Eric J. Topol, MD

Disclosures March 18, 2015

An Easier and Cost-Effective Medical Research Model?

While most Apple aficionados were looking for more information about the company's ballyhooed new watch last week, an announcement of a new platform for medical research was unveiled by surprise. By shifting data collection to consumers in their real world, on a massive scale, this potentially represents the biggest change since the origins of clinical trials—such as with citrus fruits for scurvy in 1747 or streptomycin for tuberculosis in 1946.

Until now, nearly all medical research has been conducted in clinics and hospitals. It has required a willing individual—typically called a "subject"—to provide written informed consent after a nurse or doctor reviewed the research project details. A form, often comprising several pages, had to be signed to acknowledge the risks and potential benefits.

The measurements taken, including vital signs and labs, represented one-off data points in the unnatural, contrived environment of a doctor's office or hospital setting. This is a laborious process that takes considerable time and dedicated personnel, resulting in the impaired ability to recruit individuals to participate in research projects. A major sequela has been that clinical trials are remarkably expensive and painstakingly slow. For example, when my colleagues and I performed a large heart attack trial that enrolled over 41,000 patients in the early 1990s, it took nearly 3 years and cost more than $50 million.[1]Today such a trial would probably take even longer and cost at least four to five times as much.

Enter a mobile medical research platform such as Apple's ResearchKit, with the ability to recruit patients from the public via smartphones. Some have called this a "clinical trial in your pocket." Not only does this immediately access millions of individuals (iPhones were purchased by more than 70 million individuals in the fourth quarter of 2014), but the consent process is very similar to downloading and accepting an app.

Moreover, instead of the data being entered by a nurse or research coordinator, the patient enters his or her own information or generates the data via a sensor or app. And rather than the artificial measurement obtained in a medical center, the context is the individual's real world—whether it be at work, home, or on the go. Because smartphone users look at their phones an average of 100-150 times per day, we move from one-off data capture to the potential of "all in" engagement. For example, the University of Rochester app for Parkinson disease[2] tracks tremors, voice, and gait. Participating patients would conceivably be able to have previously unobtainable information, now in real-time, about the effectiveness of their treatment throughout the day with these multiple metrics, all assayed via their smartphone.


1.Medscape, A New Mobile Model of Medical Research, Medscape: Medscape Access