This blog article continues a conversation with mHealth expert Jeremy Block (PhD, MPP), about Apple’s new rules for apps that use the ResearchKit platform. Among other new requirements, any apps using the ResearchKit platform now need an independent ethics review and an informed consent process.
Jim Gearhart: Do you have any other thoughts about Apple’s new ResearchKit guidelines?
Jeremy Block: We can draw parallels between how Apple approached ResearchKit and how the FDA approached its mHealth software guidelines. Apple watched what was submitted to the App Store and went through its review process, then created guidance when it saw how people were using the platform. The FDA took a similar wait-and-see approach, and, after a period of time, indicated it planned to use administrative enforcement and discretion.
These approaches reflect a philosophy of supporting innovation, a desire to give people space to roam and not put restrictions in too early or too heavily. They also are not basing restrictions on specific technologies. That would be an almost-instant death to innovation.
Services like the App Store and Google Play have a unique role in digital research. Google and Apple are essentially saying, We’re not conducting the research; we’re merely serving it up.
And the FDA acknowledged that when it stated it would not a priori regulate a smart phone with an mHealth app as a medical device. Instead, that regulatory decision depends on how the program operates, what tasks it performs, and what risks it creates. Which looks a lot like an assessment of a study design, and IRBs assess study designs all the time to make determinations. Very familiar process, where the app or device now represents the study design.
JG: The new guidance says an app must include informed consent, but does not provide details.
JB: It makes sense that Apple was reasonably mute on informed consent. It mandated that everyone have an ethics review, and that the app include informed consent. But Apple will rely on the review to confirm the informed consent content and reiterate any regulatory requirements.
However, here’s the kicker: This means an IRB cannot waive consent for an Apple ResearchKit study. Our regulations have carve-outs for waivers of consent and waivers of documentation of consent. Apple’s guidance doesn’t give that flexibility. For now, that’s probably a very prudent choice. We should be deliberate about making allowances for consent in digital health.
JB: My approach has been to act as though you’re going to be regulated, because it will benefit people for all sorts of reasons. I put everything through the process I would use for any study. You look through the study materials and ask, What is the program doing? What are the risks? How does this meet the regulations and guiding principles of research ethics?
At Mount Sinai, I pushed for a focus only on things that would be approvable as minimal risk and to avoid any semblance of a real or implied care relationship between participant and healthcare provider. That turned out to be prudent because risk now is part of the threshold questions for regulation. More changes, learning, and development will come when we see our first greater-than-minimal-risk ResearchKit study.
JG: How should IRBs approach these studies?
JB: Mobile health studies have specific areas to focus on for IRBs where some learning needs to happen; data and information security, confidentiality, and privacy are the biggest ones. These are all tied to the technology and how it’s the implemented, so the IRB needs specific knowledge to ask the right questions or look for the answers. We are at the regulatory threshold of needing IT and software or hardware development experts as IRB consultants or even voting members. Our regulations require we have expertise to review studies. We now need the IT consultants’ expertise.
I would like thank Jeremy Block for sharing his thoughts on Apple’s new ResearchKit requirements. It will be thrilling to see which of Jeremy’s recommendations come to life as mHealth moves to the center of the research stage.