Oct 03 2014

Are Health Apps Helpful?

There seems to be an app for everything, although we are just at the beginning of this new technology. It is a very recent phenomenon that many people in industrialized nations are walking around with a hand-held networked computer. This creates a new opportunity – to have constant access to applications that can help us run our lives. Even though there are already millions of apps, we really are just beginning to explore this opportunity.

One category of apps that seem to have a great deal of potential are health-related apps. There are apps to help people count calories, track their migraines, track their exercise, or even look up medical information or take a crack at diagnosing their own symptoms. There are apps to help you quit smoking or using alcohol, manage your medications, track your diabetes, or to provide some automatic therapy for mental illness.

The UK’s NHS maintains a list of approved health apps with hundreds of vetted apps you can browse.

In there are some useful ideas, and not-so-useful ideas. Trial and error will sort that out over time, and there are probably some killer health apps waiting to be developed.

How evidence-based are typical health apps, however? How would we even study that question?

A recent survey asked UK physicians, mostly primary care doctors, their personal experience with patients using health apps. The result are in line with my own personal experience. Basically, apps that help patients track or monitor their own health or habits are helpful. This is the low-hanging fruit of health apps.

However, apps that provide information or advice to patients are problematic. The same is true of any health information targeted to the general audience, whether in books, magazines, the internet, or now apps. Less than 5% of those surveyed found such apps helpful.

The problem is, such sources of information are rarely able to put the information into an individual medical context. What ends up happening is that patients match their own symptoms to possible diagnoses, and will often match the symptoms to some horrible fatal disease. This makes patients understandably anxious, and often may motivate them to request extra testing just for reassurance.

Not all symptoms are created equal, however. Some are non-specific and may match hundreds of possible conditions. Others require individual assessment – not every twitch is a Parkinsonian tremor. It takes years of training and experience to know how significant and predictive specific symptoms are, and how likely specific diagnoses are in turn.

I am not arguing that information is a bad thing. Rather, it is critical how information is presented. Presenting information in such a way that encourages or enables naive attempts at self-diagnosis is counterproductive. Further, useful resources will anticipate common misconceptions or misinterpretations, strongly counsel for caution in interpreting symptoms, and stress the need for expert opinion.

The possible benefits of general access to health information is that people will be better informed about their own health and theoretically will be able to make better decisions about how to manage conditions that can be managed at home, and when to consult a physician.

The possible harms are that such access to information will drive anxiety, false diagnoses, and unnecessarily increase use of medical resources.

There is another aspect to this, of which I have personally seen the positive and negative. Patients who come into an office visit armed with information they have read on the internet sometimes have a leg up, already have some useful background information, and ask well-informed questions. More often, however, they are filled with misinformation, or have questions about conditions completely unrelated to what they have.

It is difficult to estimate the net effect this has on the efficiency of office visits (something very critical to the overall efficiency of the health care system). I spend a lot of time undoing damage done by misleading websites.

The bottom line is that apps and websites that provide information are a double-edged sword. We are far from an optimal use of this technology, and there doesn’t even seem to be a way to control the information that is out there. Vetting and seals of approval are helpful, but these are difficult to execute and maintain, and often are infiltrated by special interests.

A thorough review of health apps is not possible here. A 2012 estimate put the number of health apps at 40,000.  There are already hundreds of papers assessing the utility of the apps. The ones that seem to be the most useful are those that enhance patient self-monitoring and behavior. Even there, it is not clear if they improve health outcomes.


Health apps are exploding in number and diversity, in a largely haphazard free-market way. There are some benefits to this model, and some useful apps are likely to emerge. I routinely recommend apps for tracking migraines, for example.

It will take time, however, to sort out the real effects of having a marketplace with tens of thousands of health apps, and which approaches are helpful and which are not. I do think health apps deserve more attention from researchers and healthcare experts. There is likely vast untapped potential there, but also pitfalls (I guess I can’t avoid the pun of the “killer app”).

Overall I’m excited by the possibilities, especially as the technology advances with the possibility of smart phone accessories that can actually monitor health parameters such as blood pressure, oxygen levels, blood sugar, and more. Monitoring and tracking are likely to be the most useful applications. Information will always be tricky – there is no app that can make a non-expert into an expert.

11 responses so far