Why we did this project
The last decade has witnessed a dramatic increase in citizens using the internet to find health advice and share health experiences. Sources of such advice range from officially accredited services, such as NHS Choices, to unmoderated, citizen-led chat rooms and forums. They vary greatly, from helpful sources of support and advice for patients and citizens to potentially dangerous, inaccurate sources of information.
The data generated by patients and citizens through their use of online health advice platforms is a potentially valuable source of untapped patient feedback, experiences, views and behaviours. However, at present there are no systematic efforts to collect and analyse these new data sets and no common understanding of how online communities can be integrated into wider health care. Research into precisely how and why many people seek or share information about health online remains fairly limited.
Despite advances in data analytics to collect and analyse big social data of this type in commercial settings, there have been only limited attempts to pilot the ability of these technologies to transform unregulated health data into something useful and useable. While their potential is recognised, more careful and robust evaluation of methods employed is required.
This project is supported by The Wellcome Trust.
What we did
In partnership with The Centre for the Analysis of Social Media (CASM) at leading cross-party think-tank Demos, we undertook a one year study, piloting the ability of big data analytics (specifically web-crawling and machine learning) to help tap into and make use of this new citizen-generated health data.
This study explored whether the data could be exploited to:
- spot trends in citizen-generated, unregulated health data online, using a pilot study drawn from mental health
- provide useable and robust insight for health professionals, clinicians, policy makers and commissioners that will in turn enable them to:
- quickly spot emerging trends
- provide early warning signs where there might be a lack of official provision and use online data to inform a response
- better understand public and patient attitudes towards certain types of treatment
- understand the opportunities provided and the risks posed to patients by these sources of information
- understand the potential uses of this information by clinicians and health services
- examine the ethical, regulatory and legal implications (most principally, the ethical considerations of collecting personal data of this kind, and how this can be done without posing risks to the individuals involved)
- examine the potential for further and wider application of these methods within mental health and beyond.
This project was completed in April 2017.