Guest post by Kalisha Narine, technical architect, Medullan.
In March 2015, Apple announced the next big thing for the scientific community: ResearchKit. According to Apple, the new application would help researchers gather more data, more frequently, and more accurately than ever before, all by utilizing the more than 94 million iPhones in use in the U.S. today as a strategized recruitment channel.
In a nutshell, ResearchKit makes it easier for researchers to create iOS apps for their own research, focusing on three key things: consent, surveys, and active tasks. ResearchKit provides communication and instruction for the study, in addition to pre-built templates for surveys that can be used to collect Patient Reported Outcomes. Plus, ResearchKit can collect sensor data (objective patient activated outcomes) on fitness, voice, steps, and more, all working seamlessly within Apple’s HealthKit API, too, which many users have on their devices already. This allows researchers to access relevant health and fitness data (passive patient outcomes).
ResearchKit-powered apps like MyHeart Counts, Share the Journey, Asthma Health, GlucoSuccess and mPower have shown us that people want to do their part in advancing medical research by sharing their data with researchers committed to making life-changing discoveries that benefit us all.
Five months after its launch, I’d say, in no exaggerated terms, that ResearchKit has proven to be game-changing for researchers, leapfrogging patient reported outcome studies into a “mobile first” world. However, the current framework certainly doesn’t cover the full gamut of what is needed to build a patient-centered, engaging, scaleable digital outcomes solution. If you’re planning piloting a solution around ResearchKit, here’s what you need to know:
ResearchKit offers up important benefits for medical researchers, especially when it comes to recruitment capability and the speed at which researchers can acquire insightful data to speed medical progress.
The MyHeart Counts app has been arguably the most successful example of ResearchKit use to date — it’s a great example of the recruitment capabilities provided by ResearchKit. In just 24 hours, the researchers from MyHeart Counts were able to enroll more than 10,000 patients in the study. Then they clocked an unprecedented 41,000 consented participants in less than six months (even before entering UK and Hong Kong markets). As most researchers know, recruitment can be one of the biggest challenges in building a study. But with ResearchKit, scientists are able to grow their number of participants into the thousands very quickly; it would have taken the MyHeart Counts researchers a year and 50 medical centers around the country to get to 10,000 participants.
Additionally, ResearchKit also increases the speed at which researchers are able to find the insights they’re looking for. This is mostly because people use their mobile devices constantly (most Americans clock more than two hours per day), which means that the accumulation of mass amounts of subjective (surveys), objective (sensors/active tasks) and passive (background) data happens quickly. The Asthma Health app is a great example of this, as it combines data from a phone’s GPS with information about a city’s air quality and a patient’s outcomes data, all to help patients adhere to their treatment plans and avoid asthma triggers — study participants told researchers that the app was also helping them better understand and manage their condition. The app is also assisting providers in making personalized asthma-care recommendations.
These benefits cannot be undermined — they are incredibly important and useful for our industry. But as with all new tools, ResearchKit has some stop gaps that need to be accounted for, too.
ResearchKit does not take care of data analysis, data storage, or data security.
If a study is capturing objective outcome data from patients via sensors, the app could be collecting more than 18 million data points from a single participant every day. While ResearchKit allows researchers to collect all this data easily, it doesn’t store, secure, de-identify or analyze that data in a HIPAA-compliant manner. Researchers will find themselves needing to separate the useful data from the noise if they want to efficiently gather useful insights.
Patient engagement is not a feature of ResearchKit.
So while we can tell you that ResearchKit’s initial offering provides the basic tools needed to build a successful study, the default templates and functions alone will not drive patient engagement. To get there, the experience needs to add basic features such as engaging scheduled surveys, interactive notifications and reminders for patients, personalized educational content, and visual progress tracking. The good news is that researchers and digital solutions partners like us are doing just that (and more) right now, building customized experiences on the ResearchKit developer platform.
Bottom line: Clinical research is undergoing major transformation thanks to digital advances like ResearchKit. Buckle up for the ride.
As researchers, we are certainly living at an interesting time — digital technologies are reinventing the way that patients take part in clinical trials and driving research farther, faster. So while ResearchKit alone does not cover the full spectrum of what is needed for a patient-centered, easy, real-time, secure and scalable solution, it does encourage and enable experimentation from both patients and researchers at a scale and speed we are just beginning to understand.