May 11, 2010, 10:46 am
Using Twitter to Keep Up With H1N1
By Lisa Agustin
Whenever a new disease emerges, web sites for the World Health Organization (WHO) and Centers for Disease Control and Prevention (CDC) become the go-to for the latest on epidemiology and the global implications of a given threat. But “informal surveillance sources” like Internet news sites and direct reports from individuals are becoming increasingly important for identifying early outbreaks of diseases, according to a report in the latest issue of The New England Journal of Medicine. Such is the case with HealthMap (shown above), an interactive disease-tracker created as part of the Journal’s H1N1 Influenza Center. So far, the site has collected 87,000 reports (both formal and informal) to monitor the spread of the H1N1 virus. The wealth of data collected through HealthMap enabled researchers to follow the pandemic’s spread both geographically and across a given timeframe, while enabling new areas of investigation. For example, the report’s authors compared a country’s lag time between identifying suspected and confirmed cases with its 2007 national gross domestic product. (A side note: Crowdsourcing for the greater good isn’t new; the Ushahidi platform was initially developed to map both formal and informal reports of violence in Kenya after the post-election fallout at the beginning of 2008, and has since been used to monitor federal elections in Mexico, the spread of H1N1, and relief activity in post-earthquake Haiti.)
There are both pros and cons to using informal sources. In the case of emerging outbreaks, the advantages relate to the speed with which news reports are broadcast (unusual outbreaks receive intense coverage), and the ability of individual health professionals to pick up weak signals of disease transmission across borders. However, the difficulty in confirming diagnosis “presents challenges for validation, filtering, and public health interpretation.” Validating individual sources of information will become a bigger issue with the next version of HealthMap. While the current version uses individual reports from “reliable” sources (e.g., International Society for Disease Surveillance), work is underway to draw from blogs, Twitter, and Facebook. As the ability to post and share reports from the ground becomes easier, verification processes will need to be more rigorous without compromising the delivery of timely information. The maps that solve this challenge will become indispensible.
