The Mobile Health Paradox: Why Data Isn’t Nearly Enough

Health and fitness apps are all the rage at the moment, but do they actually help us live healthier lives?

Note: This article was originally published in TechCrunch on Feb 10, 2016.

Across most developed economies, healthcare costs are rising faster than inflation. In the U.K., the National Health Service (NHS) faces an estimated funding gap of £30 billion by 2020. In the U.S., the situation looks more bleak, with total annual healthcare spending surpassing $3.8 trillion, representing an astonishing 17.4 percent of the country’s total GDP.

A key cause of the rise in healthcare spending lies in the spiraling costs of treating preventable chronic diseases (such as obesity, heart disease, stroke and cancer), which account for 88 percent of total healthcare spending. This figure isn’t surprising when you consider that approximately half of all adults in the U.S. have one or more chronic conditions. More worryingly, seven of the top 10 causes of deaths occur as a result of preventable chronic diseases, with cigarette smoking alone accounting for 480,000 deaths in the U.S. every year.

These facts suggest that many of the key healthcare challenges of the twenty-first century lie in how we tackle chronic disease. This article will explore the role of mobile technologies in meeting these challenges, why they have failed to do so until now and what a solution might look like in the future.

The paradox of digital health

With almost five billion mobile phone users in the world, of which two billion are smartphones, digital health (also known as mHealth or connected health) has been lauded as an attractive solution to address the challenges of the rising costs of chronic morbidities.

Moreover, the ubiquity of smartphones has led to a burgeoning market for m-health apps and wearable devices, resulting in more health data being collected than ever before. This has given rise to a phenomenon known as “the quantified self,” the process of tracking everyday activities to learn more about yourself.

For example, it is now possible for individuals to know their average time spent in REM sleep over three months and whether their sleep quality correlates with bad weather. One can now check their blood pressure, oxygen saturations and ECG in a single device, receive a full genetic analysis for less than $100 and soon be able to keep track of real-time glucose levels thanks to Bluetooth enabled contact lenses.

A simple look at Apple’s Health app yields no less than 79 different health records, spanning Vitamins A through E, variations in body temperature and caffeine levels. A new mobile app takes the obsession to quantify into the bedroom, by helping individuals track their sexual encounters on their smartphones. The app collects information on the sexual duration and noise levels in order to quantitatively assess the user’s performance, presenting the data in a series of attractive graphs.

Proponents of “the quantified self” phenomenon argue that the data we collect is for self-knowledge and self-empowerment. Whilst there is certainly nothing inherently wrong with the notion, insights from behavioral economics teach us that simply knowing more won’t necessarily help us live healthier lives.

Numerous experiments have shown that we human beings are not “rational individuals who engage in maximising behaviour,” but instead place little emphasis on long-term rewards, tend to prefer avoiding losses to acquiring gains and, in general, are motivated by cognitive biases of which we are largely unaware.

Failing to appreciate this, most health and wellness digital health apps have historically suffered from very short life spans, with engagement disintegrating soon after the novelty wears off — users are then free to return to risky health behaviours, negating any intended long-term health benefits.

It’s all about health behaviour

As a senior medical student at Imperial College London, I find myself exposed to a range of medical and surgical specialties during my clinical rotations — each of them presenting different problems, but managed in the same three stages: conservative, medical and surgical management.

Despite the rapid pace of medical innovations over the past century, the first line treatment for almost any insidious chronic disease involves taking a conservative approach comprising “lifestyle and risk-factor modification.” This typically involves the doctor reciting their well-rehearsed spiel; for example, instructing arthritic patients on the mobility exercises which, if completed regularly, could extend the life of their joints by 5–10 years, or young diabetic patients on the importance of good glucose control.

Yet research shows that patient recall of the medical information delivered in doctor consultations tends to be poor and inaccurate, with most patients focusing on diagnosis-related information and therefore failing to register advice, ultimately affecting their ability to later act on it.

Mobile devices offer a promising solution as a conduit for behavioral intervention programs. By combining the “big data” generated by health apps and devices with digital behavioral interventions, there is potential to pave the way for a new generation of m-health apps.

The success of such an approach requires a radical transformation — from simply collecting, storing and relaying information back to users to intelligently processing the data collected to help recommend highly personalized, evidence-based techniques, designed to nudge behaviors in the right direction.

For example, research shows that not all smokers confer the same benefits from smoking, nor have the same motivations for wanting to quit.

A simple illustration would be that person A might smoke to help manage their stress and want to quit as a result of social pressures, whilst person B might enjoy the social aspect of smoking but may be looking to quit due to health concerns.

The support required to address their reasons for smoking and motivations for quitting will vary drastically from person A to person B. To tackle this, we are developing our flagship product, Quit Genius, a behavioral intervention for smokers looking to quit, capable of intelligently identifying and adapting to all aspects of the user’s thoughts and feelings about smoking.

There are also cross-specialty implications for digital behavioral programs. A new coin-sized device developed at Imperial College now makes it possible for patients to accurately monitor their breathing and cardiac activity overnight at home to detect sleep apnea, a devastating sleep disorder estimated to cost the U.S. economy as much as $165 billion annually.

A promising application of this data could involve a targeted digital behavioral program specific to the individual, matching their weight loss to measurable improvement in their quality of sleep. Linking these lifestyle modifications to the individual’s chronic condition and consequently providing a personalized and highly structured digital behavioral intervention could have a transformational effect on the individual’s life, improving their quality of sleep and, as a result, their ability to function.

Maintaining the evidence threshold

There are currently some 165,000 health-related apps available on the market. Although that’s a highly impressive number, research has revealed many mental health apps, as well as existing digital behavior interventions, are plagued by a lack of underlying evidence and scientific credibility and, most worryingly, confer limited clinical effectiveness.

Theranos’ recent misadventures clearly illustrate what can go wrong when healthtech startups believe their own hype and refuse to engage the scientific community. As a result, there is an ongoing debate within academic circles regarding the level of clinical validation required for m-health apps. To tackle this, the Department of Health in the U.K. is looking to create a new endorsement model for evidence-based health apps.

Embracing an evidence-first approach by building up a solid evidence-base through published clinical trials will confer additional benefits. It will give healthcare practitioners much needed reassurance when recommending m-health apps to their patients, laying down the foundations for clinicians to “prescribe” digital behavioral interventions to help tackle specific “lifestyle and risk-factor changes” in the future (similar to how a clinician might currently prescribe drugs).

However, for this to occur, more work is required to help establish a framework for conducting rigorous clinical trials. A fine balance must be struck to ensure these apps can be clinically validated within a reasonable timeframe, whilst maintaining validity in the scientific approach and the evidence-base that is generated.

Towards a better future

It is widely accepted that individuals can drastically reduce their risk of premature morbidity and mortality by avoiding risky health behavior, of which lack of exercise, poor nutrition, tobacco use and drinking too much alcohol are the four main culprits.

To truly make a dent on a global scale, digital health startups must recognize the importance of creating highly structured and personalized evidence-based behavioral therapies to tackle specific risky health behaviors. To achieve this, one approach is to create a close cross-collaboration between clinicians, behavioral psychologists, designers and developers to conceptualize, build and clinically test behavioral interventions programs.

The future of healthcare lies in how we tackle chronic disease and avoid health-risky behaviors. The opportunity is clear; the part digital health interventions will play in tackling this challenge has yet to be decided.

No Spam. No Fluff. No Noise.