For Epidemics to Cross Oceans, Viruses on Ships Had to Beat the Odds

On Dec. 22, 1874, the HMS Dido arrived in Fiji from Sydney, carrying about 200 people and an invisible payload. A king of Fiji and his son, who were on the ship, were infected with measles. When they debarked, they started an epidemic that killed 20,000 people in Fiji — up to one-fourth of the population — who had no immunity to the disease.

But in those days, when people traveled by sail or steam, such events were the exception rather than the rule. A new report, published last week in The Proceedings of the National Academy of Sciences, uses mathematical models to show how viruses had to beat very long odds to be transmitted across the sea. Most often, the study found, infectious diseases burned themselves out on board before ships ever docked.

In the contemporary world, it is expected that new diseases and older infectious menaces will spread almost instantly around the globe, as happened with COVID-19. But where was the inflection point? Elizabeth Blackmore, a doctoral student at Yale University, and James O. Lloyd-Smith, an ecologist at UCLA, set out to find the moment when viral transmission started to change.

Sign up for The Morning newsletter from the New York Times

John McNeill, a historian at Georgetown University who was not involved in the study, said Blackmore’s use of sophisticated mathematical modeling “has achieved something here that no historian or anybody else has been able to do before — to quantify likelihoods of transmission.”

Kyle Harper, a historian at the University of Oklahoma who was not involved in the study, said the work “breaks new ground.”

Blackmore said she and Lloyd-Smith thought of the idea to look at shipping when she was working on her master’s degree. They learned that the first reports of smallpox outbreaks of smallpox in California were not until 1806 and 1838. And smallpox was first reported much later elsewhere in the Pacific.

She also read “Pox Americana: The Great Smallpox Epidemic of 1775-82” by historian Elizabeth Fenn. Blackmore said she was “amazed to learn that Boston experienced 20- and 30-year intervals between smallpox epidemics throughout the 18th century.”

“Both cases got us wondering about how and why it took so long,” Blackmore said. “And that, in turn, got us to ships.”

She noted that the only way a disease could be transmitted after a ship docked was if there was a chain of infections on board that lasted at least as long as the ship’s journey. Often in the days of sailing ships and even with many steamboats, that simply could not happen. The usual situation was that by the time a ship reached its destination, everyone susceptible to a disease on board had been infected and had either recovered or died.

The researchers considered transmission of three infectious diseases — flu, measles and smallpox — with a mathematical model that Simon Levin, a mathematical ecologist at Princeton University, called “beautiful.”

Flu is hardest to spread because the period of infection is so short — on average, people are infectious for just three days.

Measles, which has an average infectious period of about nine days, and smallpox, whose average infectious period is about 20 days, are more likely to be transmitted because people are infectious for longer periods of time.

Then the researchers looked at the chances of disease transmission on a list of 18 ships, including the Santa Maria, which carried Christopher Columbus to the Americas, and the Mayflower.

They calculated that if one person had the flu while sailing on the Santa Maria in 1492, there would have been a less than 0.1% chance that the disease would be transmitted to the New World. If one person had measles, the chance would have been 24%. For smallpox, 33%. The Santa Maria trip, with 41 people on board, took 35 days, so the limited number on board and the length of the trip contributed to the small chance of disease spread.

The 1620 trip of the Mayflower took longer — 66 days. So, although 127 people were on board, the chance of transmission in the New World was even lower. For flu, it was less than 0.1%. It was 13% for measles and 17% for smallpox.

Yet, ships could be hotbeds of disease, Blackmore said. Contemporaneous reports of conditions on ships in past centuries were horrific — they were suffocatingly crowded and lacked sanitation.

A newspaper report of what happened on a 77-day voyage in 1801 involving a ship called Nancy was typical. The vessel traveled from the port of Sligo in Ireland to New York with 417 passengers, most of whom soon fell ill. Blackmore and Lloyd-Smith included an account of conditions in their study: “Partly from the want of strength and assistance among the sick and partly from the want of a sense of decency, the space between decks, occupied by nearly 300 persons became the receptacle of all excremental matters, insomuch as they issued in streams from the scuppers.”

By the time the ship arrived at its destination, 90 had died and 180 were sick with an unspecified disease, according to their account.

A ship that traveled from Panama to San Francisco in 1851 crammed in as many passengers as possible before it departed, a contemporaneous report said: “It was not until it was ascertained that there was scarcely standing room for those on board that she tripped her anchor.”

People on the voyage were overcome with “fever and dysentery,” the report said.

The chance that a ship’s passengers might spread a pathogen took a sudden upturn in the middle of the 19th century, when steamships accelerated and transformed travel.

Although that makes intuitive sense, Levin said, the researchers’ mathematical analysis “shows why and makes it quantitative.”

Blackmore and Lloyd-Smith saw the effects of steamships when they looked at steamships traveling to San Francisco from 1850 to 1852, during the Gold Rush.

The steamers were much faster and carried many more people than earlier sailing ships — a median of 196 on trips from Panama, and a maximum of 1,050. Sailing ships carried a median of 53 and a maximum of 287.

And there were many more steamship trips on the Panama route to San Francisco than trips by sail. Using their mathematical model, the researchers calculated how likely it was that passengers on a steamship to San Francisco who had flu, measles or smallpox would still be infectious when they docked.

Only ships coming to San Francisco from the closer destinations such as Panama had short enough voyages to ensure that some passengers would still be infectious.

The prospect of transmitting flu remained less than 0.1%. But there was a 70% chance of transmitting measles and a 74% chance of transmitting smallpox.

The risk of transmission grew during World War I, the researchers found, when extremely fast ships could carry 1,000 to 1,500 troops to the battlefronts. Historians say that troop movements were one reason the 1918 flu spread around the world so quickly.

A surgeon general commander in the Royal Navy during World War I told of the extreme crowding in ships. When the men lay in their hammocks, their heads were “less than three feet apart.” Large numbers of men were enclosed in an area “less than one-fiftieth of a square mile.” The report concluded that “everyone on board receives a dose of the infectious agent sufficient to cause influenza.”

Blackmore is continuing this thread of research for her doctoral work at Yale, looking to test her models against “real outbreaks on real ships” described in historical archives.

“Part of what motivates me is that if we are going to convince ourselves that things like social distancing would work, it is important to have examples of infectious disease spread instead of saying ‘It spreads like wildfire,’” Blackmore said.

“We need to talk about how slowly a disease can spread,” she added.

c.2024 The New York Times Company