Hantavirus won't be the next COVID

A forecaster's breakdown of the Hondius cruise ship outbreak

About the author: Peter Wildeford is a top forecaster, ranked top 1% every year since 2022. Here, he shares the news and analysis that informs his forecasts.

-

A cruise ship trapped at sea.

Passengers with a strange, infectious disease, forced to isolate.

A diplomatic fight over which port would receive the vessel.

Sounds like the coronavirus. But this time it’s hantavirus.

Is this a big deal?

The Andes-strain hantavirus cluster aboard the Dutch-flagged expedition vessel MV Hondius is real and tragic. Three people have died. Several more are critically ill.

But hantavirus is not the “next COVID”. The way the pathogen works does not suggest that sustained spread is possible, the lack of open-ended opportunities for spread makes sustained spread even harder, and the current trends over the past 34 days since the first case suggest we are strongly on track for containment. I will say, with confidence, that this will not spiral into a new major pandemic.

To be more concrete, we can turn to Metaculus, an online forecasting platform, which currently has two questions on the hantavirus:

• The first asks “Will WHO declare hantavirus a Public Health Emergency of International Concern before 2027?” — as of the time of writing, the median of community predictions is at 2% and I am currently forecasting 0.4%.

• The second asks “Will at least 5 non-passengers be linked to the MV Hondius hantavirus outbreak before August 2026?” — as of the time of writing, the median of community predictions is at 23% and I am currently forecasting 4%.

How am I so confident? Let’s look into what hantavirus is, what happened on the Hondius cruise ship, and where we might go next.

What happened?

This hantavirus incident started not on a cruise ship, but with a Dutch couple on a bird-watching trip through Chile, Uruguay, and Argentina. After this five-month trip, they boarded the Hondius cruise ship, for what was meant to be a month of sailing. The first five days went exactly to plan — zodiac excursions, evening lectures on polar ecology, and trips to see penguins and ice. But on April 6, while onboard, the husband got sick and died April 11. With only a small ship’s clinic to work with, the cause of death couldn’t be determined. The voyage continued.

On April 24, the ship reached Saint Helena, and the widow disembarked with her husband’s body for repatriation — already carrying stomach symptoms that nobody yet recognized for what they were. Three other Dutch passengers got off with her. The widow flew to Johannesburg, deteriorating throughout the flight. As she was boarding the plane for Amsterdam, the crew realized she was too sick to travel and pulled her off the plane.

The virus causes Hantavirus Pulmonary Syndrome, which targets the lungs and heart, and has a mortality rate of 35 to 50 percent. She died in a Johannesburg hospital on April 26.

Later, a 28-year-old German woman aboard the Hondius got a fever and then died on the ship on May 2. Oceanwide Expeditions called the Dutch authorities, who called the World Health Organization. The Hondius went to its highest internal pandemic response, with cabins sealed and meals delivered by crew.

At 5:30am yesterday local time, the ship anchored off the Canary Islands. More than 70 passengers were then ferried ashore in protective suits on small boats, then put into sealed buses to the airport, and flown out to seven different countries.

However, the seal wasn’t perfect. A French passenger cleared by the ship’s doctors nonetheless developed symptoms on the flight home. And a man who left the Hondius weeks ago on Tristan da Cunha — population 200, the remotest inhabited island on Earth — is now hospitalized there with probable hantavirus, kept alive by oxygen and PCR kits that British paratroopers airdropped on Saturday.

So far, as of the time of writing, all hantavirus cases currently cluster in a 25-day window from Apr 6 to May 1, the last on-ship onset was Apr 28, and there have been zero new onsets in the 9 days since May 1. Hantavirus incubation runs 1–8 weeks, so we’re not in the clear yet.

But so far twelve days of contact tracing across twenty-two countries has turned up only Hondius passengers as having hantavirus infections. The KLM flight attendant who handled the dying widow: negative. The passengers seated near her: negative. All eight confirmed or probable cases were on the ship.

The Hondius is sailing on toward Rotterdam, with the German woman’s body still aboard.

Why the outbreak almost certainly contains

Here’s my best attempt at graphing a comparison of hantavirus vs. COVID-19 based on the 34 days so far — take note of the log scale:

Of course, this is a bit of an apples-to-oranges comparison. COVID’s first 34 days had massive reporting lag, so instead I retrospectively estimate actual infections in Wuhan rather than what was officially reported at the time¹. We may then reasonably ask: for hantavirus, could there be a quiet unknown spread too? Maybe, a few years from now, we will realize that there ought to have been more hantavirus cases reported for the first 34 days, like with COVID?

Seems unlikely. Hantavirus is in a closed system — currently only in one ship with strong post-hoc investigation. Additionally, with hantavirus there’s not known to be any spread without active symptoms. Compare this to COVID-19, which started in early Wuhan — an open system with no surveillance. Together, this means the current known case count of hantavirus is very likely close to the actual count.

But even if you wanted a strictly fair “what was officially reported by Day N” comparison, COVID had already reached over 44 confirmed and reported cases by day 34, whereas hantavirus has not yet exceeded 10. COVID-19’s doubling time in the unrestrained early Wuhan phase was somewhere around ‘cases double every 4–7 days’; whereas hantavirus on the Hondius has effectively no compound growth. The slope difference is far more notable than the absolute numbers. Put simply, hantavirus empirically lacks the exponential spread that COVID had.

So why is hantavirus not going exponential? Well, the three biggest factors that made COVID such a problem is that it was (a) quite deadly, (b) quite contagious, and (c) transmitted easily before symptoms appear making it difficult to isolate the right people.

The bad news about the Andes virus is that it is also very deadly. But in the roughly thirty years since person-to-person Andes transmission was first observed, no outbreak has ever escaped the close-contact networks in which it began. Andes virus is simply not that contagious and infection comes nearly entirely from those who are symptomatic.

I built a probabilistic model decomposing this question into exposure pools — the widow’s KLM boarding, the Airlink flight from St Helena, the JNB hospital, household contacts of disembarked passengers — applying Andes-specific transmission and incubation parameters, then conditioning on the observation that zero non-passenger cases have surfaced as of today. Running this, I get a 4% chance that there will be more than 5 cases from people outside those on the Hondius cruise, with a 70% chance of at least one non-passenger case by August, and about a 17% chance of three or more.

How hantavirus spreads

Hantaviruses live in rodents and are typically spread by inhaling parts of rodent poop that end up in the air. There are twenty known hantavirus species, but only one — the Andes virus — has documented human-to-human transmission. That discovery came from a 1996 outbreak in El Bolsón, Argentina, where treating physicians and family members of patients fell ill without any rodent exposure of their own.

The Andes virus is specifically from the long-tailed pygmy rice rat, native to southern Argentina and Chile. The Dutch couple that brought hantavirus onboard the Hondius almost certainly got it by accidentally inhaling rat poop in a rural setting somewhere in Patagonia in November.

Andes infection moves in two phases. It starts with a flu-like illness — fever, muscle pain, headache, gastrointestinal symptoms — that lasts three to seven days and is usually initially mistaken for something more mundane. But the Andes virus makes blood vessels start to leak, and then the blood pours into the lungs. Patients can move from ‘feeling off’ to requiring ICU care within mere hours. This is the phase that kills people; it is also the phase during which they are most infectious, which matters a great deal for what happens next.

Typically you can only get the infection if you’re in ‘close contact’ with someone who already has the disease. But when we talk about ‘close contact’, what does that mean? The phrase appears in every WHO and CDC document, and it does most of its work as a vague reassurance rather than a precise epidemiological parameter.

The detailed literature on hantavirus and the Andes virus is messier than the public messaging suggests.

Here, our story starts with a birthday party in Epuyén, a small town of about 2500 in Argentine Patagonia, where the Andes virus is found. The party had about 100 guests, an evening of food and drink and proximity. After attending, one of the guests developed flu-like symptoms. He died days later. By the time hantavirus was confirmed, several of his fellow partygoers were already sick. The chain spread through their households, including to pregnant women and their unborn children, before Argentine epidemiologists had stitched it together. There were 34 cases before the entire chain was able to be contained.

This is what was challenging for the notion of ‘close contact’ — some of the exposures were in fact quite casual. One secondary hantavirus case had spent only “a few moments” with an infected partygoer on the way to the bathroom; another had merely shared a short car ride. These weren’t the prolonged household contacts the literature usually pictures when it says “close contact.”

Alonso’s team estimated the number of people each an infected person infects is about 2.12 in the early, unrestricted phase. This is an infectious rate comparable to flu. But this average of 2.12 hides what’s important. Nearly all 34 cases traced back to just three superspreaders — people who had unusually high viral loads and certain immune profiles. Most of the infected individuals in Epuyén transmitted to no one at all. So the outbreak’s shape was not a slow grind of average people infecting average numbers of others like you see with a typical flu but instead a handful of extraordinary transmission events embedded in a sea of otherwise dead-end infections.

And critically, the infectious window of hantavirus is short and centered on visible illness. Andes virus infectiousness peaks on the same day a patient develops fever and drops sharply within days. Unlike COVID, pre-symptomatic transmission is minimal. By the time someone is contagious enough to seed a superspreading event, they look sick with a fever — which is why isolation works at all. This is the profile of a virus public health can contain.

This year’s events on the Hondius fit this picture. Onboard transmission is now officially attributed in part to person-to-person spread. But the case count — six confirmed and two suspected cases out of ~150 people sharing a closed environment for several weeks, including the thirteen-day period after certain exposure and before isolation measures were enacted — is meaningfully below what an Epuyén-style superspreading event in a continuously-mixing cruise environment would have produced. The most plausible read here is that the Dutch couple that started the infection were not superspreaders and most secondary exposure occurred within cabin pairs that were ‘close contacts’. This is far more consistent with a contained cluster rather than the start of a pandemic.

Additionally, unlike a wildlife-spillover event with anonymous spreaders scattered across an open population, every Hondius passenger and crew member is known by name, by nationality, and by current location. Twenty-two governments are coordinating. The 1–8 week incubation window means we’ll know by mid-July whether anyone slipped through — and we’ll know about each case the moment it surfaces. This may be one of the easiest disease outbreaks to monitor and contain.

What comes next?

Regardless of what happens next, the Hondius outbreak is a real tragedy. But tragic and “next COVID” are not remotely in the same category.

In some sense, the conclusion is not yet written. Time will tell whether I look on-point or incredibly stupid. While it’s always possible there could be some crazy mutation, these mutations are very rare and we have no evidence of this having happened.

Fundamentally, hantavirus is deadly but slow. It shouldn’t go exponential or fundamentally change the trajectory of society the way COVID did.

1. I’m calibrating to Imperial College’s January 2020 retrospective nowcast estimate of 410 cases (95% CI 130–900) in Wuhan as of January 4, 2020 — Day 34 since the first known symptom onset on December 1, 2019 per Li et al. NEJM. There were 44 COVID cases officially reported on this day; certainly an underestimate given what we now know. Compare either of these two estimates to 8 Hantavirus cases (6 confirmed, 2 highly plausible) in the first 34 days.