Off the Shelf: What Apple ResearchKit™ Really Provides Researchers and Where You Need to Customize
One of the most exciting new things to happen to research and the connected health movement is Apple's ResearchKit app framework. Released in March of 2015, Apple's ResearchKit promises to expand the scope, scale, and breadth of research by allowing researchers to conduct studies on mobile iOS devices and thereby reaching potential participants anywhere. With approximately 400 million iPhones currently in use worldwide (>100 million in the US alone) ResearchKit holds great promise for researchers to go well beyond their local vicinities and to tap into a broader demographic for the collection of richer, more granular data unbound by geography.
Like any new technology however, much of what ResearchKit offers requires additional development before it can be made study ready. Getting a study into the hands of the intended participant, thinking of participants as users, and collecting the resultant data in a meaningful way requires many researchers to involve themselves in things like front-end/back-end development, user interface (UI)/user experience (UX) design, social media, and digital marketing — things that might come more naturally to a Silicon Valley mobile app startup rather than a researcher at a university or medical institution where available resources might not have evolved enough to support them in the new world of connected health.
To ease these challenges, Apple has done some great work in determining what most mobile app studies will need. For instance, most studies will likely want to conduct surveys, perform tasks, and will require a user consent process. Apple has provided these basic study elements as out-of-the-box modules, which in turn can provide a big boost in getting started. Extending these modules, and building beyond them to create a meaningful ResearchKit app, is where the real work begins.
So, how do you get started? To determine this, you’ll need to know the answers to the following five questions:
What does ResearchKit provide off the shelf?
Let's start with what ResearchKit is. ResearchKit is a framework that provides core modules — it is not a fully functioning app right off the shelf. These modules are useful, and in most cases required, for conducting studies in a mobile app environment, as well as for satisfying IRB and HIPAA requirements. Included in the ResearchKit framework are the following:
- Survey Engine: The survey engine provides the core module for building out the shell structure of your surveys. In this form, the survey will be very basic and will likely require some additional development and design depending on the complexity and branching logic of the survey questions.
- Visual consent flow: The visual consent flow module provides a customizable template to explain the details of your study and obtain a signature from the participant. This is almost always presented in the beginning of the app experience and is typically required by your institutional review board (IRB) or ethics committee. The consent flow can also be used as an opportunity to educate and qualify your participants through a simple questionnaire.
- Active Tasks: Active tasks are used to invite users to perform activities under semi-controlled conditions, using iPhone sensors to collect data (you can view examples of active tasks here). Active tasks are where you can get really creative with your study, and typically where a lot of your additional development will be focused. For instance, when THREAD Research built EpiWatch, we incorporated a timed walking and resting activity that involved Apple Watch for gathering heart rate and accelerometer data. We see active tasks as a great opportunity to explore the possibilities of ResearchKit by using real-time data to engage the participant in new ways, such as gamification, and research as intervention.
- Examples in Objective C and Swift programming languages
- API documentation (for interfacing with your app’s back-end)
What will need to be customized in order to satisfy the study requirements?
Most Researchkit app projects will need customization and development beyond what the basic modules offer and, in fact, I recommend it. By using an iPhone app, participants are no longer in a controlled study environment, and are instead susceptible to habits of content consumption in other apps, attention spans, and user preferences.
Your app needs to keep them engaged to keep them coming back. That brings us to AppCore. AppCore is a layer built on top of ResearchKit and forms the core of the five initial ResearchKit apps. It includes:
- Dashboard with progress graphs
- JSON serialization and deserialization
AppCore is still a work in progress and is not as simple to customize as the basic core modules because it is not available in stand-alone form. The work to utilize any part of AppCore in its current state will require a much higher degree of customization and development as it needs to be extracted from one of the open source ResearchKit app projects that use it (can be found here). One of THREAD Research’s biggest focuses is to make AppCore more accessible to researchers and their institutions.
What kind of user interface and user experience design elements can be incorporated so that participants will remain engaged?
One of the best examples of how customization can improve user engagement is the dashboard. The dashboard is a great way to capture the interest of the participant by showing them a visually pleasing representation of their own data, as well as comparisons with other participants.
In the PRIDE study, THREAD Research developed a stunning dashboard which illustrates a “Participants Like Me” graph which users have found to be very engaging. Since the data shown on the dashboard is dynamic, a user can come back each day to see the change in the latest results of their own data, as well as that of their cohorts.
To read more about design and usability, please read our blog post on usability and why it’s important for your mobile study.
How will the data be collected?
ResearchKit does not include a data management solution but the framework can be used with the backend service of your choice. Consideration should be put into the provider’s data privacy and security practices, and you should ensure that those practices satisfy HIPAA compliance.
Other considerations that should be evaluated in your backend provider search:
- ResearchKit API documentation to determine how your app communicates with your backend
- How your data is collected
- How you will gain access to your data
- Where you plan to conduct your study
- What your required reach will be
- How much participation you expect to have
- What your scalability requirements are
Every study has different requirements and there are a number of services to choose from. My recommendation is to look for a service that can maintain your backend for you so that you don’t have to worry about uptime and server upgrades, and which utilizes a secure and reliable cloud infrastructure such as Amazon Web Services (AWS).
How can I get my study into the hands of more participants?
So you've built your ResearchKit app, now what? Well, now you have to submit it for review by Apple and get it released into the App Store. But then what? How do you let participants know about your study? The answer of course is spreading the word through multiple channels.
To help inform you on some interesting ways to increase study participation and engagement for your research app, I recommend the following blog posts:
At THREAD Research, our mission is to answer every human biological question, but we can’t do it alone. I encourage you to develop your study for ResearchKit and reach out to us for any questions you might have. ResearchKit may not be an out of the box solution, but with some good planning, savvy development, and well executed customizations, you can provide an extraordinary study experience with the potential to reach well beyond what has been possible before.
The ResearchKit framework, including modules and tools can be found and downloaded from GitHub here, https://github.com/researchkit/researchkit
Design guidelines for ResearchKit apps can be found here: http://researchkit.org/hig/index.html
API Reference for ResearchKit can be found here: http://researchkit.org/docs/index.html