Over the last few years, the use of smartphone apps in minimal risk research has grown substantially. Researchers test the effectiveness of apps in tracking subjects’ physical fitness, managing chronic conditions, tracking symptom progression, and use in areas such as smoking cessation. A search for “smartphone” in ClinicalTrials.gov shows 275 registered trials using the devices — up from 45 in late 2012. (For more information, see “Ensuring clinical trial compliance? There’s an app for that” in the December 2012 issue of IRB Advisor.)
Now, entire research studies are being run through a handful of apps, released in mid-March, through the iTunes App Store. These apps consent study participants, determine eligibility of participants, and run study tasks — all without the need for participants to travel to study sites.
These studies are minimal risk and include the following:
- Asthma Health, an asthma symptom-tracking app study run by Icahn School of Medicine at Mount Sinai and Weill Cornell Medical College, both in New York. The app also uses GPS tracking to provide air quality updates in users’ locations.
- mPower, developed by the University of Rochester in New York and Sage Bionetworks, a nonprofit research institution in Seattle. The app collects data on gait, voice changes, balance, dexterity, and memory in subjects with Parkinson’s disease.
- Share the Journey measures long-term post-chemotherapy symptoms in breast cancer patients. The study was developed by Dana-Farber Cancer Institute, University of California-Los Angeles Fielding School of Public Health, Penn Medicine, and Sage Bionetworks.
- GlucoSuccess, run by Massachusetts General Hospital, seeks to understand how diabetic patients’ lifestyles affect glucose levels.
- MyHeart Counts, which uses participants’ physical activity and lifestyle information as well as surveys to evaluate risks of cardiovascular disease. The study is run by Stanford University in CA, and University of Oxford in the United Kingdom.
Investigators can use apps to harness the features and data-gathering abilities of the smartphone, experts say.
“There is potential for great value with this type of data,” says David Forster, JD, CIP, chief compliance officer of WIRB-Copernicus Group (WCG) in Princeton, NJ. WCG’s subsidiary, Western IRB (WIRB) is the IRB of record for the mPower and Share the Journey studies. “A lot of researchers believe they have good methods to come up with data and remove the noise. There is potential for this type of information to grow.”
The studies can also have a massive reach: By conducting a study through an app, subjects do not have to live near a study site to participate, which means researchers could have a much larger subject pool. “You can get a cross-section of the population; you can get people from every nook and cranny [of the country] and the only requirement is that they have an iPhone and you’re not geographically limited to people who live near the research site,” says Jeremy Block, PhD, MPP, assistant professor of Population Health Science and Policy, and an IRB chair/vice-chair at multiple hospitals within the Icahn School of Medicine at Mount Sinai in New York City.
Within two weeks of the app’s March release, 5,000 subjects had enrolled in the study. “Sometimes it can take years to get 300 people for a study,” Block says. “There are a lot of potential benefits for doing research this way.”
Bringing these studies to smartphone apps was not without challenges: There is very little in the way of guidance for such studies, and issues such as data security and informed consent were at the top of the list.
“It’s absolutely a new area, and it’s been a part of our challenge,” says John Wilbanks, chief commons officer at Sage Bionetworks.
The original idea for the studies was to use an open source Web platform that would allow patients to participate more directly in research, Wilbanks says. “In working with certain patient communities with Parkinson’s disease and breast cancer, we can get into their phones to get them more into research and measure post-chemotherapy cognitive ability in cancer patients, and Parkinson’s disease symptoms. We can get quantitative data from sensors in the phones and qualitative data from surveys without patients going to a clinical site. Since [the phones] can return results easily and track progress easily, mobile was where our charge would take us.”
Back to basics
Ensuring that the studies fell within the regulations meant going back to basics, Block says. The interdisciplinary team of investigators who created the Asthma Health app approached Block about lending an IRB perspective to the development process.
“They approached my boss — the Senior Associate Dean for Research, Jeff Silverstein, MD, — and asked to have a conversation [about the app] because they wanted to talk about their idea and anticipate any issues,” Block says. “I knew right away that this would be a new and interesting area where people haven’t done anything yet. It was a purposeful choice on my part; I know the regulations and what needed to be considered during the design process of the study and the consent, and I wanted to be in it from the beginning.”
Block went step by step through the protocol to make sure all regulatory questions were answered. “For the consent process, confidentiality, security, how the individual will interact with the app — for all aspects, I asked all of those questions and if the answers weren’t enough, I said, ‘Let’s do more here and brainstorm how to accomplish it. If I was a reviewer, I would ask these 15 questions, so let’s sit down right now and try to answer them.’”
Multi-tiered informed consent
Ensuring subjects fully understand the study objectives and informed consent was another challenge, Block says. Since there would be no face-to-face interaction with an investigator to explain IC, the team took advantage of the smartphone capabilities to offer multiple IC methods within the Asthma Health app. “After reviewing IRB literature and in conversations with others, we found it’s preferable to have multiple modalities [of informed consent] to interact with a subject,” he says. “There are three different modalities for consent and interacting with the subject and making sure they are fully informed — three different ways to try to do it is pretty good.”
The three methods include:
- A brief video in which investigators explain the purpose of the asthma study.
- An interactive consent in which participants can swipe through brief explanations of the data-gathering process, time commitment, privacy, potential benefits, study tasks, and other information. Most of these screens include a Learn More button, which participants can tap to read more details about each item. There is a three-question quiz at the end of the interactive consent, consisting of true/false questions and information reinforcement.
- A longer form “that is a normal consent document,” Block says. After reviewing the consent document, participants can tap “Agree” or “Disagree” to continue. If “Disagree” is selected, the consent process is canceled.
“Something we also took into account when we discussed the consent process is that, since it’s not a treatment study and there is no clinical care in any way, shape, or form, we could not have pulmonologists on call 24/7 for the unknown number of people who would sign up,” Block says. “It’s not a reasonable way to conduct it. We made it very distinctly clear that this [study] was not care or offering care. I think in order to provide clinical care through an app, the world would have to solve the telemedicine problem first.”
The multi-pronged consent approach was also best suited for the mPower and Share the Journey apps, says Wilbanks. “If we were going to do this with a larger sample size of people, we knew we would have a consent process without a clinician talking to every patient,” he says.
Consent includes a tiered hierarchy of information, he says. “There are different layers of information — it’s not acceptable to just click ‘Accept’.”
The first level “is primarily pictorial information to communicate essential concepts for the clinical studies,” Wilbanks says. The second level, Learn More, “has the text document in there so they can follow their nose from the high level of information to the details,” he says. “The third tier is the actual informed consent. We think of the process we designed as a way [for participants] to find as much information as they want. We think it’s appropriate because these studies are very low risk.”
When approving the consent, the IRB took into account that the consent would be read from a distance without investigator interaction, Forster says.
“It’s kind of a confusing regulatory area,” Forster says. “The board has to determine whether there is sufficient electronic consent, or whether a waiver of consent is appropriate for a minimal risk study. Oftentimes what we do with e-consent is waive documentation of consent. For these studies, subjects click through [the consent] and take the test at the end, and that’s sufficient for documentation of consent.”
There was an intensive six-month period to improve the informed consent after initially presenting the mPower study to WIRB, Wilbanks says. “As we built out the apps and worked with Apple, we made the consent process prettier with better graphics, better phrasing, and more accurate details,” he says. “We worked with WIRB over six months with a series of revisions and amendments as we built out the governance and technology to make sure our protocol was consistent with the study we launched — something that brought together the cultures of independent review and software development.”
The developers also took the time to educate the IRB on mobile apps, Wilbanks says. “It was a growth opportunity to teach IRBs about what mobile devices can and can’t do,” he says.
One of the main questions in the development process was the privacy risk, Wilbanks says. The Sage-designed apps are using cloud services on the back end, industry-standard encryption, and there is no data bleed to other apps. “The privacy risk was the top-tier risk,” he says.
“One of the things that was really interesting from an IRB perspective is that we were really careful to describe in a flowchart the kind of information that would be collected, where it would be collected, with whom it would be shared, and to what level of security,” Block says. “If the people creating these apps have to go to the IRB, they have to be incredibly clear about what is going on in the app. This was a lot more detailed than any study I’ve seen.”
Mobile apps such as these are better suited for low-risk observational studies, Wilbanks says. “It might not fulfill obligations on a more complicated scale such as genetic studies,” he says. “We’re going to see quite a bit of kicking of the tires. We’re going to learn and ton and see what developers show up and what apps appear in the app store.”
Block is planning to write a paper on the outcome of the app project and what the IRB community can learn from it. “This is an all new ballgame and I hope they’re all successful and learn things from people — subjects and researchers — who didn’t have access [to studies] before,” Block says.