What makes digital data valuable?
By Carmen E Lefevre, on 27 October 2016
By: Dr Mark Cobain, co-founder and director, Younger Lives Group
We’ve been asked to consider the question of what makes digital data valuable. Is it visualisation? Insight generation? Personalisation, behaviour change etc.?
The short answer is all of the above but the ‘value’ of data can only be clear when you know what question you want to ask?
Visualisation to highlight awareness for a campaign
A year ago, we published a map of ‘heart ages’ (a more understandable measure of ‘cardiovascular risk’ aimed at the public) across US states using analyses in collaboration with the US Centers for Disease Control. We revealed differences in heart age across different states that were exacerbated by race and education. It was published in the (CDC) ‘Vital Signs’ journal and led to widespread coverage and usage across the US media outlets and supported by our campaign site americanheartage.com where we created a colour coded visualisation representation of the country. It was a good example of how metrics that are scientifically valid and predictive need to be impactful and visual o the public to raise awareness through all forms of media.
The data from our own digital heart age tools reveal more fine-grained insights about a variety of different public health issues that we feel are of equal importance. For example if we want to use digital tools to personalise health risks, do they actually reach those they are claimed to be for? What is the impact of missing user information on subsequent outputs? Do the gaps in information render recommendations irrelevant?
Having access to millions of first time users of the heart age tool in multiple countries we could analyse our data to help shed light on these questions and published the results in the Journal of Medical Internet Research two years ago. It was a smaller impact publication and not as widely covered as our CDC research. However, these are vitally important questions for digital health to answer. It’s vitally important to share more information about the impact of your program/product/service on users. For example, we discovered that users of the heart age tool were not the ‘worried well’ and had adverse risk profiles (10% smoked, the average BMI was overweight and blood pressure and cholesterol profiles were in line with population averages). However, (in some user sub-groups) missing information in the tool led to a degree of CVD risk misclassification and more often than not this was due to missing biometric data.
From insight generation to behaviour change
Insights such as this need not be a threat. We could align our business with the desire to solve these problems. The last thing we want is ‘false optimism’ (we want to stimulate lifestyle change, not falsely reassure those who need to). It’s just the opposite of what our tool is intended to do and shown to do in clinical trials. We’ve since made changes to tailor messages for people with missing information to get tested for an accurate heart age and partnered with organisations who can provide that service to them. Not all users however need full cholesterol tests and so it’s about matching the need to the right messages (and tracking the consequences over time)
What are the consequences for those working in digital health?
Providers of digital tools with a high user base have a duty to address issues of public health importance or risk being seen as damaging to population health. The good news is that their data can be used to address highly important questions AND to help their business grow by adopting new solutions. Instead of a threat, they can be an opportunity to drive collaboration between public health and digital health and an education for both sets of stakeholders.
Our business (Younger Lives Group) believes there is great value in developing health apps and sites founded on validated science (our background combines epidemiology, psychology and social marketing as well as digital), however it’s always a balance between the academic science, the user journey and analytics, in trying to avoid unintended consequences. Tracking use data is a requirement for that. We’ve learned a great deal on how to get it right as well as what to do when you get it wrong and if you want to learn more about how please visit us at youngerlivesgroup.com.
BIO: Dr Cobain has spent over 20 years working at the intersection of health psychology, epidemiology and digital health applications. He has worked across the foods and pharma industry as well as having been a research fellow in the Framingham Heart Study and is an honorary lecturer in preventive cardiology at Imperial College London. He also has a track record of publishing at this intersection in peer reviewed journals and acts as part of the editorial team for journals such as the American journal of health promotion. In 2008 he published the concept of ‘Heart Age’, which has now been adopted by the NHS and has collaborated with the CDC on its use in the United States to motivate lifestyle changes. He is also a director and co-founder of a health behaviour change company that specialises in the area of ‘healthy ageing’.