From health “big data” to consumer’s “small data.”
Powerful analytical engines are ready to mine terabytes of patients’ data collected from watches, bands or implanted sensors. Major IT vendors are offering providers their platforms such as Microsoft Health, Apple Health Kit or IBM Watson for a better insight into patients’ health conditions. Powerful pharmaceutical industries are looking at Apple ResearchKit to collect anonymous health data from a vast majority of iPhone users. Similar initiatives are driven through social networks such as PatientsLikeMe, encouraging members to share their health data with research and pharmaceutical industries.
Expectations are that the pool of Health-related BIG DATA, when crunched and analyzed, will facilitate a more precise diagnostics, better medication treatments and finally will improve healthcare processes. Consequently, in due time each individual will receive its share of benefits. But no expert is able to promise when those advantages would reach personally you.
Meanwhile millions of individuals are monitoring daily their vital signs such as blood glucose level, heart rate and medications with only one selfish hope: improve their own health and as soon as possible. Alas, even those observations one regularly collects still do not suggest how to make personal health results look better. Should I, e.g., increase my beta-blocker from 0.5 to 10.0 mg, switch to another drug or simply move to the country where the climate is better for me?
To interpret health data correctly people need to understand the dependencies between their personal health parameters (e.g., blood pressure) and external factors that can affect them, e.g., medications, stress, weather, physical activities, etc. In short: for consumer health applications to be demanded, health analytics for person’s “small data” is strongly required.
Unfortunately the majority of developers are still reluctant to add analytical engines to personal health tracking tools. Numerous mobile apps appearing on the market targeting diabetes, hypertension or cardio disorders are still presenting user silos of health-related data.
Let’s take hypertension. Over a billion people are measuring their blood pressure daily in order to avoid the possible consequences of this condition. As a user I was keen to test some of the top blood pressure application available for Android, Apple and Windows. The criteria were: rating, mobility, ease of use. Since the first time users are normally not ready to pay for something they have not checked before, mainly the for free applications were randomly considered.
Although different in colors and layouts, the reviewed apps were mainly offering the same familiar functionalities with similar data sets such as systolic, diastolic, pulse rate, weight and BMI. Additional options ( meant more for physicians rather than for consumers) were left/right hand measurements, position (seated, laying). Some apps such as iBP Blood pressure additionally contain the glucose level. The promised correlations offered by charts simply show different data (the blood pressure measurements, weight and glucose levels) presented on a single page.
The Blood Pressure Monitor – Family Lite by Taconic System LLC is seemingly more advanced promising consumers “lifetime data visualization, statistics reporting, medication correlation, email import/export, built-in reminders and much more”. Nevertheless, despite its multiple configurable pages including medications inputs and additional services such as “built-in reminders”, it basically shows static pages and not a dynamic analysis.
Among the “rising stars in blood pressure apps” one can find fingerprint blood pressure pranks that scan your blood pressure when you are pressing your finger to the designated area on the phone.
Numerous apps of this category such as the one designed by Prixlogic or Finger Blood Pressure by Onex Softech are more like an idea than a real BP app. One can only guess how these apps are supposed to operate and what exactly they measure. No surprise when you see the difference in results in systolic something about 20-30 mm Hg or even more compared to the numbers obtained from standard medical BP devices.
The Instant Blood Pressure app allegedly designed by respectable experts from John Hopkins “promises to measure your blood pressure using just your iPhone” has also received mixed feedbacks. The app claims to use the iPhone sensors to measure blood pressure without using a blood pressure cuff. It asks you to place your finger on the camera (measuring your heart rate), and place your phone’s microphone on your heart. The measurements were not really tested and were giving unreliable results. For example, a person had 170 systolic, while the reading indicated a normal BP. Finally, the University had to send to “the app’s maker, Aura Labs, a cease-and-desist letter demanding that they stop using their name”.
Although the simplicity of the usage is tempting, so far mobile BP measurement apps are in no way a substitution to the standard medical devices. The accuracy of measurements is, as a rule, totally inadequate.
To design a meaningful healthcare application for consumers, developers have to step into their shoes. To work with medical professionals is absolutely necessary, but not enough. One has to “feel with his own skin” what the priorities are for people, and what is excessive, unnecessary and can wait. Patients simple do not have neither time nor capacity to input and track abundance of parameters. However, there are ones such as the medication intake, which are absolutely key. By simply including a medication field and presenting the user with a visual chart that shows the dynamical dependencies between the drug dose and physiological parameters an individual is tracking, will help a person to understand her/his conditions better and tune the treatments. Consequently, the utilization of such applications will increase. The parameters can of course vary depending on physical disorders. For example, people with cardio -vascular problems are often meteo-dependent. The weather for them can become an important factor. So why not to integrate blood pressure apps with the numerous existing weather forecast applications?
I hope that developers will finally pay attention to individuals’ “small data” by adding extensive analytics as a standard function to personal health applications. Until that happens I will continue using my modest blood pressure tracker on my Windows Phone, guessing helplessly what is more important for my results: the drug I have swallowed or the sun shining through the window?