Millions at risk of Zika infection

As the US declares its first "home-grown" cases of Zika in Florida, we've learned that up to 90 million people, including over one and a half million childbearing women, may be infected across the Americas as the initial wave of the Zika epidemic unfolds. Andy Tatem from the University of Southampton took Chris Smith through the numbers...

Andy - Back in February, WHO declared Zika as a public health emergency, but there were no real estimates at the time of the scale of the problem and that's useful for decision makers for planning surveillance, for working out outbreak response, for working out what the demands might be on their health services. Normally, to produce these kinds of estimates you would make use of data that's coming as the endemic is going on, but often using those traditional approaches and using that data the damage can be done because the epidemic is ongoing. So we leveraged data relationships from previous epidemics that have been spread by the same mosquito, viruses such as dengue and chikungunya. We know they have relationships with things like temperature that we can utilise to be able to make predictions across the region that we're interested in, which is the Americas in this case.

Chris - And because these viruses are inhabiting the same territory, they're spread by the same mosquito vector, the same factors that influence the spread of those viral relatives will probably impact on the spread of Zika. And that means you can then make more realistic predictions about where it might go and over what  sort of timeframe?

Andy - Yes, certainly. We're assuming that Zika will be spread in a similar fashion, we know it's spread by the mosquito that has spread previous similar viruses, so we can make those assumptions at the start of the epidemic. And, obviously, as the epidemic goes on we'll have more data on Zika itself and see any differences that exist. But in the rapid spread of this virus, we need answers quite quickly to ascertain the size of the epidemic and be able to react.

Chris - The numbers which were being put forward at the time when the WHO made its announcement and, subsequently, have been really rather large; we've been talking about millions of potential cases or exposed individuals. What does your approach suggest - are those numbers reasonable?

Andy - Our estimates are of the scale of 93 million people across the first wave of the epidemic over two to three years and of those 93 million infections, we're estimating around 1.6 million childbearing women.

Chris - And where are these 93 million people?

Andy - They are spread across the Americas, and we were making predictions at 5 by 5 kilometer grid cells. So we are making really upper limit predictions of, if the virus made it to that location, what was the kind of upper limit that you would expect in an epidemic.

Chris - Where are the hotspots?

Andy - The hotspots are really around Brazil; it has suitable climate; we know it has high densities of the Aedes aegypti mosquito that spreads Zika. It's also in some of the Caribbean islands and there's been good evidence of big outbreaks of these kind of viruses on small islands before.

Chris - And if we extrapolate the number of infections that you're seeing, what are the implications for what people seem to be, rightly so, most worried about with this virus which is its impact on women who are pregnant?

Andy - I think there's still a lot of uncertainty there in terms of the relationship between getting an infection and birth defects. We are estimating here around one and a half million women giving birth who may be infected. There's a range of estimates of the proportion of those that might give birth defects and those numbers have been quite widely varying and are very small samples at present. So I think, as the epidemic goes on we'll know a lot more about that.

Chris - And what's your thoughts on the situation in Brazil given that we're about to start the world's most important sporting event there, and something like half a million to a million people will come to Brazil to support and compete?

Andy - I think there is a need to be vigilant, to improve and keep up surveillance. There's obviously a need for mosquito control in some of the most visited and highest mosquito density areas, and I think there's still a lot of uncertainty. Obviously there is a worry about people coming from all over the world taking not only Zika back, but other diseases and spreading those across the Americas. And at least now we have some estimates of, if that virus is taken to new locations of what might be the scale of the local epidemics that might occur.

Chris - Because, of course, the data you're using in order to make your predictions is based on things like dengue virus, but it based on the natural cycle of dengue, not when there's an Olympics going on, isn't there? So might there be a bit of a change to the outcome because of the Olympics? 

Andy - Yes. I mean we don't account for human movement here. We are estimating if the virus were to get into a location, what might be the scale of that epidemic. If we have good information on the number of people who are moving between locations; if we have good data on how that's connected to other parts of the world in terms of numbers of people coming from different countries, we can then extend to make estimates of how that virus might spread to new locations, and not only Zika again but other diseases.