Grants

Fund for Public Health in New York, Inc.

To evaluate and validate the use of social media for foodborne outbreak detection

  • Amount $1,044,516
  • City New York, NY
  • Investigator Romy Basil
  • Year 2015
  • Program Initiatives
  • Sub-program New York City Initiatives

The New York City Department of Health and Mental Hygiene (DOHMH) estimates that more than 1,000 restaurant-associated outbreaks of foodborne illness occur in the city each year. Outbreaks are usually reported by the victims themselves via telephone calls to 311 or the health department. Most victims don’t bother, however, and as a result the DOHMH detects only about 30 outbreaks each year. Since quickly detecting foodborne illness outbreaks is critical to implementing control measures in time to protect the public, better detection measures are needed. This grant funds a project by the Fund for the City of New York, in collaboration with the DOHMH and researchers at Columbia University to experiment with using Twitter and other social media to detect unreported instances of restaurant-related foodborne illness. The theory is that while people may be unlikely to report a foodborne illness to the health department, they are much more likely to tweet or post to Facebook about it. Real-time analysis of public data from Twitter and other social media sites may be able to reliably inform health department officials of outbreaks as they are happening. Over the next three years, the FCNY team will develop algorithmic methods for searching Twitter feeds, identifying tweets potentially relevant to foodborne illness outbreaks in NYC, and then evaluate the reliability of those algorithms in detecting actual outbreaks. Additional grant funds support efforts to increase voluntary reports of foodborne illness outbreaks by allowing NYC residents to report illness directly through Twitter. The project is experimental, but the prospective gains are large. Even a small increase in the ability to detect restaurant-related foodborne illness outbreaks would represent a significant improvement of current detection capabilities.

Back to grants database