Viva la data revolución! Free the People’s Healthcare Data!
December 23, 2019
Healthcare data could be a pot of gold for operational efficiency and patient outcomes. Unfortunately, right now, it represents ongoing lost opportunity in my view. Population-wide data technology has enabled several improvements in analytics-driven service delivery1. Healthcare is no exception, but its real potential is far from being fully realised2, as my personal research3 has revealed.
It’s true - non-traditional market players like Google/Alphabet, Microsoft, and Apple have entered the market and liberated data for use by patients and third parties. The real gold in the health data pot, however, is the ability to understand for an individual (not just a population), what their personal health will be like in the near future.
Outside hospital settings, healthcare data is rarely collected in real time. That means conclusions are based on historical data, too late for individual interventions in the case of predictable events – and are only good for generalised predictions about large groups. Conclusions for a specific sub-population are only as good as the data slicing techniques and still don’t apply to individuals. Even at an individual level, silos mean significant predictors and risk factors for health and disease might be inaccessible at times. This isn’t just an internal problem.
Patients expect their data to be generally available and are surprised and frustrated when it is locked away (even from themselves). Young patients, in particular, are dissatisfied with healthcare accessibility and efficiency, and are more likely to choose providers with significant digital capabilities. Our Accenture 2019 Digital Health Consumer Survey indicates that providers need new, digital- and data-driven operating models, or they’ll most likely lose market share. The Accenture 2018 Consumer Survey on Digital Health confirms that patients are also positively disposed towards new technologies such as AI which, again, require liberation of patient data and pooling across organisational boundaries.
We all know it’s easy to ignore things that seem distant. For individuals, general population health statistics can easily be ignored. To really make them sit up and listen, data must be representative enough to enable population health predictions, and personal enough to make prediction and disease detection possible for their own lives. This requires leveraging AI, increasing the number of data sources and maintaining an unflinching healthcare consumer focus in the development of data-driven healthcare AI applications. New sources of data include smart watches, genomics, pathology test results, and family medical history, for example. All of these must be accessed securely and with appropriate permissions.
Predicting an individual’s short-term health requires data from all sources to be harvested across large populations, with analytical results presented to individuals alongside a call to action4.
Patient data laws typically prohibit sharing of personal healthcare data with any other organisation – often even other healthcare organisations. One solution to this is data democratisation – where patients download data and share it as they see fit. At any rate, the health data revolution cannot be stopped. The question is just which providers will participate, and which will side-line themselves by trying to ignore it?
Consumers will naturally become involved as they are empowered and exposed to the obvious benefits of limiting both immediate and long-term health risks. Once data is liberated, predictions at both individual and population levels become possible.
Healthcare providers and payers need to act fast to stay relevant. If they’re not developing their own capabilities, they need to team up with experienced market stakeholders. This means payers and providers must work on making data accessible technologically, while lobbying lawmakers to update antiquated data protection laws (often from the era before the Internet!).
Get in touch if you have any comments or questions.
1 Cooper et al., 2001; Mackenbach et al., 2015; Brownstein, Freifeld and Madoff, 2009
2 Groves et al., 2013
3 This blog is based on a chapter (by the blog’s own author) in a book titled: Digital Transformation and Public Services (Open Access): Societal Impacts in Sweden and Beyond, 1st Edition (Hardback) – Routledge, 2019
4 Flores et al., 2013