By Illana A. Stanley, Ph.D.

INTRODUCTION TO APPLE’S RESEARCH KIT

Apple, the company behind the iPhone and the maker of one of the most popular mobile operating systems, iOS, announced on March 9, 2015 an addition to the iOS operating system called ResearchKit. On April 14, 2015, ResearchKit became available for general use [1]. What is ResearchKit? It is a framework (a related collection of code) that is specifically intended to allow clinicians and scientists to develop mobile apps that can be used to conduct clinical research. The goal is to take advantage of the convenience of current mobile technology and the prevalence of smartphone usage and enable researchers to improve upon the way that clinical research is currently done.

THE CURRENT STATE OF CLINICAL RESEARCH

Clinical research is used to learn, in an empirical manner, how different variables affect some aspect of human health. For example, a researcher might want to know if sitting in front of a computer screen for a certain number of hours per day is correlated with negative outcomes in overall health or correlated with a specific disease. Or a researcher might want to be informed about risk factors for breast cancer. By surveying breast cancer patients and healthy controls to see if there are common activities or factors which co-segregate with the breast cancer patient group, researchers can gain insights into risk factors for this disease. Studies like this help inform a wide range of policies from the standard protocols that doctors follow to the laws and regulations enforced by the government. Therefore, any improvements that can be made to the execution of this research has far-ranging consequences for public health.

There are many challenges that researchers currently face when conducting clinical studies. One of the main challenges is the process of enrolling participants for a study. The organizers of a study usually advertise it using flyers, posters, mailing lists, phone calls etc. Depending on the scale of the institution organizing the study, the potential pool of participants can be geographically limited to being local to that institution. In order to generate useful statistical inference, a clinical study needs a large number of participants to be enrolled. Getting enough participants to have the statistical power necessary for reaching meaningful conclusions is often a major challenge for researchers. It can take a lot of time and money to enroll enough people to satisfy this goal. Once a clinical study gets the attention of a prospective subject, it is very important to obtain informed consent from that person. In the United States, this is generally regulated at the institutional level by the Institutional Review Board (IRB) which must approve research that includes human subjects. This is to ensure that the studies are performed ethically and that data collected cannot violate a patient or participant’s right to privacy. Getting informed consent from participants in a manner compliant with an institutions IRB often requires lots of paper work that the subject must complete in person [2, 3].

Another challenge of clinical studies is that they can often take place over a long span of time (one or more years). During that time data often needs to be collected from the participants at specified intervals. To reduce the inconvenience to the participant, thus reducing the chances that they drop out, the intervals between data collection need to be reasonably long (once a year, once a quarter, etc.). However, there are many studies for which a more continuous view of the data would be beneficial [4]. And there is always the chance that participants will still drop out of the study before completion which further compounds the difficulty in getting enough subjects to make statistical inferences.

ADVANTAGES OF USING MOBILE TECHNOLOGY IN CLINICAL RESEARCH

There are several potential advantages that the current popularity of mobile internet connected devices may provide to research in general and clinical research studies in particular. In many developed markets, like the United States, smartphone usage has been increasing steadily with greater than 50% smartphone penetration in the United states [5]. This means that research which takes advantage of mobile technology could reach a large subset of the population. The iPhone makes up a significant share of the mobile market, depending on the country, with iPhone usage being particularly high in the United States at 42.6% of smartphone users [6]. The success of the iPhone, together with the success of the Apple App Store, where ResearchKit based apps would be distributed, means that researchers have the ability to reach large numbers of potential participants in a geographically unrestricted manner.

A major initial hurdle for the use of mobile technology in clinical research is satisfying regulatory requirements. When Apple introduced ResearchKit, they revealed partnerships with 5 different academic studies that are being used as a proof of concept on the utility of ResearchKit based apps for clinical research. Each of these 5 apps was able to obtain IRB approval from their respective institutions using e-consent forms in place of the lengthy paperwork that is usually required for such studies [3]. This demonstrates that a mobile research platform can be compliant with the current regulations and ethical guidelines.

RESEARCHKIT FROM A RESEARCHER’S PERSPECTIVE

For clinicians and scientists, ResearchKit offers the chance to easily develop apps that do tasks commonly encountered in clinical studies without having to write all the code for such activities from scratch. This is helpful because many of the people involved in clinical research studies are not familiar with mobile app development and do not have the financial resources to outsource this job for a custom app solution. ResearchKit offers relatively easy to use templates that are specifically tailored to the needs of an app designed for clinical studies. The code that makes up ResearchKit is open source, meaning that the source code is available for others to see and can be modified by anyone, not just by Apple. In addition, the 5 apps that have been developed thus far with ResearchKit are also open source so that other researchers can learn from the current implementation and add to it. In this way ResearchKit has the potential to grow and adapt as new needs are discovered.

To simplify app development for clinical research, ResearchKit offers templates or modules for 3 common functions. These include a module for designing surveys. The template includes surveys that accept multiple types of input (multiple choice questions, dates, times etc.). There is also an entire module to help researchers create the electronic consent form needed for approval of clinical research apps. There are templates within this module for providing information to users to describe the scope of the study. This module also contains the ability to allow users to give separate consent for the sharing of collected data with other institutions. There is even a template for quizzing users on the previously presented information to confirm that the user of the app fully understands the terms of the agreement. ResearchKit also makes it easy for the developer of an app to export the consent form as a PDF so that a hard copy can be kept with the user or mailed to external bodies that might need a physical copy of this documentation. The third major type of module that ResearchKit provides is for collecting data from active tasks. This would include game-like cognitive memory function quizzes to test spatial recall. This module also includes templates for testing motor skills using prompts that can collect user input (tapping the touchscreen) or sensor input from the iPhone such as the use of the gyroscope to determine if a subject is walking in a straight line. ResearchKit is also able to make use of the health data that is already being collected by various fitness apps that can be used on the iPhone. Interaction with other apps, however, requires explicit permission from the user [7].

Developers are not limited to the modules offered in ResearchKit, but they provide a good outline and starting point to make things easier for those unfamiliar with making mobile apps for the iPhone. Because ResearchKit is open source, developers who come up with new modules will be able to add them to ResearchKit so that the entire research community can take advantage of new ways to enhance apps for clinical research.

CURRENT STATUS OF APPS USING RESEARCHKIT

Apps developed using ResearchKit have only been active on the Apple App Store since March 9th of this year, and it is still too early to tell if this new framework will provide meaningful differences in the way most major institutions conduct clinical research. It is also too early to know if there will be any unforeseen issues. However, initial evidence from at least one of the first 5 pilot apps is showing promise. The creators of Asthma Health, an app developed in a collaboration between LifeMap Solutions and Mount Sinai Hospital in New York, have reported some promising initial observations in a blog post on ResearchKit’s official website for developers. LifeMap Solutions reports that a significant portion of participants were willing to give e-consent through the mobile app, a promising finding considering the potential hurdle of getting users to trust an app on their phones over more traditional paper consent forms. And importantly, the developers of Asthma Health noted that the majority of users also return to the app the next day. A common challenge for mobile developers in general is maintaining user engagement with their app over time. In the case of the Asthma Health app, the developers found that a significant proportion of users remained engaged with the app and returned to use it over the span of several weeks. This return engagement could be boosted by automated weekly reminders [8].

However, user reviews published on the Apple App Store for the first 5 apps making use of Apple’s ResearchKit are mixed. As of the writing of this article some apps have received fairly low scores (2-3 stars), and reviewer comments suggest some of these apps suffer from stability issues and frequent crashes. This might suggest that having apps created by researchers not accustomed to developing for iOS comes with caveats. Alternatively, because the apps are relatively new, there are still bugs that need to be fixed, which is not uncommon for newly released software. More time will be needed to determine if the early bugs in some of the first 5 ResearchKit apps can be resolved.

LIMITATIONS OF RESEARCHKIT

Although ResearchKit has the potential to alleviate several challenges associated with current methods of doing clinical research there are limitations to both ResearchKit and the use of clinical research apps in general that both developers and end-users should be aware of.

Data collected by apps that use ResearchKit must be securely transmitted back to the servers of the institution conducting the study. This secure transfer is the responsibility of the app developer. Unlike with other services, Apple does not assist or handle any secure data transfer for these apps [7]. In some ways this can be beneficial to both the study participants and Apple because sensitive data is never stored on Apple’s servers. However this puts the responsibility on the app developers and it also means that the consumer will need to be able to trust that each individual developer of a ResearchKit app can handle their data securely.

The other major limitation inherent to ResearchKit is the requirement for an iPhone running a relatively recent version of iOS. While smartphones are quickly becoming ubiquitous in many regions the percentage of smartphone users who are iPhone users is at best ~50% (in the United States) and in some cases much lower [6, 9]. Restricting a clinical study to participants who can afford the relatively expensive iPhone has the potential to introduce significant bias into a study. Researchers will need to account for this bias in their interpretation of results.

Conducting clinical research using mobile applications has the potential to increase the convenience of these studies for both researchers and subjects, however, there will always be studies for which the kind of data collected by a mobile application will not be sufficient. Studies that require medical tests or procedures will still be limited by the ability of the participants to reach clinics for testing.

Other limitations of ResearchKit include the inability to confirm the veracity of a participant’s statements. For example, 2 out of the 5 current ResearchKit apps require that participants have the disease being studied. However, there is no way for those running the study to confirm this.

As with many other aspects of smartphone and internet usage, data security will continue to be a concern for mobile applications that are used for clinical research. ResearchKit attempts to address some of these concerns by preventing Apple from handling any participant data directly and by offering granular control to the user as to how the apps can use and share their data [7]. However, privacy concerns and perceived privacy concerns may still be an issue for some potential users of ResearchKit based apps.

WHAT IS NEXT?

It is still too early to tell how much of an impact ResearchKit will make on the field of clinical research and if using these types of mobile apps will provide tangible benefits to clinical studies. Apple is not the only company investigating the potential of consumer facing mobile technology in the health and research sphere. Several companies including Google and Pebble have integrated fitness tracking into their products. From the perspective of the research community ResearchKit could be a useful concept of how to standardize an approach to doing clinical research on mobile operating systems. However, a system that was cross-platform, including at least the three major mobile operating systems, Google, Apple and Windows Phone, would likely be more helpful in reaching the largest number of potential users [9]. Regardless, ResearchKit represents an important attempt to move clinical research into the mobile and web-connected future.

REFERENCES

1. Introducing ResearchKit (no date) Available at: http://researchkit.github.io (Accessed: 8 May 2015)

2. Timmer, J. (2015) Inside Apple’s ResearchKit. Available at: http://arstechnica.com/science/2015/04/inside-apples-researchkit/ (Accessed: 8 May 2015)
3. Pogue, D. (2015) Apple’s ResearchKit Takes Medical Research Years Into the Future. Available at: https://www.yahoo.com/tech/apples-researchkit-takes-medical-research-years-117781601479.html?src=rss (Accessed: 10 May 2015)
4. Apple Inc. (no date) ResearchKit. Available at: https://www.apple.com/researchkit/ (Accessed: 8 May 2015)
5. (no date) Available at: https://www.consumerbarometer.com/en/insights/?countryCode=GL (Accessed: 10 May 2015)

6. Lella, A. (2015) ComScore Reports March 2015 U.S. Smartphone Subscriber Market Share. Available at: http://www.comscore.com/Insights/Market-Rankings/comScore-Reports-March-2015-US-Smartphone-Subscriber-Market-Share (Accessed: 11 May 2015)
7. ResearchKit Framework Programming Guide (2015) Available at: http://researchkit.github.io/docs/docs/Overview/GuideOverview.html (Accessed: 10 May 2015)

8. LifeMap Solutions (2015) Preliminary User-Engagement Learning from our ResearchKit-Enabled Asthma Health App. Available at: http://researchkit.github.io/blog.html#article-2 (Accessed: 10 May 2015)
9. Android and iOS Squeeze the Competition, Swelling to 96.3% of the Smartphone Operating System Market for Both 4Q14 and CY14, According to IDC (2015) Available at: http://www.idc.com/getdoc.jsp?containerId=prUS25450615 (Accessed: 10 May 2015)