Do health apps need endorsement? Challenges for giving advice about which health apps are safe and effective to use

Two decades ago, during the rapid emergence of the worldwide web as a popular source of health information, there was widespread anxiety among health practitioners, and especially the medical profession, that poor quality information on the internet could cause significant harm. Although there was never any consistent systematic evidence of harm beyond a few individual case reports, one response by many national health agencies was to call for, and in some cases develop, quality rating schemes. In 2000, an editorial in The BMJ by Tony Delamothe described these initiatives as ‘kitemarking the west wind’, referring to the difficulties in rating the quality of online health information and providing kitemarks for websites, and arguing that such schemes needed to demonstrate their benefit to population health before being pursued further. Just as the first generation of web technologies provided users with access to a vast quantity of online health information, now web2.0 interactivity and the rapid uptake of smartphones have led to an explosion in digital applications for health and wellbeing. An estimated 165,000 health apps were available in 2015 to the public, patients, health professionals and researchers. Health apps cover both ‘wellness’ apps, mostly for promoting healthy living (e.g. targeting physical activity, diet, sleep, stress), and ‘medical’ apps for diagnosing, self-managing, monitoring and treating conditions, providing decision-support and collecting healthrelated information. Regulation does exist for some health apps that are considered to be medical devices. In the United States, the Food and Drug Administration (FDA) launched a draft guidance on mobile medical apps in 2011, which was updated in 2015 and remains a work in progress. This guidance applies to health apps that meet the regulatory definition of a device, meaning those apps that are ‘intended to be used as an accessory to a regulated medical device, or transform a mobile platform into a regulated medical device’. In the United Kingdom (UK), the Medicines and Health Regulatory Agency (MHRA) published guidance in 2014 on health apps falling under their definition of medical devices and which should apply for CE marking (logo used in the European Union to confirm that a medical device is fit for purpose and adheres to safety legislation). While the exact number of health apps that have been registered as medical devices and FDA-cleared or received a CE mark is hard to determine, it is clear that so far only a very small proportion of health apps are regulated in this way (in 2013 there were approximately 100 FDAcleared apps in total, and in 2016 only 24 devices connected to health apps and 12 medical devices were cleared). However, most health apps are not medical devices by these definitions, and are therefore currently unregulated by organisations such as the FDA and MHRA, and freely available to the public, regardless of their efficacy and safety. Just as with the concerns over the quality of health websites, health policy makers in different countries around the world are now considering whether wider endorsement of health apps is required, as a result of concerns about risks to patient safety and privacy related to existing apps. 21 A recent review by the Commonwealth Fund showed that of 376 iOS apps and 569 Android apps for engaging patients with medical conditions, only 43% of iOS apps and 27% of Android apps appeared likely to be useful. Systematic assessments showed concerns about the clinical safety of several health apps. Concerns have also been expressed about selling data from health and wellness apps for commercial purposes. The European Commission, and national and regional governments, e.g. in the UK,, Catalonia and Andalusia, are therefore currently developing guidelines for assessing the quality of health apps that