Digital health is hot. Our take on the Eight technologies that will transform health and care was our most popular article on our website last year. We continue to support NHS engagement through our Digital Health and Care Congress and by other means.
There is much more to digital healthcare than what is done in the NHS. In early 2013, Public Health England’s ‘Sugar Smart’ app was the leading app download chart and had been downloaded over one million times.
One of the most interesting areas for me is how and whether we can get insights from unstructured chats in forums and open chatrooms. In this rich “third space,” millions of people already discuss their health and reflect on their interactions with services. The potential is enormous to gain a deeper understanding of the experience of millions of people without being constrained by time and hierarchy constraints in doctor-patient communication.
The Fund has been involved in a
Experience with specific treatments, such as cognitive behavioral therapy
The relationship between mental health and physical well-being in three areas: respiratory diseases, diabetes, and musculoskeletal conditions.
Although the technology used to count, correlate, and identify instances varied, we were able to identify and count the number of occurrences across the sample and identify rich and meaningful experiences in each of the above areas.
It is the first time, to our knowledge, that unstructured data on a complex and nuanced area of health has been collected and categorized in this manner. This method can be used in the long term to:
- Owners of forums can tailor service offerings such as self-management to the issues and topics discussed by their users.
- Help NHS and other service providers to develop a deeper, more accurate, and truthful understanding of the users’ experiences of services and thoughtful design in response.
- Give health regulators additional insights about organizational performance, safety, and security.
These posts are not intended to answer any specific questions. It is, therefore, a very strong source, as it provides unbiased, unguarded, and full accounts. This strength can be a disadvantage. Due to the lack of focus, data interpretation was complex. It required a lot of detail. This process is also context-specific and values-driven. It is also necessary to improve the sensitivity and specificity of information categorization.
The NHS decision-makers we spoke to were aware of both the benefits and risks of this type of analysis. Forum users, for example, are selected by themselves. Certain demographic groups are unlikely to be included in online research. The ethics of such work was also a major concern. The University of Sussex gave ethical approval for this study. We followed guidelines when retrieving, storing, and handling these data. As this technique evolves, we must ensure that the research ethics and codes for traditional health research can be adapted to this new type of knowledge.
We are still at the beginning of the journey to apply machine learning to complex issues in health, and there are still many technical and ethical obstacles to overcome. This study has shown that we can identify and understand millions of unstructured and complex online conversations that are about issues that impact our health.
- Online support: Investigating the role of public forums on mental health
- Digital Health and Care Congress 2017,