What You Need to Know About Apple ResearchKit

Last week, Apple unveiled ResearchKit, an open source software framework for medical and health research, to much interest among Apple watchers, health tech and clinical research experts.

ResearchKit is Apple’s foray into clinical research based on the backbone of its HealthKit, released  in June 2014, which offered the ability for health and fitness apps to communicate with each other. mHealth News reported Apple CEO Tim Cook said, “As we worked on HealthKit, we came across an even broader impact that iPhone could make, and that is on medical research.”

With the launch came five apps developed with ResearchKit. They are as follows:

• Asthma Health-developed by the Icahn School of Medicine at Mount Sinai and LifeMap Solutions

• Share the Journey-developed by the Dana-Farber Cancer Institute, Penn Medicine, Sage Bionetworks and UCLA’s Jonsson Comprehensive Cancer Center

• MyHeart Counts-developed by Stanford Medicine

• GlucoSuccess-developed by Massachusetts General Hospital

• Parkinson mPower-developed by Sage Bionetworks and the University of Rochester

According to AppleInsider, Stanford’s cardiovascular study received 10,000 additional participants overnight, after the trial was featured on stage during an Apple's "Spring Forward" event. And according to this TED blog post, the Parkinson mPower app had recruited 8,000 participants in a week.

And therein lies the potential power of ResearchKit for trial recruitment. As reported in MedCityNews this video on Apple’s ResearchKit web site features Kathryn Schmitz, an epidemiologist at the University of Pennsylvania’s Perelman School of Medicine. She explained that a mailing of more than 60,000 letters resulted in just 305 women qualifying for clinical trials at Penn. The new technology should widen the pool of potential participants.

In addition to the recruitment factor, data collection could also benefit. Because of the features of the iPhone, ResearchKit can request access to an accelerometer, microphone, gyroscope and GPS sensors in iPhone to gain insight into a patient’s gait, motor impairment, fitness, speech and memory. In addition, the compatibility with HealthKit gives the apps access to data such as weight, blood pressure, glucose levels or asthma inhaler use, which are measured by third-party devices and apps.

According to yet another blogger, data collection in clinical trials is completely changed. Instead of collecting data from subjects every three months, researchers will collect data every three hours. Instead of collecting data from 100 subjects, they will collect data from 100,000. These benefits should improve research and outcomes, as well as decrease the costs of clinical trials.

ResearchKit Challenges

But as with any new technology, there are potential negatives with ResearchKit, first of which is just with Apple in general. Rahlyn Gossen, founder of Rebar Interactive, a company focused on patient- and clinical research professional-directed solutions on varying digital platforms, and Applied Clinical Trials Editorial Advisory Board member, said, “Apple has a long and gloried history or disproving naysayers. They also are incomparable in their attention to detail and marketing prowess. These traits will serve Apple well in creating a platform that researchers will want to use and the masses will want to adopt. Furthermore, Apple often releases flawed first generation products with the intention of iterating and improving on them. These same initially flawed products end up becoming huge successes. For example, the iPhone did not have copy and paste until the release of iOS 3. The absence of copy and paste, which is a critical feature, certainly did not prevent the iPhone from becoming a great success.”

So what are the flaws with ResearchKit? They can be divided into two buckets: exclusion and ethics.

Many blog posts note that pointed to the fact that ResearchKit is for iOS only, which significantly excludes the 81.5% of Android smartphone users. There are other issues around the iPhone vs. Android that lead to exclusion, or a lack of population diversity. While the white population is the largest percentage of both iPhone and Android users, the African-American population disproportionately prefers the Android*. Further, the median income for an iPhone user is $85K a year; and iPhone users are younger and better educated. The initial apps cover diseases that pre-dominantly affect African-Americans (diabetes, CV); children, lower-income families and minorities (asthma); with the older demographic effected by Parkinson’s.

After exclusion, the issue of ethics was discussed foremost in this blog,which noted gaps in participant verification. Gossen said, “ResearchKit does not provide researchers with adequate means to ensure the accuracy of user-supplied data. What's to prevent a minor without parental consent from entering a study? How do we know if someone truly meets the inclusion / exclusion criteria? Apple will need to overcome limitations like these for ResearchKit to be truly viable.”

The five ResearchKit apps do make clear that participant data is not shared with other entities, unless they give their consent at the beginning the study. Additionally, since ResearchKit is open source, it would allow a pharma to develop its own app for App Store download availability-essentially a portal for pharma to gather its own information from iPhone users.

Gossen concluded: “Regardless of what happens with ResearchKit, Apple's announcement is certainly a win for clinical researchers. At the very least, we'll benefit from the greater awareness about clinical research that's been generated. Given the general public's lack of awareness about clinical research, that's something we can all celebrate.”

*thanks to Lynne Nyugen, who contacted me via Twitter @disparityreport, and is Director, Population & Community Core, Center for Community-Engaged Translational Research, MD Anderson Cancer Center, who explained the statistics from the MarketProfs link on iPhone vs Android use. The stats show that the responses are largely from the white respondents, which is representative of overall white, African-American and Hispanic population of the US. Of total Android users, that number is predominantly white. But of the African-American users, they report a preference for the Android.