ResearchKit: A framework for developing medical research apps

At Apple's Spring Forward event two weeks ago, the star of the show was Apple Watch which will be available on April 24. However perhaps the most intriguing (and unexpected) announcement was for a new developer platform called ResearchKit.

ResearchKit is an open-source software framework that provides tools for researchers to create apps that can be used to gather data for medical studies. The goal is to help researchers conduct medical studies using data collected by the iPhone (and peripherals) with the consent of the user. In typical medical studies, it is challenging to enroll enough study subjects and to collect data from the subjects in a reliable fashion. The sensors, computing power, mobility, and communications ability of a smartphone make it an ideal tool for facilitating medical data collection.

One example highlighted during the event was an app created by Stanford doctors that assesses a person's cardiovascular risk. The app collects data on movement from the smartphone accelerometer like a standard activity tracker, and presents a series of surveys to be filled out by the user providing lifestyle information. From this input, the app calculates your heart age and your 10-year cardiovascular risk. Presumably the study will monitor cardiovascular health of the subjects that can be correlated with activity and lifestyle choices inferred from data collected by the app.

Within a few days of launch there were at least five of these medical research apps that could be downloaded from iTunes, and thousands of people had signed up (Apple Insider):

"Stanford's study, a joint effort with the University of Oxford, is one of five live in the App Store. Others include an asthma self-management program from Mount Sinai, Weill Cornell Medical College, and LifeMap; a Parkinson's study from the University of Rochester and Sage Bionetworks; a diabetes analysis tool from Massachusetts General Hospital; and a breast cancer study from the Dana-Farber Cancer Institute, UCLA Fielding School of Public Health, Penn Medicine, and Sage Bionetworks."

The smartphone (iPhone) furnishes the primary sensors for this data collection:"[P]eople take tests like saying “ahhh” to detect vocal variations, walking in a line, or tapping in rhythm to test for Parkinson’s Disease." However eventually, these capabilities can be greatly extended by using  iPhone peripherals. For example, the Apple Watch can monitor your heart beat and could eventually perform blood glucose measurements; ear pods can monitor respiratory rate and blood pressure, and even the iPhone case can measure your heart's EKG. The possibilities are literally endless.

ResearchKit is similar to Apple's previously announced HealthKit framework. To remind you, Healthkit is a developer platform that expose an API that allows "different apps and devices to talk to each other." As a result, your internet-enabled scale by one manufacturer can communicate your weight information to the fitness app developed by another company. The goal is to produce an ecosystem of devices and apps that can exchange information while taking advantage of the various resources on the platform. ResearchKit probably shares many of the same API tools as HealthKit i.e. they probably have overlapping functionality.

The primary difference between the two is that HealthKit is directed toward developers writing apps for the general consumer, whereas ResearchKit is directed toward medical researchers who want to make apps to help collect data from subjects for medical studies.

However, as Apple points out: "With a user’s consent, ResearchKit can seamlessly tap into the pool of useful data generated by HealthKit — like daily step counts, calorie use, and heart rates — making it accessible to medical researchers." Thus, a ResearchKit app can collect data from a HealthKit app, but because of the patient privacy concerns it probably doesn't work the other way around.

Google and Microsoft have also released their versions of HealthKit (Google Fit and Microsoft Health, respectively) and so in the future they may set up something similar to ResearchKit for Android and Windows Phones. Regardless this new initiative is a great way for recruiting more people into medical studies and collect more data with the goal of remedying "small sample sizes and inconsistent data collection" with a process "that is more automated and less subjective." Already more than 11,000 people have signed up to use the Stanford cardiovascular disease app (Figure 1).

Figure 1. The Stanford Medicine MyHeart Counts app which will use surveys and tasks to collect data on cardiovascular disease risk.