Skip to main content

Apple's ResearchKit generates reliable health data — at least for asthma patients

Apple's ResearchKit generates reliable health data — at least for asthma patients


Good news, since there are so many smartphones

Share this story

Apple ResearchKit photos

Health data collected entirely from smartphones can be reliable, research from Mount Sinai Hospital claims. The researchers involved found that Apple’s ResearchKit platform and an app for asthma were fairly accurate when compared to existing patient studies.

Finding and recruiting participants is a big hurdle for medical studies. In recent years, people have started collecting health data from smartphones, which seems sensible given how common smartphones are. But this raises questions about whether data gathered this way can be trusted. Today’s study, published today in the journal Nature Biotechnology, suggests that health care apps may be reliable, at least in regards to asthma. This is good news since smartphone usage is only increasing — there are supposed to be 6 billion smartphones used worldwide by 2020 — and collecting reliable health data from them could be very good for research.

Apple launched ResearchKit, a software medical platform, in 2015. It helps researchers recruit participants for studies; participants can enroll in trials and take surveys or provide other data. Early research partners included big names like the University of Oxford, Stanford Medicine, and the Dana-Farber Cancer Institute. The asthma mobile app from today’s study was one of the five disease-specific apps that Apple launched with the initial release of ResearchKit.

In today’s study, nearly 50,000 iPhone users downloaded the asthma app. Of these, about 7,600 people enrolled in the six-month study after completing the consent form. People in the study took surveys on how they treated their asthma; the app also provided information about location and air quality. The scientists then looked at how this patient-reported data measured up when compared to external factors. For example, around the time there were fires in Washington state, patients in the area reported worse asthma symptoms, as might be expected, suggesting the data was fairly reliable. They were similarly able to correlate data related to heat and pollen.

Of course smartphone-based studies have limitations, too. The researchers say that this way of gathering data is good for studies that only last for a short amount of time and that need to quickly enroll participants across the country.

And they acknowledge that there are several weaknesses. First is low retention rate. In this study, about 85 percent of people completed at least one survey, while about 30 percent completed more than one over six months. The lack of sustained interest meant incomplete information.

That’s not the only problem. It’s possible that the kind of people who voluntarily download these apps might not be representative of the general population, in part because many people have Android phones rather than Apple ones. That makes the data limited in scope. And because people were filling out their own surveys — a notoriously unreliable source of information — there may be problems with the data they generate as well. Others have pointed out that ResearchKit can run into ethical issues when it comes to data security and making sure people understand all of the risks of volunteering their information.

Even with those constraints, though, researchers are likely to find today’s study reassuring. If self-reports match closely with the local environment, that gives scientists using ResearchKit more confidence their data is good — within the platform’s limitations, of course.