Jim Gearhart

by Jim Gearhart

ResearchKit Studies: What Might an IRB Think?

In a previous post I looked at the five studies that heralded the arrival of Apple’s new ResearchKit. In this post I’ll consider some regulatory and ethics issues surrounding these five studies. (I should say that I work at Quorum Review, but I am not a member of the IRB. I have had CIP certification, and an attorney on our regulatory team helped draft this).

IRB Review: Each of the five ResearchKit studies did have approval from an IRB (Institutional Review Boards are ethical committees that review and approve medical research in the United States). For three of the studies, an IRB at the hospital or university running the study did the review; for the other two an independent IRB approved the research.

Practically speaking, the interactions within the apps all had similar steps, which could mean the four IRBs reached similar conclusions about their studies. The structure of the five studies suggest that the IRBs all decided:

  • the research was of minimal risk to its subjects
  • the measures taken to ensure privacy and confidentiality were sufficient
  • a remote informed consent process was acceptable

Minimal Risk Research – And Beyond: If the reviewing IRB finds that a medical study does not create risks greater than the experiences of a typical day, then the regulations surrounding the study offer more flexibility. This is called Minimal Risk Research and could explain the relatively short and informal informed consent processes for these first ResearchKit studies. It will be interesting to see how IRBs respond once someone uses the ResearchKit to support a more risky study, such as one that tests an investigational drug or device.

Privacy & Confidentiality/HIPAA: The ResearchKit apps raise the prospect of gathering a person’s name, age, height, weight, and medical conditions. In a healthcare context, that kind of information quickly becomes Protected Health Information (PHI) and so invokes HIPAA’s stringent privacy requirements. It is something to assess in any medical research.

Apple appears to have distanced itself from the HIPAA fray, at least for now. “Apple will not see your data,” read the huge display behind CEO Tim Cook, and the ResearchKit’s Technical Overview passes the HIPAA buck along. In its Current Limitations section, the Overview states that the ResearchKit does not comply with “international research regulations and HIPAA guidelines,” adding, “These are the researcher’s responsibility.”

But it does look as if the HIPAA rules were considered not to apply to this first batch of ResearchKit studies, for a couple of reasons. First, the data being gathered does not involve a treatment relationship and so HIPAA would arguably not apply. For another, in each of the studies an intermediary will scrub the data after it leaves an iPhone and before it reaches a researcher. That de-identification step can essentially remove the research and the data from the HIPAA regulatory framework. (In practical matters, many are quick to observe that true anonymity is not possible.)

Informed Consent: One notable feature of all these studies was that consenting could happen remotely and without a signature. Again, this probably is because the IRBs all found the studies to be Minimal Risk. The risk and benefits sections of the studies were all brief (Typically: the survey questions might make you uncomfortable, versus researchers might learn something that could help others), but riskier studies will require lengthier explanations and personal discussions with a qualified professional from the research team. Regarding electronic signatures, Apple’s Technical Overview says that the ResearchKit can handle them, when they’re required, but details are still forthcoming.

To me, the study apps did a good job of informing. The short explanations with graphics were instructive; videos added dimension; and in the end the final consent form document contained no surprises (each study delivers a PDF version of the full consent form). One shortcoming (that an article in The Verge pointed out) was that the process does not allow for real-time questions.


Apple has crafted an appealing launch for this app-builder, and the research institutions helping out appear to have followed ethics guidelines. Time – and more complicated research – could test out the possible weaknesses. For me, the ResearchKit marks a significant step in improving technologies for clinical trials, and is something for researchers and IRB reviewers to watch closely.

Tags: , , , ,