Burying ebola, and other emergent disease

By Tom Stocker @ 2026-05-29T14:47 (+26)

A position from a non-expert mid-career health manager. 
Comments and mutual aid welcome. Opus/Gemini used for search, research synthesis and drafting - but these words are nearly all my own (sorry for you dear reader).

There is an Ebola outbreak unfolding now in eastern Democratic Republic of the Congo. As of 26 May 2026, DRC's health ministry reports 121 confirmed cases (17 deaths) and over 1,077 suspected cases (246 suspected deaths) across Ituri, North Kivu and South Kivu, with confirmed spread to Kampala, Uganda. The WHO has declared it a Public Health Emergency of International Concern and raised its national risk assessment to "very high." This is DRC's 17th recorded outbreak since 1976 — and it began barely five months after the last one ended.

This post argues something the EA community has said in the abstract since its earliest days: organising to fund and support early intervention in epidemics is among the most impactful things we can do.

However, EAs probably also say that ebola won’t scale widely, this already has significant funding (~$350m) and attention to control the disease, and the potential scale of its impact is nothing near the X-risks associated with engineered pathogens.

I'm offering a counter: its just so tractable that we can do a lot here to improve DALY style outcomes, and there are things we can do to learn how to respond better to potential disease-based x-risks too.

1. Ebola hurts

Ebola kills up to 40% of those infected with the Bundibugyo strain, and historically up to 90% for the Zaire strain. It kills fast and frighteningly, and the people most likely to be infected are those who did the most loving thing available to them: caring for a sick relative, or preparing a body for burial. Funeral rites that involve washing and touching the deceased — acts of dignity and devotion — are among the primary transmission vectors.

The harm radiates outward. Households lose breadwinners and caregivers. Survivors carry stigma and long-term sequelae. And the wider community suffers a second epidemic of disrupted ordinary healthcare: during the 2014–16 West African outbreak, the collapse of routine obstetric, malaria and TB services produced excess mortality rivalling Ebola's direct death toll. In the current outbreak, this lands on a population already enduring one of the world's most neglected humanitarian crises — nearly 10 million people across the affected provinces face acute hunger, amid active conflict.

This is suffering of a kind, scale and concentration that a relatively small, well-timed intervention can meaningfully reduce.

2. Funding early intervention is probably extremely cost-effective

The core economic fact about epidemics is that costs compound exponentially while response capacity scales linearly. Every week of delay multiplies the eventual caseload, the eventual death toll, and the eventual price of containment.

A peer-reviewed compartment-model study of Sierra Leone found that intervening just four weeks earlier would have averted roughly 10,000 cases and 8,800 deaths — about 456,000 DALYs. That implies a cost-per-DALY in early response that is plausibly competitive with GiveWell top charities, before even counting the averted macroeconomic damage. And that damage is the larger share: the World Bank's analysis of West Africa found that most economic loss came not from treating the sick but from "aversion behaviour" — healthy people withdrawing from markets, travel and work out of fear.

While there is a lot less certainty than non-emergency scenarios, early epidemic response is plausibly in the same tier as leading global health interventions, and the "go early, go hard" logic holds across multiple modelling approaches and disease types. 

3. Central Africa's outbreak is different

Three differences from West Africa 2014 or DRC 2018–20:

The virus is different — and there's no vaccine, and no treatment for Bundibugyo. Most DRC outbreaks have been Zaire ebolavirus, for which we have two licensed vaccines and a ring-vaccination strategy that worked in 2018–20. This outbreak is Bundibugyo, which has caused only three outbreaks ever (Uganda 2007, DRC 2012, now). LSHTM's vaccine centre calls this a "technology blind spot": commercial and research funding followed the common strain, leaving the rarer one neglected. This is one of the market failures EA biosecurity folk have pointed to — and which Open Philanthropy's Sherlock Biosciences bet and others tried to pre-empt.

Tough times in the DRC The epicentre is Ituri — a high-traffic mining region with intense cross-border population movement, embedded in active armed conflict (ADF, CODECO, M23). Insecurity directly degrades the response: it restricts the movement of surveillance and contact-tracing teams and limits humanitarian access — the precise conditions under which the Sierra Leone modelling showed that rapid, well-resourced scale-up becomes most life-saving.

Detection was slow, and underreporting is high. WHO received the first alert of an "unknown illness with high mortality" on 5 May; confirmation took ten more days, partly because standard rapid field tests are calibrated to Zaire/Sudan strains and can miss Bundibugyo. The gap between ~120 confirmed and ~1,000 suspected cases reflects diagnostic scarcity and likely substantial underreporting — echoing the verification challenges GiveWell-adjacent retrospectives documented in West Africa.

4. We know what activities to fund

The response toolkit is well-established, and with no vaccine available, the WHO has been explicit that strategy "will rely heavily on comprehensive public health measures." things like:

Isn’t our collective ability to deploy all of this relevant to a potential higher-mortality human engineered biorisk?

5. We have some critical information gaps, again…

I can tell you what works. I cannot yet tell you, with confidence, which specific organisation a marginal pound should go to right now for the best expected value, or where a marginal week of skilled volunteering or lobbying does most good. 

I’m embarrassed about this gap, given the movement’s focus on pandemic response as a priority over a decade ago. SoGive's 2020 survey of biorisk orgs found donor attention barely moved even after COVID; we keep re-learning the "panic-then-neglect" dynamic rather than building standing capacity.

Would you, dear reader, be interested in helping to bring together:

  1. Donation routes. There are options for great orgs that tick boxes on safe and dignified burials e.g. IFRC and their local partners, and isolation units like MSF but who don’t have ebola / DRC specific funding routes I can find. There are local NGOs but no associated evaluation of their impact or efficacy. Then there is Alima who may or may not have trust and presence on the ground. Is one of the things we can do lobby for better donation routes in early epidemic responses - or create an active fund?
  2. Lobbying activities. Pandemic response and preparedness advocacy - presumably we want a link between active epidemics and how they are responded to alongside pandemic preparedness and innovation such as https://cepi.net/ ‘s work? Is now the right time? Are there other EA groups, people or orgs engaged in this that want help?
  3. Volunteering avenues. Especially for those of us with some relevant skills - where, how, how much impact?

Please do share your thoughts in the comments.

6. Are we big enough as a movement to build a response capability?

The EA case for pandemic preparedness as scalable, tractable and neglected is old and, I think, correct — Toby Ord's The Precipice puts engineered pandemics among the leading existential risks, and the biosecurity community has produced serious work on surveillance, next-gen PPE, and precision response. But most of it is upstream — prevention and detection. There is comparatively little organised response capacity that can mobilise people and money fast, at the right moment, in the right way.

Conclusion / discussion

We have a live outbreak of a horrible disease with the potential to get much bigger.

We have a lot of uncertainty, but some pointers towards give-well level or better cost-effectiveness if we can interrupt transmissions earlier in the response or do at least as well as the overall NGO effort in the West African ebola epidemic on the margin, as Give Well stated in its blog:

“We estimate that the several billion dollar effort, taken as a whole, could easily have saved lives as cost-effectively as our top charities – something that we don’t believe is usually the case in a disaster”.

We don’t have a clear donation or support route.

The constraints faced by early outbreak responses - funding, insight, coordination - are things many EAs are plausibly well positioned to resolve.

This kind of situation will repeat itself.

There are probably things we can do if we organise ourselves for all outbreaks, using this one as a catalyst.

Tom Stocker @ 2026-05-29T14:53 (+12)

Post script: Personal reflection, and biases-from-experience

I’ve tried to be effective and altruistic for nearly 20 years. I'm no star performer by any stretch of the imagination - but there are probably only a couple of things I’ve done that beat my GWWC pledge’s estimated impact. Both were in a more permissive, emergency context.

The first was supporting with a mere £~1k and 2 weeks during a surge of cholera in south west Bangladesh following a cyclone in 2009. With that money and having a few wits about me, it was possible to pump and clean Shyamnugar’s contaminated water supply (mini reservoir) right in the middle of the outbreak that was cholera contaminated, and add ~20% to emergency clean water provision by providing transport to move existing clean water supply to families waiting for water-aid by the roadside.

This was possible for a non-expert 21 year old, working creatively with local government, and allowing quick partnership and leadership from a local NGO, and Oxfam’s willingness to share their water. Being in the right place at the right time with the right values & resources is something I encourage readers not to under-estimate - but I might be over-egging in my thinking.

I find it more motivating and easier to connect inputs to outcomes in situations like this, than AI or theoretical hazards with very small p values. p=1 for this being a problem, and even if it doesn’t scale to the whole world itself, I can’t think of a better way outside the world of life science R&D to address disease risks as a class than build general capability in responding to them. I am also very sceptical of our ability as humans to anticipate large-scale tail risks effectively AND to address them before they happen - I worry about accidentally making things worse in those situations. In an actual outbreak, I believe there is better and more feedback to work with.

Finally, I joined GWWC in 2013 because I was motivated to make the most difference possible to human flourishing/suffering & my focus is still there.

Guinevere S @ 2026-06-10T14:51 (+1)

Thanks for this. I want to defend the principle in your section 2 and push two threads further: a near-term funding case that has shifted in your favour since 29 May, and the longer-term response-capacity case in your sections 5-6, which I think is where the real opportunity sits. From an impartial view it doesn't change things that this is unlikely to go pandemic, or that the suffering is in central Africa rather than closer to home; it counts the same.

1. The near-term case is stronger than when you wrote it

The "it's already funded" objection is being falsified in real time. On 5 June, WHO and Africa CDC launched a consolidated six-month plan asking for $518m, not ~$350m, and pledges are soft and falling: Africa CDC put them at ~$290m in late May, down from ~$500m a week earlier, with the headline figure still bundling in-kind and repurposed budgets. The structural reason matters for an EA audience: the US withdrawal hit this exact response (health-worker layoffs, supply shortages) and no other donor has filled the gap it left. The "vivid disaster, then money floods in, so low room for more funding" dynamic that made GiveWell cautious in 2014 is not operating this cycle.

That is the sharp point. GiveWell judged even the multi-billion-dollar 2014 response plausibly as cost-effective as its top charities, yet still declined to recommend marginal donations, because governments were flooding in. That single reason is exactly what is failing now. Honest limits: this is no longer an "early" intervention in your section-2 sense, so the case is stronger on room-for-more-funding, not on timing; and whether it matches 2014's cost-effectiveness depends on how large the averted caseload turns out to be, which we cannot yet know. With R still above 1, every week of delay keeps multiplying that caseload. The toll is also worse than a DALY count for the outbreak alone suggests: Ebola disproportionately kills health workers, and each loss thins an already-thin system for years, well beyond Ebola.

Where a marginal donation can and cannot go: no individual can move a shortfall in the hundreds of millions sitting at the sovereign and multilateral level (WHO, Africa CDC, national plans). MSF, the most visible operator, is ~98% privately funded and largely independent of that shortfall, so I doubt a marginal MSF gift is the binding input either, though I cannot show its outbreak-specific position. But the burial-and-community-engagement layer, the most agile and scalable bit and the slice the US defunded, is retail-reachable: IFRC's CHF 29m regional appeal funds exactly this, with a dedicated donation page and ~200 DRC Red Cross volunteers already working in Bunia and Rwampara. The honest gap is not the route; it is that no one has evaluated it GiveWell-style. Which is the bridge to the bigger point.

2. The longer-term case: the response-capacity gap EA keeps re-noticing

Your sections 5-6 are the more important half, and I would push them past "should we build something?", because most of it already exists. That is the more useful answer to your "what's already there so we don't reinvent it" question. The counter-cyclical fund meant to disburse before the appeals are written is the WHO Contingency Fund for Emergencies, and the funding collapse this cycle is partly the story of that fund being starved: it has released only ~$3.9m into this outbreak so far. The layer that places skilled responders into a live outbreak is GOARN, which already deploys epidemiologists, surveillance, data and risk-communication experts within 48 hours. So the gap is not building those two; it is that they are chronically underfunded and disconnected from EA money and people.

What genuinely doesn't exist is the third piece: independent, EA-style evaluation of whether a marginal dollar does most good through the CFE or through an operational appeal like IFRC's. That sits squarely in this movement's comparative advantage. The work is about money, not bodies, and it starts with that evaluation: do it in advance rather than mid-crisis, and money can flow to whatever clears the bar. The failure mode that should worry us is the reverse: channels that would clear the bar never get funded, because no one has done the work to show that they do. That, not new infrastructure, is what "a vetted route from day one" actually takes.

On the x-risk link you raise: I would not claim this outbreak is existentially relevant in scale. Its value is as a drill. Engineered-pandemic preparedness suffers from tiny probabilities and absent feedback, and the most recent real stress test was sobering: COVID was neither the Big One nor engineered, yet wealthy governments still failed on testing capacity, surge coordination and public trust. A live, p=1 outbreak is the cheapest available place to build and test exactly those muscles, plus surge logistics and diagnostics for a strain the standard field tests miss. The transmission dynamics differ, so the clinical specifics don't all transfer, but the operational and trust layer that broke in COVID is the layer a real outbreak exercises. Building response capability this way is plausibly the most tractable thing we can do for the engineered-biorisk class outside life-science R&D, and it pays off even if the tail risk never arrives.

Net upshot: if you're giving now, the IFRC appeal looks like the cleanest retail route; the larger and more durable prize is standing evaluation of these channels, so the next outbreak starts with a vetted route instead of this scramble. Is anyone already doing that, formally or informally? If not, it seems worth a few people scoping, which is the question your section 6 is really asking.

(Disclosure, in the same spirit as the OP: I used Claude for research synthesis and a first draft; the framing and judgements are my own.)

Namana @ 2026-06-08T09:46 (+1)

Am new to EA but have over 7 years experience implementing public health projects this is my 2 cents,global health starts and ends with the community and sadly we tend to overlook the community even here at EA,why has no one ever thought of a cost effective way of funding or doing community health projects?This is the missing piece even with this Ebola outbreak,stop only concentrating on research with data that was not directly collected on the ground,have people on the ground to talk to the community include their input in your research.You will have all the know how and equipment but will not make any meaningful impact,yes you got the numbers but how was the reception?Did you build any meaningful constructive,sustainable bond?If such outbreaks were to occur does the community trust you enough to listen to your advice,or you just dropped whatever and left,was it even needed,do they even know how to use it and use it well?All am trying to say factor the community and everything works out.You wanted to know where to send your donations well am sure the areas mentioned have community based organizations that can be used to directly reach the affected communities with activities like health promotion things like encouraging health seeking behavior,basic awareness of the disease how its spread,handwashing etc all this can be done by using public address systems or the local radio stations,establishing community health units if non exist or strengthening the existing units.

Jmd @ 2026-06-06T09:35 (+1)

My quick thoughts - Do you know if Coefficient giving GHW or Biosec teams have some sort of reactive capacity to look into these things? My intuitive assumption is yes? (Though honestly maybe this got lost because hiring so many new people as they have been doing is a lot of work in itself)