The Rio de Janeiro Olympic Games pose little additional risk of causing new outbreaks of the Zika virus, the U.S. Centers for Disease Control and Prevention recently wrote. The report came on the tails of a letter signed by 2401 international experts, including physicians, academics and bioethicists, asking that the Olympics be postponed or moved, as well as the news that numerous athletes, many of them golfers, have decided to sit these games out because they’re concerned about the mosquito-borne virus.
I don’t envy the athletes trying to figure out their individual risk of getting Zika at the Olympics — which open Friday — even though that risk is relatively low. There’s consensus among scientists that Zika is dangerous for pregnant women, who are at risk of giving birth to infants with neurological defects as a result of the infection, and that it can cause Guillain-Barré syndrome, a condition that leads the body’s immune system to attack the nervous system, often causing paralysis. And it can be transmitted sexually. But there are still a lot of unknowns about Zika, including which mosquitoes can carry it and how long it stays in the body.2
So how did the CDC land on this conclusion? The agency’s modeling wasn’t about whether any of the Olympic athletes will get Zika; the experts assume that some might. Instead, officials wanted to determine whether holding the games risked causing an outbreak in countries that don’t currently have one. (Zika is already spreading without the Olympics: The CDC has travel advisories for more than 50 countries and territories, as well as Miami, where at least 15 people are thought to have acquired Zika locally from mosquitoes.) By the agency’s estimations, most countries’ risk of developing an outbreak as a result of travel to the Olympics by delegations and spectators is slight, either because Zika is already circulating or because so many people already arrive there from Zika-affected countries as part of normal travel.
The rough model used to make this calculation is based on a five-layer cake of data from a variety of sources, including the CDC’s own Zika travel warnings, earlier modeling on Aedes aegypti mosquitoes (believed to be the main carrier of the virus), a country’s potential for a dengue outbreak (the closest proxy the agency had for whether a country is susceptible to a Zika outbreak), international airline logs from 2015 and a video of the opening ceremony for the 2012 London Games.
But of all the variables included in the CDC model, the determination that most countries weren’t at risk essentially came down to a simple equation:
(People traveling from Rio for the Olympics)/(People traveling from all countries experiencing an outbreak if there were no Olympics)
In other words, a 30-person Olympic delegation and 100 spectators don’t present much additional risk to a country that’s already seeing 50,000 visitors a year from areas with Zika.
The CDC team started its research into which countries are at risk of a Zika outbreak because of the games with the 206 countries planning to participate in the Olympics as of June 30. It then removed 39 countries that already had a Zika outbreak at that point. An additional 148 were excluded because they do not have Aedes aegypti mosquito populations that are active in August or September;3 are not thought to have the conditions necessary for outbreaks;4 or have had evidence of Zika outbreaks in the past, which means they likely have some population immunity5 or Zika is already endemic. That would make it hard to rule out sources other than the games in the event of an outbreak. That left just 19 countries.
Of those, however, 15 already have a lot of people traveling from areas with ongoing Zika outbreaks, so the additional risk created by those who are traveling for the Olympics is relatively small.6 That left just four countries at significant risk of spreading Zika back at home as a result of the Olympics: Chad, Djibouti, Eritrea and Yemen.7 The populations of those four countries total about 45 million, but they had only 64 people travel from any area currently experiencing a Zika outbreak in 2015. That makes the estimated 57 people traveling to Rio as part of their Olympic delegations a significant increase. The CDC said that this information was passed on to the World Health Organization and that the WHO is working on specific plans with those four delegations on how to reduce the risk of a Zika outbreak.
I had wondered what it would take for the CDC to decide the Olympics posed a threat to health. To lead study author and CDC health scientist Ardath Grills and her team, the answer was pretty clear: a less global world. If people traveled less, only then would the additional risk of travel to Rio for the games be concerning.
Grills was quick to point out, however, that the team’s calculations are rough estimates. “All the limitations are listed in the back [of the study], instead of up front where I would like them to be,” she said. The government group she works with has a mandate to operate on what she calls “the firefighter model of quick and dirty analytics.” It gets the job done in an emergency — in this case completing the study in only a week — but works only with data that’s readily available, often relying on assumptions for information that would ideally be quantified if there were more time. It’s back-of-the-envelope math compared with the painstaking modeling that epidemiologists perform over months under less pressing deadlines.
“We’re trying to shine a light of what we do know in the darkness of what we can’t know,” Grills said.
The model has many assumptions and limitations, including how the size of each country’s delegation — which includes athletes, coaches, staff and others — was estimated. Those figures were based on a rough formula of the number of athletes competing compared with the number of people participating in the Parade of Nations from the opening of the 2012 London Olympic Games. But it’s hard to say what percentage of each delegation walks in the parade. Perhaps the biggest assumptions the research team made were that the threat of Zika and the Olympics wouldn’t change the overall number of people traveling to and from countries with ongoing Zika outbreaks in 2016 and that travel to Rio for the Olympics would mirror travel to Rio in 2015. Also, the researchers excluded countries with an ongoing Zika outbreak, or where the virus is endemic, even though athletes could carry the virus to communities within those countries that aren’t currently affected.
But most of the modeling is based on worst-case scenarios and therefore overestimates the risk: The researchers assumed that people wouldn’t protect themselves from mosquitoes in Brazil or their home countries and that they would have the virus when they returned. They also assumed that the threat of Zika would be ongoing even though August and September are winter months in Rio, so the mosquito populations are at their smallest.
Another study, recently published in the journal Annals of Internal Medicine, estimated that relatively few spectators and athletes8 traveling to the games would get Zika. The researchers estimated that under worst-case-scenario conditions (in which people have a risk of getting Zika that is equal to Brazilians’ and take no precautions to prevent infection), somewhere from three to 37 visitors will return home infected with the virus, mostly to countries that aren’t at risk for an outbreak or already have one. They also estimated that the entire African continent, which has been the object of much concern for people asking to postpone or move the games, will see a total of about one to 14.5 person-days of infection, the total number of days anyone would have virus in his or her blood, when Olympic travelers return home.