Written by John Wilbanks and Erin Holve.
Early last month, Apple announced ResearchKit, a framework for running clinical studies on the iPhone, as well as the first five applications that use ResearchKit. These studies are being run for Parkinsons disease and more than 50,000 people have enrolled in the first two weeks. Each of those participants enrolled via a novel e-consent process developed by Sage Bionetworks as an EDM Forum Collaborative Project.
There’s a tremendous amount of excitement about these new tools in ResearchKit because they have the potential to make it far easier for more individuals to participate in patient-centered outcomes research (PCOR). Doing so will ideally lead to more interest (and perhaps investment) in research questions prioritized by individuals, patients, caregivers, and communities, as well as the faster uptake and use of new evidence that comes out of these studies. At the same time, privacy and data security are paramount concerns that must be addressed to preserve the public’s trust in research.
Consent is an important part of participant engagement and PCOR for many reasons. From an ethical perspective consent is an important strategy to ensure research participants preserve their autonomy and are fully informed partners in research who understand any risks of participating in research. Consent also asserts an individual’s right to their own data, as authorized by the Health Insurance Portability and Accountability Act (HIPAA) and so can support access and engagement with participants’ own data. However, lengthy consent documents may not promote comprehension of key concepts such as research risks, or facilitate engagement. This is where participant-centered e-consent comes in.
At the beginning of our collaboration to build the e-consent process, we co-developed a multistakeholder process that began with stakeholder interviews from throughout the EDM Forum’s collaborative network, including ethicists, technologists, scientists, patient advocates, clinical data specialists, and more. These interviews fed a user-centered design process that led to the Sage Bionetworks design for e-consent: a tiered, hierarchical information “stack” that use iconography and simple language, paired with a quiz-based assessment of comprehension post-consent before data collection begins. The goal of the approach is to help those running studies to convert a static consent document into processes that extract the most important concepts of an e-research study, and make those concepts easier to understand and remember than rendering large quantities of text on a small screen.
The first “tier” of the information stack is icon-dominant. The screen is divided into elements indicating progress through the consent process, with the primary element devoted to a visual indication of the essential clinical study concept. Concepts for the final RK consents include: study activities, phone sensor data, data processing and use, data protection, potential benefits, risks to privacy, withdrawal, and more. It is essential for study designers to stop and consider: what are the concepts that participants must know to make an informed choice? Those concepts then must be presented at the top “tier” of the consent process.
The second “tier” is text-dominant. This is where clear language about the concept is expressed, providing more detail for those who wish. In future releases we plan to expand the text-dominant tier to include hyperlinks, multimedia, and other, more interactive ways to communicate essential clinical concepts to participants.
The slides below are from the Parkinson mPower app, but reflect a fundamentally shared experience across all five apps at the first tier. Each app of course has completely different text in their second tier of “learn more” screens, and different consent document. Each app was approved by the IRB for the sponsoring institution.
After we developed the first implementation of the concept, as part of our collaborative methods project, the EDM Forum convened a workshop featuring bioethicists, privacy experts, patient advocates, technologists, clinicians, data scientists, and more. Participants came from government, non profit, academia, and corporate sectors. During the workshop we collectively reviewed the core concept of tiered information access as well as the first set of designs for an actual e-consent process.
Thanks to extensive feedback from the workshop and other stakeholders, we began a phase of refinement and regulatory submission. Our Parkinsons study was approved by Western IRB in the fall of 2014, but rather than launching the study immediately, we continued showing our designs to stakeholders and refining. We gave a series of public lectures and posted slides to the internet, using social media and other systems to attempt to gain the widest possible feedback.
As we showed the designs, many parties expressed interest in using one part or another of the design methodology. So we decided to release all of our work as an open source “toolkit” for anyone interested. We released the alpha version of the toolkit during the Trans-NIH Workshop to Explore the Ethical, Legal and Social Implications (ELSI) of Citizen Science ().
All five of the studies launched on March 9 and have achieved impressive enrollment numbers, but the work is far from done. We plan to study the “informedness” of our cohorts over time, and to continue developing additional methods and practices for using technology to assist in making informed consent more “informed” – although the key fundamentally is not technology, but the forcing function that makes a clinician really think about the key concepts in her study, and how to represent them visually.
AcademyHealth’s EDM Forum is committed to bridging diverse perspectives to collaborate and advance these discussions that proved valuable to the development of the e-consent process. The community is welcome to join in the discussion on twitter (@edm_ah) and participate in our ongoing activities by emailing email@example.com.
The EDM Forum was created through a cooperative agreement from the Agency of Healthcare Research and Quality (AHRQ) grant (U13 HS19564-01). Ongoing support for the EDM Forum and the Concordium meeting comes from AHRQ U18 HS022789-01. Meeting materials, presentations, and eGEMs publications do not reflect the official view of AHRQ or the United States Department of Health and Human Services.