Monday, March 9, 2015

Flash Thoughts on Apple's ResearchKit Program and Citizen Science

Today's big news in technology circles was Apple's unveiling the new iWatch.  But for researchers and citizen scientists there was an even bigger set of announcements, Apple's introducing the new ResearchKit suite of tools.

Don't get confused, ResearchKit is not a new app.  Instead it is a toolkit that lets medical researchers create their own applications using functions and data from a user's Apple device. But not just professional developers.  This will be an Open Source toolkit allowing anyone with a computer and interest in medical research to utilize it.

At the most basic level ResearchKit creates an electronic connection between researchers and their test subjects, allowing the constant collection, and reporting, of data.  It blossoms from there though, as researchers can use exercise data from fitness applications and Apple's own HealthKit app, as well as pulling personal health information it.  Other functions are available too, such as collecting GPS data (to connect medical incidents to locations) and voice recording.  All a researcher needs is some coding experience and a willingness of people to join.

In promoting ResearchKit Apple discussed a number of initial apps designed to help existing researchers.  One in particular (Asthma Health) not only collects data from asthma sufferers and interfaces with bluetooth-enabled inhalers (to analyze that person's lung function), but it also maps any asthma attacks to maps of known pathogens and potential environmental toxins.  So researchers can test which of those potential triggers are really connected to asthma, and which may not be. Other examples included:

  • mPower - An app that analyzes a user's voice, walking motion, and finger function (by tapping on the phone) to determine the disease's progression at any given moment.
  • GlucoSuccess - A diabetes app that monitors behavior and exercise levels from the phone and compares it to user-reported glucose levels.
Although ResearchKit won't be available until next month, these apps are all published and ready to download now.


Benefits to Citizen Science
The promise of these individual programs are obvious, and Apple's Keynote (available here at the 16:23 mark) goes through many of them.  But what most interests me is that the larger problems ResearchKit tries to solve, and those that are highlighted by Apple itself, are many of the same issues brought up in my own Keys to Successful Citizen Science Projects research.

  • Limited Participation: Recruitment is often a large problem with clinical research on human participants.  It is often hard for researchers to find willing subjects with the specific conditions one they are looking for.  This is compounded by the difficulty in working with those patients if they are not in the researcher's geographic area, or if it requires frequent visits to a doctor.  These obstacles add up as major roadblocks to research.  But this is drastically lessened when millions of people have an iPhone and you Make Participation Convenient by integrating your app with it.  Massive recruitment campaigns are no longer necessary, nor are lengthy sign-up questionnaires when downloading the app is a click away. ReserchKit apps can help solve this by Making Participation Easy through the iPhones existing health monitoring data and user-friendly interfaces.  They more you can automate or require fewer steps by users, the more people will remain as participants through the end.
  • Communication Flow: Providing Feedback is important for keeping people involved in projects after they sign up, improving retention and reducing recruitment costs.  The more they see a benefit to their work the more they want to stay involved.  But it's also just fair.  If they give you so much time and effort (or in the case of health data, trust), they deserve to at least watch what researchers are doing with it.  Similarly, Interacting in Real Time with participants also meets that feedback requirement, but int he case of medical data, can provide information they can use on a daily basis to improve their health.   

There are many more that are not addressed by Apple but which I think are strong contenders for future programs.  For example, research apps that also include a gaming component, or a challenge component, will keep people involved by Benefiting the User.  Gamification has worked in many other citizen science projects and there's no reason it can't work here too.  Researchers could also use the tracking components of each phone to provide rewards, or connect to the many learning applications to help educate participants on the conditions being researched.  And some budding entrepreneur could make participation easier by using HealthKit information not only as the data for the research, but also to help you identify which of the many ResearchKit apps to use (in the future there may be MANY to choose from).  So there are opportunities for scientific gain as well as financial gain.


Of course there are many issues that must still be sorted out.  For one, the rules for obtaining informed consent can vary greatly between research institutions, and although Apple has come up with an easy way to handle it for these projects, there is no guarantee that other researchers will be able to do the same thing.  Additionally, we have already seen through 23andMe and other cases that entrepreneurial firms with big health ideas often run into problems navigating the labyrinth of regulations surrounding privacy, medical ethics, and patient protection in the health care industry.  So any newcomers, entrepreneurs, or venture capital firms hoping to take advantage of HealthKit would be well advised to work through those issues first.

There is a lot of potential here.  Obviously Apple likes to hype it's own creations and there are still many obstacles to overcome.  But if we design apps carefully and learn from the existing keys to successful citizen science projects, the potential this brings will be even greater.






No comments:

Post a Comment