It was with interest I read a recent Viewpoint article in the Journal of American Medical Associations (JAMA) titled: Social Determinants of Health (SDoH) in the Digital Age, Determining the Source Code for Nurture authored by Dr. Freddy Abnousi, the head of healthcare research at Facebook, along with a couple of other authors, Dr. John Rumsfeld, Chief Innovation Officer at the American College of Cardiology (@DrJRums) and Dr. Harlan Krumholz, Professor of Medicine at Yale (@hmkyale)
They rightly point out the major contribution of social determinants of health – a fact highlighted as far back as to 1946 and the World Health Organization (WHO), but the research has been hampered by the inability to capture accurate granular data which is mostly self-reported (with the associated unreliability). We do need better approaches and the social networks offer a tantalizing look into data of this nature with a peek into online behavior, data that is posted by the millions of users who engage daily online.
They offer an intriguing potential to pre-identify suicidal ideation, “with enough advance warning and accuracy to stage a peer-driven intervention“. The opportunity to identify high risk for opioid addiction or finding those at highest risk of cardiovascular mortality and engaging with the users corresponding social network who would be “tasked with responsibilities”.
There is much to applaud in the concept but it raises some serious and challenging issues in my mind
1) Informed Consent is a major challenge and history and recent revelations do not engender any confidence that this data or insights would not be used against the patients or their families
2) De-Identification of data is already problematic – when you consider Intensity Analytics ability to identify individuals and behavior simply from their interaction with a keyboard
3) Trust is broken across so many areas and the current system is working as designed – a business. It is highly unlikely that users would ever *knowingly* give their consent
4) Healthcare consumers in the United States are struggling while the business of healthcare continues its march towards profit. Intuitively any insights from an SDoH program would have to focus on the best economic solutions which are mostly non-healthcare solutions (food, housing, income, education)
We need insights and data to provide the data to support and effect change and this idea has merit – but without some real changes to the business of healthcare, it will struggle to take off or deliver value to our population. I’d suggest a better incremental step would be to look at this data to show the underlying struggles of the users and creating a catalyst for change
