Journal article
Prediction Market for Disease Surveillance: A Case Study of Influenza Activity
Journal of prediction markets, Vol.10(1), pp.68-82
09/01/2016
DOI: 10.5750/jpm.v10i1.1162
Abstract
We conducted a pilot study on the use of prediction markets to aggregate private information for disease surveillance. Influenza activity in Iowa, North Carolina, and Nebraska between 2008-2010 was forecast through prediction markets operated on the Iowa Electronic health Markets (IEhM). We found that prediction markets were well utilized by participants, and they achieved high level of forecasting accuracy as far as 4 weeks before actual influenza-level outcomes were announced. Trading activities indicate that new information continuously flowed into the markets during the trading window, which further improved prediction accuracy as contracts drew down to expiry. This project demonstrates that a prediction market is a practical infectious disease surveillance mechanism that provides low-cost useful information for public health administration in a timely manner.
Details
- Title: Subtitle
- Prediction Market for Disease Surveillance: A Case Study of Influenza Activity
- Creators
- Anson HoPhilip Polgreen - University of IowaTim Prendergast
- Resource Type
- Journal article
- Publication Details
- Journal of prediction markets, Vol.10(1), pp.68-82
- DOI
- 10.5750/jpm.v10i1.1162
- ISSN
- 1750-6751
- eISSN
- 1750-676X
- Publisher
- The University of Buckingham Press
- Language
- English
- Date published
- 09/01/2016
- Academic Unit
- Infectious Diseases; Epidemiology; Injury Prevention Research Center; Internal Medicine
- Record Identifier
- 9984363184002771
Metrics
28 Record Views