamic that allowed Bundibugyo to go unvaccinated for nineteen years in the first place. If the outbreak expands to multiple countries and generates the kind of visible, high-mortality crisis that dominates international news for sustained periods, the probability of genuine structural change rises meaningfully — perhaps to 30–35% — but still does not become a likely outcome even under that scenario. Breaking a structural cycle requires changing the incentive architecture, not just the narrative.

The structural bright spot in the long-term analysis is Africa's own emerging biopharmaceutical manufacturing capacity, and I want to give it the credit it deserves while being honest about its timeline. South Africa's Afrigen Biologics, designated as the WHO's mRNA technology transfer hub following COVID-19, represents a genuine institutional inflection point in the history of vaccine equity — the first time a major vaccine production technology has been deliberately transferred to an African institution with a regional mandate. The African Union's Partnership for African Vaccine Manufacturing has established a target of 60% continental vaccine self-sufficiency by 2040, and if that target is even approached, it changes the economics of pathogen-specific vaccine development in a fundamental way: pathogens like Bundibugyo that are commercially unviable for global pharmaceutical companies could be developed and produced to serve regional epidemiological demand, bypassing the market logic that has failed the continent for decades. The vision is right and worth building toward. The honest assessment is that Africa currently produces roughly 1–2% of the vaccines it uses, and closing that gap requires technology transfer, cold-chain infrastructure at continental scale, regulatory harmonization across 54 national regulatory bodies, and trained biopharmaceutical workforce development — a multi-decade project under even the most optimistic resource scenarios. The PAVM trajectory offers hope for the 2035–2040 timeframe. It offers no relief for the DRC in 2026, 2028, or likely 2032.

Three scenarios bracket the realistic range of outcomes. The bull case — roughly 15% probability in my assessment — sees the 2026 outbreak become a genuine policy inflection point: pan-Ebola vaccine development is accelerated with committed multi-billion-dollar public funding at Operation Warp Speed scale, WHO emergency response procedures are formally revised to reduce PHEIC activation time for non-mainstream pathogens, and cross-border surveillance networks receive durable, multi-year replacement funding commitments that partially offset the USAID structural withdrawal. The base case — roughly 55% probability — sees the outbreak contained within the DRC and one or two neighbors by year-end, a meaningful but incomplete international response that falls short of structural reform, CEPI's pan-Ebola program proceeding on a slower-than-projected timeline, and surveillance gaps persisting through at least 2027–2028 before alternative funding architectures can be partially assembled. The bear case — roughly 30% probability — sees multi-country spread across three or more nations, the absence of species-specific diagnostics and vaccines allowing outbreak scale to exceed anything seen in the region since 2014, fragmented and delayed international response due to the degraded surveillance and funding architecture, and a preventable humanitarian catastrophe that demonstrates in the most costly possible way what the structural underinvestment of nineteen years actually costs. I want to name the bear case probability explicitly: 30% is not a tail risk. It is a one-in-three outcome that should make policymakers deeply uncomfortable.

Two wildcard factors could materially shift these projections in ways that my base case doesn't capture. First, if field data confirms that the existing Zaire Ebola vaccine provides meaningful cross-protective efficacy against Bundibugyo in emergency use settings — some preclinical data has shown partial cross-species protection, and regulatory agencies are actively evaluating emergency use authorization options — even 40–50% efficacy in high-risk healthcare workers and close contacts would substantially alter near-term mortality trajectories and shift the short-term scenario toward the bull case. This is a genuine wildcard because the data could arrive relatively quickly, and its clinical impact could be disproportionately large relative to its technical imperfection. Second, if CRISPR-based field diagnostic platforms — SHERLOCK, DETECTR, and related approaches — achieve field-deployable scale faster than conventional diagnostic development timelines, the architecture for adding new pathogen sequences to these systems is dramatically more flexible than cartridge-based platforms, meaning a Bundibugyo-specific rapid diagnostic could reach prototype stage within six to twelve months. I hold these wildcards with genuine openness but also genuine caution: technology advancing faster than expected does not automatically mean technology reaching the communities that need it faster than expected. COVID-19 taught us that vaccines can be developed at record speed and still reach Africa twelve or more months after high-income countries. The speed of scientific innovation and the equity of access to that innovation are two completely separate problems that require two completely separate solutions. Solving the first is necessary but not sufficient for solving the second.