Archive
2014
January
March
April
May
June
July
2013
January
February
March
April
May
June
July
August
September
November
December
2012
January
February
March
April
May
June
July
August
September
October
November
December
2011
January
February
March
April
May
June
July
August
October
November
December
2010
September
October
November
December
Jul. 20, 2012

Using Twitter to Put Disease Outbreaks on the Map

by A. Tomasulo, C. Freifeld, J. Brownstein

Click to enlarge images
Many of the millions of people who visit London for the Olympics might bring home more than just souvenirs, since mass gatherings are an opportunity for travelers to swap infectious diseases. The founders of HealthMap explain how monitoring online discussions in all kinds of forums--from news feeds to Twitter--can lead to early detection of disease outbreaks, thereby giving health officials a chance to quickly respond to imminent epidemics.
 
About 16 years ago, public health organizations such as ProMED and Canada’s Global Public Health Intelligence Network began exploring Internet-based disease surveillance. These systems searched for infectious disease outbreak information and provided alerts to subscribers. In 2006 we founded HealthMap at Boston Children's Hospital, a freely available online platform that mines informal Web-based sources (local media, mailing lists, chat rooms, etc.) and user-submitted reports to track infectious disease outbreaks worldwide and in real-time. Today, thousands of users access HealthMap daily. Equally important is how the HealthMap system produces its real-time view of infectious disease.
 
{"input":{"width":"150","photo":"app","row":"4285","table":"DOCUMENT"}}
First, to acquire data, the HealthMap team carefully developed language specific search strings to sift through news aggregators (such as Google News, AllAfrica, Baidu, and others), RSS feeds, mailing lists, and discussion forums to collect all data relevant to disease outbreaks. HealthMap also relies on direct reporting from users as a form of crowd-sourcing information about disease outbreaks. Users report via the web, Twitter, Facebook and even through the Outbreaks Near Me mobile application. Data is then submitted to a characterization engine that automatically determines the disease category and location of the outbreak based on extensive disease and location dictionaries. Next, the HealthMap system uses Bayesian filtering to determine the relevancy of each disease alert, and the likelihood that a given news item is reporting on an ongoing outbreak, an imminent health threat, or simply irrelevant information. Then, alerts are grouped together by outbreak. The end product, http://healthmap.org, is a highly organized data set that allows public health officials, travelers, and curious minds to see a real-time image of infectious disease activity worldwide. And nearly the entire process is automated.
 
The structural hierarchies of traditional public health reporting, while important, introduce delays. A typical public health report may move from local health worker to local official and then to a laboratory. Once a pathogen is confirmed, the report moves to national and international officials. The entire process can take weeks. And within weeks, thousands of people could be sick or dead. Meanwhile, strange symptoms may be tweeted or covered by local media. This real-time information can indicate emerging threats.
 
This contrast was evident in 2009. While much of the world focused on the threat of avian influenza in Asia, HealthMap detected a disease alert via the astute reporting of journalists in La Gloria, Mexico of acute respiratory infections in Veracruz, Mexico—the beginning of the H1N1 pandemic. Through informal, Web-based reports, HealthMap was able to track the pandemic from the beginning through its global spread.
 
Over the years, the platform has proved useful to many different organizations, specifically during times when the risk of infectious disease outbreaks is heightened. In 2010, HealthMap teamed with Crisis Mappers and Humanity Road to create a map tracking cholera outbreaks in post-earthquake Haiti. The resource, http://healthmap.org/haiti, proved useful during a humanitarian crisis as it provided information on clean water sources, health facilities, and location of outbreaks.
 
{"input":{"width":490,"photo":"haiti","row":"4285","table":"DOCUMENT"}}
 
Among the many reasons disease risk is heightened during humanitarian crises is that catastrophic events represent an aggregation of people in a new place – typically a place without the resources or capacity to handle such an influx. Mass gatherings, such as Hajj or the Olympics, also amplify disease risk.
 
As Dr. Kamran Khan, of St. Michael’s Hospital, Toronto explains, the global population is increasing and the “number, frequency, and scale of human congregations has increased dramatically during the past half century”. These mass gatherings increase the risk of a rapid international spread of infectious disease. In 2009, Khan’s group, Bio.Diaspora (a collaboration studying population movement and its impacts on health), and HealthMap collaborated to create an early detection system for the 2010 Vancouver Olympics. Bio.Diaspora studied worldwide air traffic patterns from February 2007 and February 2008 to learn who would be traveling to the Olympics, and to where and from where they would be traveling. The HealthMap system increased sensitivity for disease outbreaks in the 25 “cities of interest.” Relevant alerts were placed on an Olympics map so travelers and public health officials had a real-time view of potential health risks and could therefore prepare for them. In preparation for this summer’s Olympics in London, HealthMap is collaborating with the European Centre for Disease Control and Prevention and the UK’s Health Protection Agency to develop a similar system.
 
In essence, HealthMap takes information shadows of disease outbreaks and augments official public health reports with real-time information. But what makes HealthMap truly unique is that it makes this information automatically and immediately useful to those who can act upon it.
 
{"input":{"width":"75","photo":"bio1","row":"4285","table":"DOCUMENT"}}
Anna Tomasulo, MA, MPH is currently Editor-in-Chief of HealthMap’s publication, The Disease Daily. She received her Bachelor’s and Master’s degrees from Bryn Mawr College in French and Francophone Literature. Through Bryn Mawr, she studied in France and Senegal, where she discovered an interest in global health. Anna pursued studies in international health and journalism at Boston University School of Public Health, from where she received her MPH. Her studies focused on HIV/AIDS, marginalized populations, and the relationship between health and human rights. She recently returned from a fellowship with the Pulitzer Center on Crisis Reporting for whom she reported on child marriage in Nepal. Her pieces were published at the Pulitzer Center and Huffington Post.
{"input":{"width":"75","photo":"bio2","row":"4285","table":"DOCUMENT"}}
Clark Freifeld, M.S., is a Research Software Developer at Boston Children's Hospital and Harvard Medical School. His research interest is around applications of artificial intelligence and crowdsourcing to problems in global health. In 2006, Freifeld co-founded HealthMap.org, recognized by the Library of Congress and the Smithsonian Cooper-Hewitt, National Design Museum. He has published articles in PLoS Medicine, the New England Journal of Medicine, and other journals, and his work has been featured in Wired Magazine, Scientific American, CNN, NBC Nightly News, as well as a range of other local and national media outlets. He holds an MS degree from the Massachusetts Institute of Technology, a bachelor's degree from Yale University, and is currently a PhD candidate in biomedical engineering at Boston University.
{"input":{"width":"75","photo":"bio3","row":"4285","table":"DOCUMENT"}}
John Brownstein, Ph.D. is an Associate Professor at Harvard Medical School and directs the Computational Epidemiology Group at Boston Children's Hospital Informatics Program. He was trained as an epidemiologist at Yale University. His research agenda aims to have translation impact on surveillance, control and prevention of disease. He has advised the World Health Organization, Institute of Medicine, the US Department of Health and Human Services, and the White House on real-time public health surveillance. He recently was awarded the Presidential Early Career Award for Scientists and Engineers, the highest honor bestowed by the United States government to outstanding scientists and engineers. His work has been reported on widely including pieces in the New England Journal of Medicine, Science, Nature, New York Times, The Wall Street Journal, CNN, National Public Radio and the BBC.
 
About A. Tomasulo, C. Freifeld, J. Brownstein

See bios above.

The views expressed are those of the author and are not necessarily those of Science Friday.

Science Friday® is produced by the Science Friday Initiative, a 501(c)(3) nonprofit organization.

Science Friday® and SciFri® are registered service marks of Science Friday, Inc. Site design by Pentagram; engineering by Mediapolis.

 

topics