The Food Standards Agency (FSA) estimate that 1.7 million cases of foodborne illness are contracted each year in the UK, of which 22,000 cases result in hospitalisation and 700 result in fatality (Tam et al., 2014). The burden of foodborne illness on the population and economy remains unacceptably high and there are several problems associated with the way traditional data is used to monitor the problem. As many victims of food poisoning neither report their illness nor visit their GP, national surveillance data severely underestimate the incidence of foodborne illness. GP data publication takes approximately two weeks following an appointment. Subsequently the data is unsuitable for the timely identification of harmful pathogens (Achrekar et al., 2012).
In recent years, Twitter has emerged as a useful source of information for monitoring and mapping both disease and public health outbreaks. Available in near real time, Twitter data can be attributed to a specific location and can include a wealth of information not available in traditional data. Studies have ranged from mapping influenza outbreaks (Culotta, 2010) to monitoring dental pain (Heaivilin et al., 2011), however studies focusing on foodborne illness are limited.
This study used systematic scoping methods to retrieve and analyse all primary research concerned with monitoring public health and disease outbreaks using non-traditional data such as Twitter and restaurant reviews. Emerging from this study were methodological recommendations for using non-traditional data to calculate the incidence of a public health ailment or disease. This paper will discuss these recommendations, and consider the use of non-traditional data to more accurately quantify the incidence of disease and to further inform food establishment inspection procedures in the UK.
Acknowledgements: This research is supported by the Economic and Social Research Council [grant number ES/J500215/1].
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)