Depending on where you live, you look at your mobile phone on average between 40 (UK) and 120 (US) times a day. The likelihood of one of those interactions involving a health site is very low, despite the large number of applications available. In 2012 there were over 40 000 mobile device applications for health care, now the number has quadrupled.

They are all flat screen attributes, designed to catch the roving eye and hold it at least to purchase or single use.

The most accessed health apps deal with weight reduction. Even there, the overall use is low and rarely sustained. In one very large study (189,770 people) of a free mobile app for dietary self-monitoring only 2.58% used it actively and most people who started did not continue. Those who did were healthy eaters or already compliant with strict diets.

So why are more and more apps, WHO calls them m-health, arriving on the scene when there is so little scientific evidence supporting their use? Perhaps it is the way we measure effectiveness of apps. Health apps reviewers have established a set of 5 criteria by which they judge them. The MARS system looks at five items: engagement, functionality, aesthetics, information, and subjective quality.

They are all flat screen attributes designed to catch the roving eye and hold it at least to purchase or single use. These five are not the kind of characteristics that you would judge the clinical acumen of your regular physician by, especially not their appearance. So apps that replace regular clinician contact are unlikely to be successful.

On the other hand, for organisations interested in global health, the widespread reach of mobile communication makes some apps attractive. In a cash strapped world where funding for global initiatives is a low priority, m health is an attractive and more affordable option to propagate health promotion, disease prevention and monitoring messages.

When devices attain an International audience, regulators such as the FDA, take an interest. The FDA has recently classified mobile medical apps according to their risks to patients: high, medium and low. High risk applications are ones that can directly affect patient care such as urine analysers and blood sugar testers. Where clinicians are sparse and treatment options may only be available on line, such as assistance to determine drug dosages, the risk is assessed as moderate. Many applications are low risk, designed for mass consumption such as thermometers and blood pressure monitors.

These regulations do not address the problems of communication that are intrinsic in rapid and short interactions such as occur in the mobile environment. Health communication over time is complex and there are also privacy concerns. More dialogue is required to define a role for m health to improve patient care, before our miniscreens fill up with apps we use once and never again.