I got distracted when I was writing last time about “The New Epidemiologists” because we started our descent into Minneapolis. I wanted to tell you that one of the consequences of new methods to do disease surveillance and data analysis is that we have a batch of “new epidemiologists” that may not have proper public health training at all. That’s not necessarily a bad thing. Lord knows we need a ton of people to address the ton of problems we’re dealing with in public health. Citizen scientists can actually help detect problems early and/or solve the problems that plague us.
It’s a brave new world, to be honest. I’m very excited to see where we’re going to go with all of this.
TRADITIONAL INFLUENZA SURVEILLANCE
Last time, I told you all about disease surveillance and its different tenets. It needs to systematically collect and analyze data, distribute the findings, and something needs to be done with those findings. When it comes to influenza surveillance, the way in which public health agencies has been doing it has not changed a whole lot in the last few decades. It has basically gone through two complete phases and entering a third.
The first phase of influenza surveillance is the one that has been around for hundreds of years. We call it “syndromic” surveillance because it looks for the influenza syndrome (the signs and symptoms of influenza). It’s not very specific because so many other infections mimic the signs and symptoms of influenza, but it can be helpful in making some informed guesses as to what is going on with regards to influenza activity. Or, if you put a program in place to contain influenza, you can use syndromic surveillance to see if your program is working.
However, not just anyone can do syndromic surveillance. To get the best results, you want people trained to identify the syndrome to be the ones counting the cases. In the United States, this is done by a network of “sentinel” healthcare providers who report cases of influenza-like illness to the Centers for Disease Control and Prevention (CDC) every week. The network is called “ILInet”, and it requires health departments at the local level to report healthcare providers who do the reporting to CDC. In my experience, the providers didn’t get much back for their efforts. Most of the time, they got a thank you certificate from the secretary of health.
The second phase of influenza surveillance started when scientists discovered that influenza was caused by a virus, and that the virus could be grown in a culture. As the years went by, other laboratory methods were developed to identify true cases of influenza infection. As I’m writing this, there are traditional virus culture methods, rapid influenza testing methods that you can do at a doctor’s office or a clinic, antibody testing of blood of exposed and sick individuals (though this only detects an immune response and not active disease), and nucleic acid (RNA, in the case of influenza) through tests like PCR. The costs of these tests have started very high but gone done as the technology expands and becomes more attainable. Right now, you can walk into a doctor’s office and get a rapid flu test for about $100, including the visit and specimen collection. In general, state public health labs, do viral culture or PCR, which are far more accurate than rapid tests.
Together with syndromic surveillance, laboratory surveillance helps detect the true cases of influenza and keep a better eye on what is happening in the community. For example, if you see a bunch of people with ILI and they are all testing negative for influenza, you might start looking at other pathogens as the cause for so many people being sick. These two systems complement each other and help public health officials draw appropriate conclusions as to what is going on with regards to influenza in the community. But then game a third phase…
THE MARYLAND RESIDENT INFLUENZA TRACKING SURVEY, BROUGHT TO YOU BY AUSTRALIA
When I first started working at the Maryland Department of Health and Mental Hygiene (DHMH), one of my first jobs was to get the flu surveillance system in fighting shape. The first year that I was there, I added on rapid influenza testing from several urgent care centers and hospitals to the system. (The system was only running ILInet reports.) It was a great complimentary set of data that helped us identify the true number of cases reporting to the different sites. It also allowed us to see what type (A or B) of influenza was active and when it was active. (On average, the flu season ends in activity 6 to 8 weeks after type B becomes predominant over A.)
For the second year, I looked around at different new things that others were trying around the world. I stumbled upon “Flu Tracking,” a program run by the Australian Health Ministry. They signed up a few thousand people all over the country (continent) to report if they had any flu-like symptoms. The best part was that the people reported every week, and enough of them did it to contribute a significant amount of information to the overall surveillance strategy. People who would otherwise not have contact with healthcare — and thus not be counted as cases of influenza or influenza-like illness — were now being counted.
Yes, there are some pros and cons to this strategy, and I’ll discuss them later.
After a few, very expensive, phone calls with the Australian researchers, we put together the “Maryland Resident Influenza Tracking Survey” or MRITS. We began by asking people in the state and local health departments to participate and to invite people to participate. For that first season, we had about 700 people participating. It was just in time, too. The H1N1 pandemic hit at the end of that flu season (May 2009), and we had a tool to estimate how the pandemic strain was affecting people who did not come into contact with healthcare. (It also gave the governor and secretary of health something to hold up to the public and say, “See? We’re looking out for the flu.”)
FLU NEAR YOU
A couple of years after MRITS went online, I got a call from a group inviting me to a conference in San Francisco. The group was the Skoll Global Threats Fund, a foundation devoted to addressing issues around the world that threaten all of us. This particular branch of Skoll was looking into pandemic early warning systems, and one of their projects was called “Flu Near You” or FNY. FNY was very much like the Australian project and like MRITS, so I was invited to give my point of view on what FNY wanted to become.
FNY is part of a larger strategy of what we are calling “participatory epidemiology,” the kind of epidemiology where those who are being looked after have a say in how we do things. It’s kind of how people have weather stations in their backyards and help meteorologists forecast the weather by proving even more data than what is usually available. Or think of it as people with television boxes that report to the networks what the favorite television shows are, or are not.
THE FUTURE OF PARTICIPATORY EPIDEMIOLOGY
The hope of programs like MRITS and FNY is to get the public more involved in telling us, the epidemiologists, what is going on at a very “granular” level. Influenza surveillance has suffered from not being able to tell very well what is going on at level smaller than, say, a county or a city. Sure, we can tell what is going on at a hospital or school, but that’s because those are institutions that have traditionally participated in reporting diseases and conditions to health authorities. With participatory epidemiology, we may very well be able to tell what is going on at a neighborhood level, and we may be able to do so for things well beyond influenza and well beyond infectious disease. (More on that later in October, as I discuss with you my newest thesis idea.)
Of course, there are going to be challenges to this phase of influenza surveillance, just like there were challenges to the other two phases (syndromic and lab surveillance). The biggest one that we have identified so far is poverty, and it affects all phases of public health efforts in general and influenza surveillance in particular. People who are poor or extremely poor don’t have access to internet-connected devices in order to participate, and they have little to no access to healthcare in order to be counted with the “traditional” surveillance systems.
In a more perfect world, if you wanted to truly count all the cases of a disease, any disease, you would make sure that everyone had access to the points of contact where the disease is counted. If the countries in Africa where we recently saw Ebola take such a toll were not so poor, the outbreak could have been detected early and maybe even contained. Don’t believe me? Just look at what happened in the United States with the people who were exposed here. We knew who they were, and we were able to contain them. In the poorest places in the world, the “undesirables” can go for weeks with an outbreak among them before anyone cares. (And it might actually take someone in a higher strata of socioeconomic status getting sick for the world to take notice.)
Could participatory epidemiology be that one thing that gets everyone counted? Not if people have to decide between paying the water bill or paying the internet bill. (Though there could be other ways for people to participate beyond the internet, but the internet is probably the most efficient.) And if they don’t have the education necessary to recognize the need for them to participate in the protection of the public’s health, they won’t know what to do or why they need to do it.
THE FUTURE OF DISEASE SURVEILLANCE
I’m an optimist, so I think that the future of disease surveillance is bright. I would like to think that we are going to overcome the sociological issues that keep people from being counted and/or from seeking the care they need in order to be counted. We are also likely to overcome the technological obstacles and get better ways to communicate, better ways to diagnose disease, and — from all that — better ways to protect the public’s health.
Who knows? Maybe 100 years from now we will all be so interconnected via social networks online that we will be able to know exactly who has what, when, and how they got it… Or we can all be reverted to the stone age because some zoonotic went pandemic and we failed to detect it in time.
THE FUTURE EPIDEMIOLOGISTS
Not to long from today, I envision an even younger, newer group of epidemiologists looking at what happens with regards to disease in the community and, of course, doing something about it. These future epidemiologists will be armed with the latest, greatest technology. Real-time systems will feed them all sorts of data, and they will be well-trained and well-equipped to sort through all of it and reach the proper and necessary conclusions in order to keep the public’s health in good shape.
I met some of them during the “EpiHack” meeting. It feels good to know that there are plenty of people out there who want to save the world and are intelligent enough and committed enough to get it done. It feels even better when they invite you into their inner circle and listen to your ideas. There is still a lot of training of other epidemiologists, future and present, though.
I cannot emphasize training enough. An untrained epidemiologist will be overwhelmed by data and not know what to do with it. What’s worse than that? An epidemiologist that does something completely wrong. The consequences of that go from no harm and no change to a ton of harm and no change. And that’s kind of the worry in these early days of participatory epidemiology… That we get the wrong information and make the wrong decision from it.
But, if fear of failure drove us only to inaction, we as a species would have called it a day a long time ago.