The Promise (and Limitations) of Apple’s ResearchKit
Fifteen minutes into the glittering product parade presented during Apple’s “Spring Forward” event on 8 March 2015, the presentation focused on ResearchKit, a non-glittery, pro bono initiative that seemed strangely out of place yet provided a real, and exciting, example of what mobile smart technology can do beyond allowing you to collect green chiclets in the Candy Kingdom.
Until now, much more has been heard from Google and Microsoft than from Apple in supporting academic research, so it’s nice to see Apple stepping up to the plate to help make the world a better place in ways other than just producing compelling products.
The Promise — As Apple glowingly describes it, ResearchKit is “an open source software framework that makes it easy for researchers and developers to create apps that could revolutionize medical studies, potentially transforming medicine forever.” Yes, the language is somewhat overblown, but, in fact, may turn out to be literally true.
Consider: today’s iPhone, which, in addition to lots of computing power, communication capabilities, and data storage, contains sophisticated sensors like an accelerometer and a gyroscope, along with a special processor to monitor and manage the data they provide.
Also consider: iOS includes HealthKit, which enables iPhone users to record physical activity through both automatic monitoring of the device’s built-in sensors and user input, and does so behind a sophisticated security system that protects the user’s privacy (see Rich Mogull’s Macworld article, “Why Apple really cares about your privacy”).
Next consider that much medical research involves studies that use human participants and that track certain facets of participants’ physical condition and behavior over time. In most cases, the more participants, the more useful and reliable the research findings — but also, the more expensive and difficult the research is to conduct.
Finally, consider that hundreds of millions of people already own devices that contain sophisticated sensing, security, and communication capabilities, devices that could be used to collect data from study participants, and at a scale and level of detail that heretofore has been, if not impossible, certainly expensive to deploy and difficult to manage.
If researchers could easily and reliably plug into the rich motion-sensing and health-related resources the iPhone offers in a way that protects participants’ privacy (de rigueur in any medical study), it could revolutionize medical studies and transform medicine.
That’s what ResearchKit promises.
In the long run.
What We Have Now — For the moment, though, ResearchKit exists as a relatively small software framework for building iOS apps that collect and manage medical research data, along with a few apps that researchers have developed with the framework.
The apps serve as demonstrations of what ResearchKit can do, and you can download them yourself if you like. But they also are actual research tools, and have already begun to yield substantial benefits to researchers: within 24 hours of the Apple event, 11,000 people signed up for one study, something that would normally take a year and cooperation among 50 medical centers.
The ResearchKit framework itself, due to be released as open source in April 2015, is relatively small and unambitious. Currently, it consists of three modules:
- Surveys: Many studies involve surveys to which participants respond. ResearchKit provides a localized user interface with which researchers can specify their questions.
- Informed Consent: All medical studies require the informed consent of participants whenever personal health information is collected; ResearchKit offers a number of templates that researchers can customize to provide the information required in order to request such consent, as well as methods for acquiring consenting participants’ signatures. Researchers can also create comprehension tests with this module that help make sure participants understand just what they are consenting to.
Active Tasks: This module currently provides five subsidiary modules that researchers can use to collect data from the iPhone’s sensors. These include Motor Activities, such as gait and coordination, making use of the accelerometer, gyroscope, and touchscreen; Fitness, which uses the accelerometer; Cognition, which employs the touchscreen; and Voice, which uses the microphone. Support for data returned by the sensors in the Apple Watch (which can, among other things, detect heart rate) is also provided by some of the Active Tasks modules.
However, ResearchKit itself does not provide for secure communications between the deployed app and the researchers’ server, it doesn’t give access to passively acquired background data, nor does it provide the capability to schedule tasks or surveys; iOS has other programming frameworks that can handle those needs.
Also lacking is any automatic compliance with federal or international research regulations and guidelines: researchers are responsible for implementing such compliance.
What Doctors Think — Many of those in medicine are thrilled by the possibilities that ResearchKit offers. For example, Dr. Norman Cohen, a professor at Oregon Health Sciences University, told me, “ResearchKit has the potential to be a real game-changer. As an anesthesiologist, I am always looking for ways to help patients have improved surgical outcomes. ResearchKit could provide a method to measure the impact of interventions, such as diet modification, physical pre-conditioning, glucose control, and medication management on outcomes across large numbers of surgical patients. The
framework may also be able to improve patient education, engagement, and compliance with care pathways. These are exciting possibilities as we institute Perioperative Surgical Homes in an effort to improve the value of care we deliver.”
Others are less sanguine. Dr. Phil Jones, Associate Professor at the University of Western Ontario, has his doubts, saying, “You cannot simply ‘crowdsource’ medical research. There must be some minimal checks and balances, and those checks and balances are precisely what ResearchKit is currently missing.” He provides several examples of where today’s ResearchKit falls short, including design holes that could allow selection bias, attrition bias, and observer bias to corrupt a study’s findings.
Nonetheless, even Dr. Jones isn’t a complete naysayer, concluding his analysis of ResearchKit’s flaws on an upbeat note: “However, the future is bright! I have no doubt significant improvements will be made that will enable high quality, high fidelity research to be done using smartphones and ResearchKit.”
If researchers, and Apple, actively devote the time, resources, and ingenuity needed for ResearchKit to fulfill its promise, the future may indeed be bright: the results of that study are yet to be compiled.
Looks like all the apps are only available in the US. mPower, MyHeart Counts, etc do not show in the Australian iTunes Store. :-(
That is unfortunate, but not surprising. Each country has its own laws that govern medical research conducted within its jurisdiction. The apps most likely have not been checked for conformance with non-US medical research guidelines and statutes.