Four Months After the Yellow Fever Alert
What Is Happening Off the Map

← Back to Latin America News

On 13 March 2026 the Pan American Health Organization (PAHO) issued an epidemiological alert about sustained yellow fever transmission in South America. What drew attention was not the case count but the geography: human cases had been confirmed in São Paulo state in Brazil and Tolima department in Colombia, neither with any prior record of transmission, both far from the Amazon basin long treated as the endemic core. Four months on, the southern hemisphere has entered its dry season. Here is what those months revealed.

From the March alert to WHO's June update

According to PAHO, seven countries in the Americas reported 346 confirmed cases and 143 deaths during 2025, a case fatality rate of 41%. Brazil accounted for 120 cases and 48 deaths and Colombia for 125 cases and 46 deaths — together roughly seventy percent of the regional burden. In epidemiological weeks 1 to 7 of 2026 alone, four countries reported a further 34 cases and 15 deaths. In an outbreak update published on 24 June, WHO reported that six countries — Bolivia, Brazil, Colombia, Ecuador, Peru and Venezuela — had confirmed 79 cases between January and May. More countries and more cases than the March snapshot, but a gentler slope than the steep climb of 2025; WHO now assesses the risk as moderate in regions with a history of transmission.

Colombia: an extreme concentration in Tolima

Colombia carries the heaviest burden. The health ministry puts the outbreak that began in September 2024 at 201 confirmed cases and 89 deaths. Of those, 53 cases and 26 deaths fall in 2026, and 52 of the 2026 cases are in Tolima alone; the remaining one was reported in Villavicencio, Meta. Cumulatively Tolima accounts for 180 cases and 74 deaths.

In late June, ahead of the mid-year holiday season, the ministry identified 162 municipalities as very high risk and recommended a booster even for people vaccinated more than a decade ago if they were travelling to specific municipalities such as Chaparral. The vaccine is free at over 3,000 sites nationwide. Bogotá administered 162,885 doses in the first half of 2026, yet as of 30 June the city had confirmed 16 infections linked to exposure in Tolima, six of them fatal. Those who fell ill were not residents of the outbreak zone — they were visitors to it.

Brazil: the numbers fell, but did not vanish

Brazil's picture differs. São Paulo state confirmed 57 cases and 35 deaths in 2025; in 2026 it stood at 11 cases and 6 deaths as of June. A steep decline on its face. But nine of those cases cluster in the Vale do Paraíba region, and the small municipality of Lagoinha alone accounts for roughly 80% of the state's cases this year. And according to the state health department, every confirmed case involved someone who had not been vaccinated. The outbreak has not so much shrunk as condensed into the places where the immunity gap remains.

Why it happens where the map says it shouldn't

Yellow fever in South America is not an urban person-to-person disease. It circulates in the forest between monkeys and mosquitoes, and people are infected when they enter that cycle — the sylvatic pattern. The vectors are not Aedes aegypti but canopy-dwelling mosquitoes of the genera Haemagogus and Sabethes. The risk map therefore follows forests and primates, not population density.

When logging and conversion to farmland move the forest edge, both the monkey troops carrying the virus and the places where people enter the woods shift with it, while a changing climate pushes mosquito ranges to higher elevations. The Bosque de Galilea in Tolima is exactly the kind of highland forest once assumed to be safe. Then the immunity gap compounds it: in areas never classified as endemic, yellow fever vaccination was never part of routine practice. People outside the line on the map were never urged to get vaccinated, and when the line moved they were unprotected. The transmission mechanics differ from dengue in the Americas, but the underlying problem is the same: the map has gone stale.

The debate: rewriting the procedure itself

The yellow fever vaccine confers long-lasting immunity after a single dose and is highly effective. That people keep dying is a delivery problem, not a medical one. PAHO asks countries to sustain coverage above 95%; press reports indicate Tolima has yet to reach that threshold. In Colombia the ongoing EPS crisis is straining frontline operations and could complicate campaign delivery.

There has been movement, though. On 25 May, PAHO convened specialists in epidemiology, entomology, clinical management and immunization from more than ten countries in Bogotá to review revised regional guidance on yellow fever surveillance. Using a Delphi process they updated the operational definitions of suspected cases and of outbreaks, and validated a matrix organising interventions across pillars — surveillance, vector control, clinical care, immunization, risk communication — differentiated by transmission cycle (sylvatic or urban) and by phase (preparedness or response). That is an attempt to rewrite the decision procedure rather than to run one more campaign.

A personal view

There is a connection here that is easy to miss from Japan. Yellow fever is essentially the only disease for which countries may still require proof of vaccination from arriving travellers under the International Health Regulations. Costa Rica is one of the countries that does. Arrivals from Bolivia, Brazil, Colombia, Ecuador, Guyana, Peru and Venezuela, among others, must present the international certificate — the yellow card.

Before I first travelled to Latin America I was vaccinated against yellow fever at a quarantine station in Japan and received that card. In Japan the vaccine is available only at a handful of designated sites, by prior appointment. Yet arriving in Costa Rica directly from Japan, there is almost no occasion to show the document; it matters only if you route through South America. The same person, the same body — and the administrative treatment changes with the itinerary. Requiring a certificate from an individual at a border and recommending vaccination to residents of a risk zone use the same vaccine under entirely different logics.

Studying Japan's assistive-device benefit system, I run into this shape constantly. Eligibility has to be drawn somewhere; those inside the line are served generously, those outside often never hear the system exists. The line is not malice, it is what makes administration possible. The trouble starts when reality moves faster than the line. That is precisely what yellow fever has done — the virus has outrun the definition of a vaccination target area. Most of those who died in Tolima were not in some place the system had failed to imagine; they were simply there before the map was updated.

Three things to watch from here: when PAHO formally publishes the revised regional surveillance guidance, how close Tolima gets to 95% coverage, and the next season. Yellow fever in South America rises with the southern rainy season, roughly December through May, so the real verdict will come in the figures from late 2026 into early 2027. Whatever coverage is built during the dry months is what the next wave's death toll will reflect.

Glossary

fiebre amarilla / febre amarela = yellow fever, in Spanish and Portuguese. ciclo selvático = the sylvatic cycle, transmission circulating between monkeys and canopy mosquitoes. epizootia = an epizootic, an outbreak among animals; monkey deaths serve as a leading indicator of human outbreaks. certificado internacional de vacunación = the international certificate of vaccination, commonly called the yellow card.

When people die of a vaccine-preventable disease, the question is not medicine. It is how delivery was designed.

※ This article is the author’s commentary based on public information. Please confirm the latest figures, dates and procedures with governments and primary sources. Quotations are kept minimal and sources are cited.