MR. KOJO NNAMDIFrom WAMU 88.5 at American University in Washington welcome to "The Kojo Nnamdi Show," connecting your neighborhood with the world. Ten years ago this spring the world learned about a new virus that had quietly been sweeping through China infecting hundreds and jumping into international borders.

MR. KOJO NNAMDISARS would eventually spread to thousands of people and kill nearly 800 before it died out. Scientists say China's initial efforts to hide the disease insured its global death toll. But SARS came before tools like smart phones, GPS and social media made it easier to identify and potentially stop viruses in their tracks.

MR. KOJO NNAMDIIn fact, in 1996, it took almost half a year to discover a potential pandemic. By 2009, it took only 23 days. Global virus hunters are using this technology to create programs that are both proactive and predictive in identifying threats in hotspots. It's a job that takes them from the world's most remote jungles to the most advanced labs and it's already showing results.

MR. KOJO NNAMDIJoining us to talk about the way new ways scientists are tracking pandemics is Jonna Mazet. She's a professor of medicine and epidemiology at the School of Veterinary Medicine at UC Davis. She's also principal investigator and co-director of PREDICT. That's a global surveillance system for emerging pandemics. Jonna Mazet, thank you for joining us.

DR. JONNA MAZETThanks for having me, Kojo.

NNAMDIAnd joining us by phone from Harvard University is John Brownstein. He is a professor of pediatrics at Harvard Medical School and also the co-founder of HealthMap, an online tool that monitors disease outbreak. John Brownstein, thank you for joining us.

MR. JOHN BROWNSTEINThank you very much for having me.

NNAMDII'd like to invite our listeners to join this conversation by calling 800-433-8850. Do you think we are in fact prepared for a pandemic? 800-433-8850, you can send email to kojo@wamu.org

NNAMDIJonna Mazet, this marks the 10th anniversary this year of the SARS epidemic and right on cue, we're now grappling with not one, but two cousins of the SARS virus that are circulating in China and the Middle East.

NNAMDISo before we dive into talking about the innovations and tracking these diseases, allow me to read from a news story on Reuters today which says: "The World Health Organization on Monday urged health workers around the world to be on the alert for symptoms of the deadly Middle East respiratory syndrome coronavirus MERS which has the potential to circle the globe and cause a pandemic.

NNAMDIThe United Nations Agency which issued new long-awaited guidance to countries on influenza pandemics said the world was also in the same alert phase for two human strains of bird flu H5N1 which emerged a decade ago and H7N9 first detected in China in March." So can you update us on these viruses called H7N9 and MERS respectively, Jonna Mazet?

MAZETSure, yeah. I think we could think of this as being kind of a scary time, but I prefer to think of it as a time where we're really benefiting from the advances of technology that you mentioned, Kojo. So we have the amazing lessons learned from SARS and ten years later we're seeing China really taking those advances and using them to the benefit of their country and the world.

MAZETAnd I think with that H7N9 avian influenza outbreak that's ongoing right now, we're seeing a rapid influx of information from the world and a push-out of information from China that has probably been responsible in part for the reduction in the epidemic that's going on there and the slowing and maybe the halting of that epidemic.

MAZETWe can contrast that with what we see with MERS, if you will, the coronavirus that we're seeing in the Middle East right now and we're seeing a little less information coming out on that one, which is a little disturbing especially because we're less familiar with these coronaviruses that might be very divergent from each other, which is different than what we see with the better-studied flu.

MAZETIn the MERS situation, we know that it's transmissible from person-to-person and with the upcoming Ramadan and the Umrah, the migration that might come with that we need to kind of think about how we're going to control that and how much information we may have in the world to help us understand that disease better.

NNAMDIBecause millions of people will be descending on Saudi Arabia for Ramadan, is that correct?

MAZETYes, I think so and you never know people might choose to converge at a different time but this is the cultural practice and when we have an infectious disease that we're worried about person-to-person transmission and we don't know where it's coming from then we need to worry.

MAZETThat said, again, with the technologies that are out there now, communication technologies being a big one and our other guest and my collaborator can talk about how that can be really, really helpful. But we also have advanced technologies which are improving our ability to detect and predict where we might see these kinds of outbreaks and hopefully not go into epidemic or pandemic stage.

NNAMDIIndeed, John Brownstein, ten years ago the world did not find out about SARS until it jumped international borders. How would you assess the global response to these newest outbreaks?

BROWNSTEINYeah, it's a completely different world right now and it's really amazing to see the transformation in the flow of information. And it's true ten years ago with SARS there was very limited information coming out from China about this mysterious respiratory condition and only months later did we really recognize it to be SARS.

BROWNSTEINThere was some trickling of information coming through various, some backend channels, through systems like ProMED, which is the International Society for Infectious Diseases system for tracking infection through a network of experts. But really there was limited knowledge of what was happening.

BROWNSTEINYou fast-forward ten years and now the world is completely different with the advent of social media and mobile technology. And the availability of information is going at a speed that we can barely keep up with. If we look at what is happening with China with H7N9 it really is incredible how we were able to track the spread of H7N9 through a social media site called Weibo which is very similar to Twitter that we use here.

BROWNSTEINAnd in fact, in that case, we were seeing people, even patients or friends of patients reporting in about cases and even health records being posted, I mean, in incredible fashion. Obviously some privacy concerns there, but it was actually sort of pushing the envelope and actually the Chinese government was really upfront in confirming these cases, you know, on the hour, you know, that same day.

BROWNSTEINJonna is very correct in that the situation with MERS in Saudi Arabia is a little bit different. There's definitely been a slower progress of information coming out from the region unfortunately and it's made it more difficult to assess risk.

BROWNSTEINSo although, you know, we're utilizing these new technologies and the discussions and news media and in fact the news media and reporters have been an incredible asset to understanding what is happening with these two events. Helen Branswell is an example. The Canadian Press has been an incredible resource that everybody should follow.

BROWNSTEINBut in reality, there's definitely a disparity in terms of the information coming out from different parts of the world. So China seems, to me we were getting real time information. The Middle East, it's probably less so. So the world is changing, but not at the same pace everywhere.

NNAMDIWe're having a conversation about identifying and preventing global pandemics and inviting your calls at 800-433-8850. Do you think we're prepared for a pandemic? What resources do you use to find out about flu outbreaks in your area? 800-433-8850, you can send email to kojo@wamu.org. Send us a tweet @kojoshow Here is Mike in Fairfax, Va. Mike you're on the air. Go ahead, please.

MIKEThank you, Kojo. Recently or two weeks ago, the director of the World Health Organization said the Middle East Respiratory Syndrome raised alarm bells for her. In The New York Times recently, the Centers for Disease Control and Prevention were quoted as saying it would take six months or more to develop a vaccine for the H7N9 bird flu virus and more time to develop a vaccine in the millions of doses for the Middle East Respiratory Syndrome.

MIKEDo either of your guests know exactly what is the CDC actually going to do to protect the American people if either of these viruses start to spread easily from person-to-person and we get a global pandemic in the near term?

NNAMDIAnd frankly, Jonna, we can add to that SARS. We know a lot about it, but it might surprise people that there are still no vaccines or antivirals that have been developed for it or for any of the cousins that Mike was mentioning that are circulating right now. Why?

MAZETYeah, that's true. I think, well, first of all, I don't want to speak for CDC. I do -- we do work with them quite a bit and I do know that they have great people that are working on these issues. But I think part of what is the next phase is we want to move away from creating vaccines which always take time and then the distribution and the safety testing and all of those things.

MAZETWe want to get upstream from that reactive place and technology is helping us do that. Instead of reacting to viruses and epidemics we want to get to the place where we know what might be coming and what the transmissibility factors and pathogenicity factors might be so that we can control those outbreaks at their source like we're seeing with the China situation.

MAZETThat way it's less important for us to be ready and spending billions of dollars on vaccine as it is to controlling the transmission and potentially understanding the reservoirs and so that's what we're trying to do with the PREDICT.

NNAMDIWith the PREDICT program, which was started five years ago, almost five years ago by the U.S. Agency for International Development you work with a huge host of partners including labs, technology firms and international health organizations, the goal being to detect and prevent potential outbreaks before they happen.

NNAMDIThat seems like, well, a Herculean task so how does this surveillance system work?

MAZETSo we work with a combination of factors and, as you mentioned, some of my great colleagues, like John on the line with us today at Harvard, but also we work with lab teams at Columbia University and the University of California San Francisco and then our global consortium which includes EcoHealth Alliance, Wildlife Conservation Society, Smithsonian Institution and Metabiota.

MAZETAnd then we work with about 20 countries around the world, lower to middle-income countries where we have targeted from our modeling hotspots for emergence of new diseases. Now those are lower and middle-income countries because this is a program of the U.S. Agency for International Development, but also the science is pointing us in that direction.

MAZETSo we start with that modeling to get us to look in the right geographical locations and then we work with our teams on the ground that are people from those countries, scientists, educators, ministry officials to really understand the biology and the transmission interfaces where people might be interacting with animals and allowing the spillover of virus.

MAZETAnd that's really important because we're less likely to have a pandemic from a virus that's a human virus with which we have evolved so it generally evolves if it's a smart virus to a lower pathogenicity. We're more likely to see a pandemic from a spillover of a virus from an animal where we haven't seen that virus either at all or too much in our evolutionary history.

MAZETSo when that spills over, we're susceptible and we could have severe disease. And then if it's also transmissible human-to-human we could see it go sort of around the world very quickly. So that's what we're looking for, instead of maybe spending billions of dollars to react to an outbreak and I think the World Bank estimates that we spent in just six zoonotic -- that's animal to human or human to animal transmission of virus outbreaks -- in just six of those we spent about $80 billion between 1997 and 2007. So instead of that, we target the interfaces, the drivers that allow these things to spill over. And then we look in the right place and in the right species of animals and in the people that are interacting with those species. So that's the first step, which is a big step but that's the first step. It allows us to efficiently target.

MAZETThe next step is being able to detect and diagnose these viruses. And then if they become pathogenic they're a disease. And so we need better diagnostics and that's where technology is really helping us. New technologies and diagnostics though, very expensive. So we have a combination of very expensive technologies that help us in these real emergency situations.

MAZETBut we're also putting together sort of diagnostic platforms for mystery diseases that are really inexpensive. And that probably comes from the fact that we're working with a bunch of wildlife professionals who haven't had a lot of funding in the past to work on the diseases, so they find clever ways to look for new things. As well, we're working in lower and middle income countries and so we don't have necessarily the resources in those government ministries to be able to sustain the really expensive platforms, at least not yet.

NNAMDII was going to ask you.

MAZETSo we've -- go ahead.

NNAMDIYou said you're working in a lot of low and middle income countries. What hotspots are you working in right now and does the U.S. have hotspots?

MAZETSo the U.S. might have hotspots. The models show us that there are potential hotspots here in the U.S., but that's a little bit clouded with the fact that we do tons of great research in the U.S. So it kind of clouds our models by making it a little trickier for us to detect the difference between how much effort we put into research and detection and how much there really are those drivers in the U.S.

MAZETI think the more dangerous situation for the U.S. is probably the movement of these pathogens out of other hotspots and into the U.S. And so that's why we're targeting the Amazon Basin, the Congo Basin in Africa and then the sort of Gangetic Plain region of South Asia and then the southeastern Mekong -- southeast Asian Mekong region.

NNAMDIGot to take a short break. When we come back we'll deal also with the technology side of this with John Brownstein. But Jonna Mazet will still be with us and we're looking for your calls at 800-433-8850. You can send email to kojo@wamu.org. We did invite the Centers for Disease Control to participate in the broadcast. But the CDC declined. Question for you, 800-433-8850, do you think we worry too much about the possibility of a pandemic? Why or why not? Or you can shoot us an email to kojo@wamu.org. I'm Kojo Nnamdi.

MAZETWelcome back. We're talking about identifying and preventing global pandemics, inviting your calls at 800-433-8850. Do you think the government is doing enough to prepare for an outbreak, 800-433-8850? Or you can send your questions or comments to kojo@wamu.org. We're talking with Jonna Mazet. She is a professor of medicine and epidemiology at the School of Veterinary Medicine at UC Davis, and also a principal investigator and co-director of PREDICT, a global surveillance system for emerging pandemics.

NNAMDIJohn Brownstein joins us by phone from Harvard. He is a professor of pediatrics at Harvard Medical School, and co-founder of HealthMap, which is an online tool that monitors disease outbreak. John, on the technology side of this equation you're a co-founder, as I mentioned, of HealthMap. How does this tool work? It works to predict and -- it works with PREDICT and the CDC, among many other organizations, to track outbreaks online. How does it work? Where do you get your information?

BROWNSTEINYeah, well, so HealthMap was founded about seven years ago in 2006 with the idea that there was all this information online, all these discussions about events happening around the world, around public health and outbreaks. But they're despaired across a variety of different sources. There was news media and social media and discussion groups, blogs and chat rooms and now social media. But that information was locked away in different parts of the web, but very useful information. Even news media reporters, experts are talking about outbreaks.

BROWNSTEINAnd that was data that if we could harness, it could actually have an impact on global public health because it would shed light on events happening around the world. So we created this tool with our co-founder Clark Freifeld with the idea that if you just mine the web through tens of thousands of sources -- and we do this in 15 different languages -- we could create a world map of disease outbreaks around the world. And that's really what we've been trying to achieve over this time.

BROWNSTEINAnd of course information has evolved and now we rely heavily on crowd sourcing so we have an app called Outbreaks Near Me, which people can report into the system. We even run a system in the U.S. called Flu Near You where people can report weekly about their symptoms. And all these new data sources are providing and shedding light in a new way compared to traditional surveillance sources which often can be underfunded or limited in scope.

BROWNSTEINWhile still they're important, these kind of data streams are actually what are driving a lot of the disease investigations around the world.

NNAMDIJohn, if I log onto HealthMap, what kind of outbreaks could I track?

BROWNSTEINYou can track pretty much anything. We track over, you know, 200 disease events. They could range from, you know, the new coronavirus in the Middle East in H7N9, but to some -- things that are highly local, to food-borne outbreaks that are happening in the community to just general disease events that may only be of importance to the individual or potentially to the traveler.

BROWNSTEINSo really the idea is it's become your weather.com for disease outbreaks essentially. And that's really -- what we're trying to do is build awareness to -- not only to help public health agencies but to build awareness to the general community about disease outbreaks and how people can better prevent disease in their community.

NNAMDIJonna, how did the MERS and H7N9 viruses show up on your radar screens -- it's a question for you too, John -- when they started percolating late last year?

BROWNSTEINYes. So with H7N9 the earliest reports were coming out of very local media sources of China and also the social media site webo where we're seeing lots of discussion on cases. And they started coming up -- popping out three of these sort of what we call informal channels. Eventually we were getting official reporting coming through the Chinese government but this is a very classic situation where we see sort of rumors bubbling up from experts. This was very similar to H1N1 in Mexico where the first reports came out of local media sources.

BROWNSTEINMERS is slightly different in that information was held back for a number of months. In fact, though, the first reports came out of a system called Promed which is a network of over 50,000 disease experts. And one basically -- I wouldn't say whistleblower -- but an Egyptian scientist was actually -- put out this information into this network. And so it really came once again through a back channel communication as opposed through official government.

BROWNSTEINAnd working with PREDICT and Jonna's effort, the idea is that we send this information -- as soon as we get reports coming through these informal channels we send it up the channel to the efforts of PREDICT that have access from the ground that can actually do very detailed investigations. And I'll let Jonna explain to you what they do from there.

NNAMDIJonna Mazet, can you add to that?

MAZETSure. Yeah, so for -- as John said, it's different in both of these situations, but in the China situation with the H7N9 influenza, we are working in China. And we have Chinese partners, including provincial CDCs and the national CDC. And then our primary effort is organized at East China Normal University and the Wuhan Institute, which is a great virological diagnostic institute.

MAZETSo we're working already in the countries. And so what we were able to do, because China was open and because we are working with our Chinese team to work on PREDICT, we're able to add some capacity to what was happening in China already as part of their detection and then surveillance. So we could start to look at environmental factors, look for other flu viruses that might be there together with the H7N9 that might allow it to remix and become more transmissible human to human. So we want to be watching and be prepared for all of those things.

MAZETAlso look and help with the animal side of the picture together with FAO, the Food and Agricultural Organization who is helping out a lot with China. So the Chinese ministry of agriculture working to look at all the live bird markets and chicken farms and really start to get a handle on the epidemiology and how the virus might be moving. And I think that's been really helpful with the control.

NNAMDIJoining -- go ahead, please.

MAZETOh, I was just going to say, with MERS, it was kind of the backdoor, as John mentioned, in that the U.S. government was talking with the Saudi government. And then the Saudi government was also reaching out to some of our team members and saying, what should we be looking for. So in that situation we were able to look at what we were finding as far as new coronaviruses around the world, and give them some information about what we were seeing in other places of the world for new coronaviruses, and compare that to help them target their surveillance. But still they're kind of keeping it a little closer in the country.

NNAMDI800-433-8850, do you worry about catching a virus when you travel? Do you find yourself taking extra precautions against getting sick when you board a plane, 800-433-8850. Joining us now by phone from Cambridge, Mass. is Ruben Juanes. He is a professor of energy studies in the civil and environmental engineering department at the Massachusetts Institute of Technology MIT. Ruben Juanes, thank you for joining us.

MR. RUBEN JUANESHey, my pleasure. Good afternoon.

NNAMDIRuben, we've been talking about some of the new technologies, some of the new tools that have been developed to track and predict disease outbreaks around the globe. At MIT you've created a model which shows how likely the 40 largest U.S. airports are in spreading contagion. What kind of information did you use to create this model? After all, there are a huge number of variations in how people travel.

JUANESYes. There are two key elements to this model. One is to rely on global traffic data. So from airport A to airport B, what the yearly traffic is and that gives us an idea of simply how many people, how many passengers will go through the different segments of the air transportation network. And the other type of data we relied on was individual mobility data. So based on (word?) route taken by individuals we can recreate the individual behavior.

JUANESAnd then we match these two types of data to create a model in which individuals move. And then we can see an outbreak at a particular point at a particular population. And then estimate or forecast the spread of that outbreak.

NNAMDIRuben, you'd think that the biggest airports would be the most influential spreaders of disease. Did you find that to be true?

JUANESSo to some degree, that is true. So again we have data for the 40 largest U.S. airports. And perhaps not entirely surprisingly JFK and LAX came out on top. But we found that traffic alone, such as the number of passengers per year alone was not a good predictor in our minds for the potential for spreading a disease by a particular airport. So we found out that a better estimator would have to take into account not only traffic, but also connectivity of the airport and the ability with which that airport would connect to airports through long range travel. So...

NNAMDII was about to ask, is this a model that you'll be applying to international airports?

JUANESYes, very much so. So at the time we have data for the U.S. bound airports. We now have data for the entire world. And as you can imagine we're actively working on extending this study to the world air transportation network.

NNAMDIRuben, it's pretty scary to watch contagions spread on the video simulation that's on your website. What practical use do you hope comes from this study? And listeners can find a link to that video at our website, kojoshow.org. But go ahead, Ruben.

JUANESYes. So first of all, what I would like to say is that this is very much a physicists view on disease spreading. I'm not an epidemiologist. So really what you see there is perhaps some new insight as to what are the factors, what are the features that make, you know, potential disease spreading important or relevant for potential disease spreading.

JUANESSo our study does not reproduce or replicate any one particular outbreak, so that's the first thing that I should say. But again, what our study suggests is that perhaps the allocation of resources in terms of prevention or rapid mitigation of a disease should perhaps be thought of as being affected by more than just traffic alone. And in particular -- no. We propose a metric that combines long range travel, traffic as well as connectivity. So there's notion of your potential is high if your connected to airports that, say, have a high strength. You're well connected if your neighbor's well connected is very much a part of our metric.

NNAMDIJohn Brownstein, Ruben made the point that he is not a trained epidemiologist. These online tracking programs and tools can be pretty scary if we are not trained epidemiologists. So what are your recommendations to users who would like to separate the noise from the truth when they're looking at outbreaks online?

BROWNSTEINYeah, it's very difficult. There's a lot of information out there. It's very difficult to process this information. That's not to say that untrained people can still not understand what's happening in the world. And we often feel that more knowledge is better. And oftentimes there's this sense that the public can't process sort of complex public health issues. We tend to feel that actually additional knowledge, you know, breeds better sort of understanding and awareness of disease. And, you know, so we tend to be -- push on the side of transparency.

BROWNSTEINOf course the idea of what to do with this knowledge once you have it is very difficult. And of course vaccination, as one caller brought up, is a very key element. But there's -- the traditional public health measures of hand washing and practicing social distancing, those are all important public health practices that everybody can play a role in.

BROWNSTEINAnd so from our perspective, you know, what we try to do in our system is provide all the information that's out there. We do curate it and have public health experts that are looking at this information on a daily basis to make sure that, you know, some potential false alarms are left off. But in general we think that the public can really contribute and be part of the solution.

BROWNSTEINThat's partly why we developed our tool Flu Near You, which means that anybody can be part of public health by contributing how they're feeling, how their family's feeling. We can build a flu surveillance system that's unlike any other. And so we see everybody as an epidemiologist. And it's not category reserved for someone with just an advanced degree.

NNAMDIHere is Dan in Frederick, Md. Dan, you're on the air. Go ahead, please.

DANIt's an interesting conversation. The one part that is difficult is that there's no quantitative stuff going on. I mean, how many pandemics have we had? How many people are affected? You know, it would be a lot easier to get excited about doing more if there was some idea of what the actual threat level was. I like the approach where we're using technology to determine what might happen as opposed to the knee-jerk reaction to crank up, you know, X number of millions and or billions of vaccine unit doses. So, I mean, it'd be nice if we had some numbers.

NNAMDIAnd Jonna Mazet, what do you say to that?

MAZETYeah, I can speak to that. I go back to the World Bank sort of assessment, and their suggestion that we use what we like to call our one-health approach where we think about the people's health as well as animals that could be the -- both the reservoir or also get sick like our food supply, and the environmental factors that make that, if we use that approach, their estimate is that we could save about $6.7 billion a year in our responses to these epidemics.

MAZETSo I'm not able to at this time tell you exactly how many epidemics are going on, and that's really not only because we're not tracking it that way, but because we don't have great surveillance in most of the world, including in the U.S. We have surveillance after we see disease, then we ramp up surveillance. Instead what we want to do, like I mentioned, is get that proactive approach, and with PREDICT, we've been able to identify 200 new viruses that are in families that have caused epidemics and could cause a pandemic.

MAZETSo we've identified these 200 new viruses, we've been able to show that some of those are causing disease in people or animals, sometimes death unfortunately, and so we know those ones we need to really watch out for, and then the majority of the rest of the 200, we're able to say these are ones we need to be now looking for. We need to be ready for. We need to be knowing that they're out there. So if we had known with MRSA pathogen was there, we might have been able to respond more quickly. We might have been able to know how to diagnose it.

MAZETAnd so what we'd really like to do is move to a situation where we can diagnose new things without having to even, you know, catalog them in advance. And so I think that's what some of the technologies are getting us to that place where we can rapidly diagnose new things, and then help to stop them.

NNAMDIAnd Ruben Juanes, we may not all be epidemiologists, but many of us do in fact use airports, so it's better to know exactly where airports -- what role airports may play in this than the guessing games that we've been used to before.

JUANESWhat -- what -- sorry, I...

NNAMDIOn the basis of the studies that you have been doing, I guess we now have a better knowledge of the extent to which airports play a role in this.

JUANESSo I would say so. I think that we're starting to understand better some of the elements that play a role across the board, irrespective in a way of the particular outbreak. And I think there's still a lot to learn. So for instance, one of the key questions that remains in many ways is to what degree these predictions are constantly being affected by the response of the travelers themselves. So this feedback between the propagation of the disease and the propagation of the awareness of the disease and then how the -- how the two feedback each other. I think there's still a lot that remains to be understood.

NNAMDIRuben Juanes is a professor of energy studies in the civil and environmental engineering department at the Massachusetts Technology. Ruben Juanes, than, you for joining us.

JUANESIt's my pleasure. Thank you.

NNAMDIWe're going to take a short break. When we come back, we'll continue our conversation with Jonna Mazet and John Brownstein, and take your calls at 800-433-8850. Do you worry about catching a virus when we travel? Do you think the government is doing enough to prepare for an outbreak? 800-433-8850. I'm Kojo Nnamdi.

NNAMDIWe continue our conversation about identifying and preventing global pandemics with Jonna Mazet, professor of medicine and epidemiology at the School of Veterinary Medicine at UC Davis and principal investigator and co-director of PREDICT, a global surveillance system for emerging pandemics. Also with us is John Brownstein, professor of pediatrics at Harvard Medical School. He's also the co-founder of HealthMap, which is an online tool that monitors disease outbreak.

NNAMDIJohn the Internet and social media have provided new ways to track the spread of disease around the world. We now have Google Flu Trends which estimates spread by counting flu-related Internet searches, but Google Flu Trends was found to overestimate the latest flu season. How reliable are sites that are not getting data directly from the field?

BROWNSTEINYes. It's difficult to validate these types of information that come out of sort of non-traditional sources. We should say that it's still unclear how these systems behave in sort of major flu epidemics. There's definitely a thought that Google Flu Trends overestimated what took place because of changing in behavior and increased interest in flu which led to a spike. At the same time, we don't have great gold standards, and even the surveillance systems that we rely on from the CDC and others also have biases.

BROWNSTEINAnd so we never have a true picture of what's taking place. The best way we do that is through multiple data streams, and yes, we have problems when we rely on search data. We have problems when we rely on hospital data, because not everybody goes to the hospital when they're sick with flu. So what we try to do is just integrate this data into one picture of the flu, and based on laboratory data, hospital visits, data from our mine, we can build a picture of the flu season. And yes, it was a more severe season, but integrating those systems together really is what provides us the best picture.

BROWNSTEINAnd clearly the idea of using the web and social media, it's not a replacement for global public health. We need shoe leather epidemiology. We need astute clinicians and public health infrastructure, but our data through the web becomes very valuable when that kind if infrastructure doesn't exist, and that's really where it can start to fill in the gaps. But really, clearly what we need is all these different data streams to paint the best possible picture of public health.

NNAMDIJonna Mazet, many people may not think that, well, building a highway, a bridge, or a building would prompt an outbreak, but part of your work is looking at how land use impacts the spread of disease. Talk about that.

MAZETYeah, absolutely. So we want to figure out the underlying drivers that allow diseases to emerge, and some of those are the things that bring us into contact with new species, or allow that spillover. So if you can imagine, if people are going into cave that's never been explored before, maybe to get some products that we need for the new cell phone that's going to come out next, they may be exposed to a bat or some other animal in that cave that human beings have just never seen before, and the viruses that are contained in that.

MAZETSo we think about that. It may sound a little, you know, farfetched, but it isn't. If you think the virus -- Nipah virus, and the example that sort of that -- that virus was used to exaggerate into to movie contagion, but really, the urine and feces from bats was carrying this virus, and increased pig farming in Malaysia, and especially in areas where there was mixed with fruit orchards or maybe even they put the pigs under the fruits trees so that the pigs would have shade, and it seems like a great way to do farming, but the bats live in those fruit trees, and they eat the fruit and drop the fruit and then the pigs eat that fruit or get exposed to bat's urine and feces, and that allows the food chain to be infected with this virus, and we saw an outbreak -- multiple outbreaks.

MAZETSo land use is one of those big changes that we're seeing all over the world, including moving new highways into areas right through the middle of forest that's potentially not even been explored before. We have the transoceanic highways going in in South America right now, going right through the Amazon. So we're really potentially putting ourselves at risk, but if we're being careful about it, as John said, using sort of the old-fashioned, but yet most useful hand washing, cook your food well, all of those kinds of public health precautions, together with the technologies of knowing what's there and knowing how to detect and protect us from those things, I think we'll be okay.

NNAMDIHere is Jessica in Washington D.C. Jessica, your turn.

JESSICAHi. I have a question about antibiotic resistant bacteria. I've recently become involved in a case with a dog where they have an antibiotic -- a resistant to the first level antibiotic, and so had to have an antibiotic compound for the infection. Also, the owner is MRSA positive and I want to know what you concern is about antibiotic-resistant bacteria. I also want to know what we're doing to track any MRSA in hospitals and what we're doing about trying to get hospitals to not cover up those issues. So I'm going to take your answers off the air.

NNAMDIWell, don't take it off the air, because I have to make sure that Jonna Mazet knows exactly what you're talking about, because frankly, I don't. Jonna Mazet?

MAZETWell, I do know this subject. I'm certainly not an expert in it, but I am a veterinarian, so I can speak a little bit to the dog-human issue. And certainly the overuse of antibiotics in our world is driving a lot of antibiotic resistance and development, which means, as we use antibiotics, even when we don't need them, we put them out there and we allow the bacteria that are -- maybe were once susceptible to continue to evolve as good bacteria would, to be able to be resistant to those antibiotics, and then we have to get you stronger ones or get out there and develop new ones.

MAZETKojo, you asked me before about dangers of emerging infectious diseases in the U.S., and in much of the higher income world, antibiotic resistance is probably one of the really big worries. And because that's where we have the money for health care, and we aren't very good about saying we went to the doctor or the veterinarian and that doctor or veterinarian sent us home without a drug, because we wanted to be fixed. Now, a lot of times we go in and we have a virus, and we walk away with an antibiotic that's targeted at bacteria, and that helps create this antibiotic resistance.

MAZETSo to the caller's question, certainly overuse of antibiotics in animals can drive resistance for people and the other way around, which is probably even more common over use in people driving resistance in bacteria that can be problems for people and animals. So it is a big issue. It's one of the ones I think for the U.S. we really need to be watching. I think all the listeners can be good about learning not to ask for antibiotics and trusting their medical professionals a little bit better to decide, and we're seeing a huge movement in physicians and veterinarians not to over prescribe antibiotics because of this problem.

MAZETAs far as tracking and hospitals, this one of those problems that people already know about, as I mentioned. We get really good about our surveillance once we know it's a big problem. So it is being done. I think that the sensitivity to MRSA or methicillin-resistant antibiotics, I mean, staph bacteria, is very good right now. As far as how well the communications are going for that, I can't really speak to that.

NNAMDIThank you very much for your call. Jonna, Nature published an analysis last year that painted a pretty bleak picture of how animal flu viruses are being monitored, namely that a tiny percent of flu samples are being collected and not necessarily from the most high-risk places. Obviously, you're trying to lessen that problem in your field work, but can you talk about some of the practical and logistical problems you encounter obtaining samples in the field and getting them to the lab?

MAZETSure. So it's a big task to get out there and safely capture animals. Safe for the people, safe for the animals themselves, and get not, you know, get samples. We swab them, like a Q-Tip to get the virus and then get it back into the laboratory and analyze it. And if you can imagine, it takes many people to collect a big group of birds or bats or rodents. And so it's an undertaking, but it's not just that.

MAZETWhen we think about it, when we think about flu for example, if I take you back to this H7 and 9 flu, the H7 and 9 flu has been known for a couple decades. We knew it was out there, but it was incredibly rarely detected, and actually, my colleagues have recently just done an analysis to show that just to detect it, we probably need to sample about 30,000 wild birds, because we do know it does circulate in wild birds. It's once it spills over to the domestic birds or the birds that are being treated in live bird markets that we start to worry about it, and so maybe we can target some of our surveillance activity there, and that's what happened in this China example.

MAZETIn fact, over 700,000 chickens their blood was tested, and over 200,000 birds swabs were taken. But even with those huge numbers, only 35 serum sample were positive, and only 53 of those swabs were positive. But that little bit of information was enough to cause some changes. Reducing the live bird trade in China, and culling the farms and the live bird markets where the virus was detected, and now we're seeing drop off in human prevalence.

NNAMDIJohn, we're almost out of time, but we talked earlier about Google Flu overestimated the latest flu season. What kind of factors can trip up those models?

BROWNSTEINYeah. Well, in terms of the models that rely on the interest of people and the particular terms that they search for on Google, the idea of Google Flu trends was that, you know, people search for thermometers, they search for cold remedies when they're sick, and that's an indicator of flu, but obviously, if there's a lot of new media, and a lot of discussion about influenza just generally in the news, and people's interest in the subject will increase, those terms will increase. And because we can't confirm that when someone searches for flu, or in the case of when we use Twitter, someone's tweeting about being sick, we can't confirm that they really have flu, there are potential for these systems to go out of whack, and for us to get signals from these data that aren't actually reflective of any pathogen, or last influenza in the population, and...

NNAMDIAnd I'm afraid that's all the time we have. John Brownstein, thank you so much for joining us.

BROWNSTEINThank you very much.

NNAMDIJohn Brownstein is professor of pediatrics at Harvard Medical School and co-founder of HealthMap, an online tool that monitors disease outbreak. Jonna Mazet, thank you for joining us.

MAZETThank you, and thanks for your attention to this important topic.

NNAMDIJonna Mazet is a professor of medicine and epidemiology at the School of Veterinary Medicine at UC Davis. She's also principal investigator and co-director of PREDICT, a global surveillance system for emerging pandemics. And thank you all for listening. I'm Kojo Nnamdi.