Yesterday, during Apple’s “Spring Forward” event, a small announcement was somewhat unceremoniously dropped into the presentation between a thinner MacBook and details about the Apple watch. Apple announced their entry into a space that has the potential to be a game changer for health care: a toolkit for developing medical research apps. ResearchKit is Apple’s attempt to empower medical researchers with a platform to develop standardized apps for iOS. This platform will allow for the collection of exponentially more health data than most researchers can currently collect today.
The majority of clinical trials and longitudinal medical research studies are conducted the same way they have been run for decades. Participants meet at scheduled intervals with researchers to receive treatments or medical tests to monitor their health and progress. The large gap between the points of data collection is simply par for the course. Many trials use paper-based self-reported tools to track and monitor changes in symptoms or overall health between visits. These tools, while helpful in increasing the rate of collection, are often intertwined with perception issues, consistency difficulties, adherence problems and human nature-driven biases.
So what’s changed? Millions of people now carry an Apple-built phone in their pockets (Apple has sold over 700MM of them). These iPhones are jam-packed with sensors and nearly constantly connected to the internet. When combined with intelligently designed research protocols, they make the possibility for real-time or near real-time data collection and reporting a reality. Bridging the gaps between the trial participant meetings with researchers, these new apps will start to standardize the mechanisms for real-time research-driven medical data collection.
This, platform has the potential to upend an outdated, outmoded research model. At the very least, it could augment existing research models with a significant increases in data collection frequency and volume. For researchers who spend years trying to collect enough data to analyze, report and ultimately bring new interventions or treatments to patients, this could improve the speed at which research is completed.
What do we know about ResearchKit so far?
At the moment, there has not been a tremendous amount shared. Apple SVP of Operations, Jeff Williams, presented a vision for what ResearchKit could become, but very little else has was revealed. Even the released technical overview (see the technical overview below) is somewhat limited in its explanation of how Apple intends to achieve its long term vision to improve medical research. Here’s what we do know:
- ResearchKit will be released to developers next month. It is likely that some of the components of the toolkit have not yet been worked out which is why it’s not quite ready for prime time. It will be interesting to see how, if at all, Apple adjusts its plans based on feedback from scientists and developers alike between now and April.
- ResearchKit will be Open Source This means Apple will give free access to the underlying blueprint for ResearchKit to anyone who wants it. They will rely on a broader community of developers and researchers to help evolve the platform to make it better over time. Apple has made Open Source development a part of their software strategy for years and this is precisely the kind of program that will benefit from the crowd sourced evolution.
- ResearchKit will not rely on Apple servers for data storage. This means that developers will likely have to solve the concerns around compliant data storage and transmission themselves. Companies that are considered Covered Entities or Business Associates under HIPAA (hospitals, insurers, etc.) will have to pay special attention to this consideration as Apple is not offering solutions as part of ResearchKit (so far). On the other hand, this also means that concerns about Apple knowing more than they should about our health can be mitigated by closed-loop data collection outside of their control.
- ResearchKit will integrate with external data too. This means that Apple’s toolkit will allow for 3rd party integration. This opens up opportunities for research using the hundreds of health and fitness accessories that connect through Bluetooth to your iPhone. The expansion of the sensors to additional monitors allows for opportunities like an asthma monitoring app that can study a participant’s breathing with a Bluetooth enabled inhaler (see What’s been done with research kit already?)
- ResearchKit has global potential. This means that research can be conducted virtually through apps where historically participants needed to be in close proximity to researchers. Not only does this improve the rate of recruitment for research, but it opens up the range of participants to include anywhere a person can use an iPhone. Current research can be biased by the participants lacking racial or geographic diversity. ResearchKit has the potential to improve that significantly.
- ResearchKit has the potential to standardize. This means that Apple is considering the user experience of app-based research as being fundamental to its adoption. Normalizing the way people collect data for scientific research can improve consistency and IRB review of the protocols.
What’s been done with ResearchKit already?
Apple included in its announcement, five apps developed in partnership with early testers of ResearchKit. These apps, which are available on iTunes (only available in the US) showcase the potential power of the toolkit in a variety of condition categories. From neurology to oncology, these apps provide insight into how ResearchKit could be implemented in a number of intelligent ways.
- Asthma: An app was developed to help participants self-manage their asthma by identifying areas with poorer air quality. The app monitors symptoms of individuals across thousands of geographic areas.
- Parkinson’s Disease: An app to measure changes in dexterity, gait and balance using the built-in sensors on the iPhone. The app will help researchers understand more about how symptoms manifest in Parkinson’s patients over time.
- Diabetes: An app called “GlucoSuccess” to better understand how patients’ activity level, diet and medications influence the glucose levels in their blood. The app provides a feedback loop to help patients make appropriate changes in their lifestyle.
- Breast Cancer: “Share the Journey”, an app to help monitor how patients respond to chemotherapy across a number of dimensions including mood, energy level and cognition.
- Cardiovascular Disease: “MyHeart Counts” app which uses regular surveys to understand how participants’ lifestyles are correlated to their cardiovascular health. Comparing users’ biological age to their “Heart Age” gives a risk assessment percentage for both heart disease and stroke.
Technical overview – how might it work?
Apple has already released a technical overview for ResearchKit in advance of the actual kit becoming available to developers. It’s only version 1.0 but here’s what it tells us: ResearchKit will likely be released with only three modules but that number will grow as the Open Source community contributes to future releases.
The ResearchKit survey module will give developers a pre-built front-end user experience to create surveys for research participants. The module will have a series of preset survey instruments from which researchers can choose. The survey module is localized to allow for customization of question and answer sets at the geographic level.
Informed Consent Module
This is a flexible module to allow for multiple types of opt-in to participation in research and data sharing. These visual workflow templates provide opportunities to customize language and authorization mechanisms. Pre-approved IRB consent templates can be integrated easily into apps.
This is the actual collection of health data from the iPhone’s sensors. Currently the types of active tasks are limited to:
- Motor Activities – using the touch screen to monitor dexterity by tapping or using the accelerometer and gyroscopes to monitor gait or movement
- Fitness Activities – a 6 minute walk to measure distance travelled over time. Included in the list of fitness sensors is a mention of heartbeat which likely indicates integration with Apple’s watch also launching in April
- Cognition – using the iPhone’s touchscreen to measure spatial memory through a variety of tasks
- Voice – Phonation tests administered using the iPhone’s microphone and speakers
Many of these tasks are aligned to standardized and validated tools or measures to make them more easily integrated into research protocols.
Current limitations – get your hopes up, but not too high
- The toolkit doesn’t manage secure connections between your app and your server. This means that Apple isn’t offering to solve one of the more complex and labor intensive components of health data collection – the secure transmission from a participant’s iPhone to the researcher’s server
- The toolkit does not yet have the functionality to schedule surveys or active tasks for participants. This is a limitation as most research protocols will rely on a cadence of scheduled activities and self-reported data
- The toolkit does not provide compliance with HIPAA or any other privacy regulations. According to the Technical Overview, this is the responsibility of the researchers and their oversight bodies to ensure. This is logical, but is concerning as many researchers may infer incorrectly that the control of PHI is managed by the toolkit in some way
What could this mean for pharmaceutical companies and clinical trials?
ResearchKit is not even born yet, let alone in its infancy. A few companies have already made strides in the use of smartphones for clinical trial data collection (e.g., Vertex Pharmaceuticals conducted research as early as 2011 on the use of smartphones to monitor the progression of Multiple Sclerosis There is a lot of room to grow and experiment using this platform but it will take time to iron out the details.
- If you are interested, it’s worth starting the process early to ensure you have the ability to influence the protocol design for the trial
- Consider what you are trying to achieve with your research. Not all trials or medical research will be suitable for data collection through smartphones
- Work with your IRB or ethics review committees to discuss the research implications of collection of data through an app and how best to manage participant privacy
- Leverage your internal and external HIPAA compliance experts to determine how best to manage data security and Protected Health Information (PHI).
- Consider the use of ResearchKit for patient recruitment as well as for those already enrolled in trials
In conclusion (but reserving judgment)
Even though ResearchKit as a technology on its own may not change research as we know it, when Apple does something, people take notice. These sorts of real-time data collection programs have been possible for years but Apple might be able to make them ubiquitous.
ResearchKit is not (yet) a panacea for helping researchers get over the hurdles of privacy and security inherent in apps. What it does do, is start to get researchers thinking about standards for protocol development that include app-based data collection. These standards may help IRBs (and potentially the FDA) review research more easily knowing that the collection mechanisms (surveys and actions) are based on the same underlying repeatable frameworks.
When the first version of ResearchKit is released to the developer community next month, we will provide additional guidance and analysis of the toolkit. Stay tuned.
If you have any questions about these developments, reach out to your account representative or contact:
- Ari Schaefer: VP, Group Account Director
- Email: firstname.lastname@example.org
- Global Mobile: 416-992-4663