SENTENCES: Social media analysis to promote cancer screening

Cancer screening is a key government tool for the early detection of cancer in order to reduce cancer mortality and morbidity. Informed decision making – where people move from issue awareness and information search and evaluation to decision making – is increasingly complex in the current media environment. 

That is, the public nowadays has a large quantity and variety of information sources at its disposal (such as news media articles shared on social media, or online discussions and threads on fora) from which to select and evaluation screening-related information. Not only has this extent information led to perceptions of ‘cancer information overload’, the content of cancer and screening-related information in that media is generally also not conducive to promote trust in government and government programs, such as cancer screening. 

For instance, news media reports tends to be sensationalistic through headlines such as ‘Top doctor argues why breast cancer screening mammograms do more harm than good’. Moreover, research shows that a large proportion of tweets about breast cancer (screening) can be classified as misinformation. As a result of exposure to those news articles and social media posts, the public is unsure which cancer prevention recommendations to follow, less likely to believe that cancer is preventable, and have suboptimal levels of trust in government programs and national health services. Success of cancer screening programs can no longer be guaranteed in that media environment. 

In this ZonMw-funded project, we investigate how news media content and social media discussions affect the public’s perception of screening benefits and harms and how this is different for different social groups. To this end, we will automatically code news topics and social media discussion using Natural Language Processing techniques. Based on these content analyses and longitudinal relationships, we will develop and validate state-of-the art simulation models that will inform how the government and national health services can effectively communicate in a dynamic media environment to promote informed decision making. In this project, we collaborate with key stakeholders in the Netherlands for cancer information, cancer prevention and cancer screening, such as KWF, RIVM, and

The four-year project awards a PhD position (start 1/2/2021), a post-doc position (start early 2023), and a technical support position.

Research team:

Status: ongoing since 01 February 2021
Funding: ZonMw Prevention Program 2019-2020
Link: ZonMw website