OR WAIT 15 SECS
Paul W. Glimcher holds the Julius Silver, Rosyln S. Silver and Enid Silver Winslow Chair of Neural Science at New York University (NYU) and is Professor of Neuroscience and Physiology at NYU’s Langone School of Medicine. He is also Professor of Economics and Psychology at NYU. He is also the chief Executive Officer of Data Cubed LLC, a pioneering healthcare technology company that collects accurate real-world data and electronic Clinical Outcome Assessments eCOA solutions tailored to everything from remote patient engagement to virtual clinical trials. By tapping into what makes humans human, Data Cubed is changing how healthcare approaches big data to create a healthier world. Find out more by visiting www.datacubed.com.
FDA's proposal to build a modern system to gather real-world evidence (RWE) from about 10 million individuals could have profound implications, but it will realize its potential only if FDA takes a more expansive view of what RWE can be, writes Paul Glimcher.
In a sign of the increasing role real-world evidence (RWE) will have in healthcare research, FDA recently announced a $100 million proposal to build a modern system to gather RWE from about 10 million individuals. The good news is that this could transform clinical trials for a generation; the bad (or at least worrisome) news is that as currently planned this dramatic initiative might wind up being largely limited to data from electronic health records.
The FDA announcement comes at a time of growing consensus that the future of clinical trials will depend heavily upon the gathering and analysis of RWE. What exactly is RWE? By definition, RWE in healthcare consists of information about patient health status, outcomes and care delivery from “real-world sources”. That data could come from just about anywhere: from data gathered in a physician’s office, to automated recording of the exact time at which a patient takes medication in her home, to the number of steps a person takes each day, to the quality of air in a person’s home, to the amount of tobacco smoke to which they are exposed. With the rapidly expanding technological base being developed today, the possibilities are just about limitless.
The critical idea for many researchers is that RWE analyses can go beyond data entered by physicians and technologists into an electronic medical record and can record health-related events happening in the real-world outside the hospital or doctor’s office. These new data could be used to generate truly novel insights about how a drug is actually used and how it interacts with the observable health-related behaviors of study participants. RWE from outside the traditional medical environment will undoubtedly let us explore new research questions, complement and refine the findings of traditional clinical trials, and fill knowledge gaps which prevent us from seeing why a medication works for some patients and not for others.
For all these reasons the FDA proposal is important and could have profound implications, but it will realize its potential only if FDA takes a more expansive view of what RWE can be – genuinely moving beyond data from billing codes and medical records.
So what kinds of data should the FDA consider including in the RWE initiative?
Validated Consumer Medical Device Data: The low-hanging fruit in RWE would be the use of external data sensors of the kind consumers already buy. Consider the many scales in the marketplace that gather heart rate data each day. Or the newer devices that can even gather electrocardiograms. From devices measuring blood-pressure to devices quantifying indoor air quality, if FDA were to develop guidelines for incorporating these kinds of existing measures in to studies this could provide a simply and impactful way for big pharma to new class of readily available data.
Proprietary Consumer Devices: A bit more challenging – but totally doable-would be gathering data from consumer devices that have not yet been standardized or validated. RWE from accelerometers can be used to measure fitness and activity levels, is one example of this class of data. Of course, the challenge with proprietary data sources is that each brand of device being sold today uses different proprietary algorithms. As a result, a person who walks a mile wearing two different brands of accelerometers will find the two devices report completely different number of steps taken. What needs to be done to use this kind of data is for FDA to help industry to develop data standards that can them be made available to device manufacturers. It is not that hard to imagine a company like Fitbit or Garmin offering a version of its devices that implement a standardized medical algorithm. This would be a boon to researchers and players in the pharmaceutical industry, players that have already expressed an interest in such a standard.
Novel Data Classes: Probably the most mature of the new data classes is geotrack data. Using smartphones to compute how much time a person spends walking each day, and at one speed, is surprisingly easy. Measures that approximate how much time a person spends smoking or stationary are also possible. Even measuring how much time a month a subject spends at health care providers of any kind is easy to do. Better yet, these kinds of measures can (at a technical level) be made without the study sponsor ever having access to the private geotracks of the participants. What we would do with this kind of data is obviously an open question but knowing how much time a subject spends at health care facilities is surely real-world evidence we want and need.
FDA deserves praise for its decision to embrace the potential of RWE. Many of us in the academic and corporate mobile health space hope that FDA’s definition of RWE is as broad and exciting as the field that they hope to invigorate.
Paul W. Glimcher holds the Julius Silver, Rosyln S. Silver and Enid Silver Winslow Chair of Neural Science at New York University and is Professor of Neuroscience and Physiology at NYU’s Langone School of Medicine. He is also Professor of Economics and Psychology at NYU. He is also Chief Executive Officer of Data Cubed LLC.