Public health and medicine are in a moment of digital euphoria. We have convinced ourselves that mhealth (mobile phone) technologies will improve the health of millions. After all, there are 6 billion smartphones out there and more than 300,000 stand-alone health apps on the market ready to be uploaded. Suddenly, the promises of behavior change to improve the care of highly prevalent conditions (obesity, diabetes, anxiety, insomnia) are scalable. But as behavioral health facilitators, will phone apps work and how will we know if they do?
Health apps monitor, measure, and manage. (Let’s set aside for another time the apps that connect wearable devices to the electronic medical records that Apple and Google and others are creating). The first generation of apps were about self-monitoring. These “do it yourself” kinds of apps had users record and consider their own data—food ingested, mood, glucose levels—and offered users opportunities to modify their behavior. Next came measuring, quantifying apps, which soon became sensor-based, including recordings of heart rate, steps walked, sleep stages, food labels. Such sensors limited manual input and offered novel “health” markers to monitor. We have recently entered the era of the “prescribable” app. An app for the treatment of managing depression at home, at your fingertips, at any time, some even recommended by health providers.
The most commonly-trialed apps have been designed to address conditions with the largest global health burden: diabetes, mental health, and obesity. But of all the health apps on the market, only a few dozen have ever been testedusing randomized trials. And these trials, as with most new treatments, involved few participants, assessed for short periods of time, and in general were poorly designed, performed, and analyzed. This should not be a surprise; funding sources for app development and testing are rare, and a full-scale clinical trial can be long and expensive, and risks a negative result. Waiting for the demonstration of positive benefits is not yet the modus operandi of this new field.
In the digital world, faulty anecdotes of health success can go viral. We need a means for efficacy assessment and a reliable source of quality information, a Consumer Reports for health-related apps, and a recognized national organization that can evaluate and decide which are safe and which are worthy. Even a modestly potent app, if available to millions of users, could have an enormous impact on the public’s health. But the search for app evidence is time-consuming and challenging, and in the digital age, none of us has enough time.
Warmly,
Michael Stein & Sandro Galea