The Yale School of Medicine and Mayo Clinic are experimenting with a personal health record platform that allows patients to gather real-world data to share with researchers for post-market surveillance of medical devices.
Speaking to the NIH Collaboratory Aug. 10, Sanket Dhruva, M.D., an assistant professor of clinical medicine at the UCSF School of Medicine and a cardiologist at the San Francisco VA Medical Center, described some of the limitations of the current mechanisms used by the U.S. Food and Drug Administration for medical device surveillance and the potential for real-world evidence to help clinicians better understand the safety and effectiveness of devices.
Currently the FDA uses passive surveillance such as reporting of adverse events to the Manufacturer and User Facility Device Experience (MAUDE) database and does studies when safety concerns need to be investigated. Researchers at the Yale University-Mayo Clinic Center of Excellence in Regulatory Science and Innovation (CERSI) are seeking to complement that approach with real world data. They are conducting a demonstration projects for NEST, the National Evaluation System for Health Technology, which was established by the Medical Device Innovation Consortium and funded by FDA.
Dhruva said the best data sources would be prospectively planned and offer continuously updated longitudinal follow-up for a comprehensive set of clinically relevant outcomes. “Ideally it would include patient-reported outcome measures as well as patient-generated data and would integrate seamlessly with existing data systems,” he said.
Both claims and EHR data sources are commonly used for device surveillance. But they have limitations of their own, he said. Claims data are ubiquitous but not collected for supporting research. They are complete only if people remain in the same health system, so there tend to be gaps in data. Also they have time lags in availability. “They cannot identify use of a specific medical device,” Dhruva added.
EHR data offers very rich clinical information, but EHRs also are not designed to support research. The data is complete only if the patient remains in the same health system, so there may be many important gaps in data. “EHR data rarely includes patient-reported outcomes in a structured format,” Dhruva said. “Parts of patient-reported outcomes are in clinician notes or scanned documents,” he added. EHRs are rarely able to identify the use of a specific medical device.
In 2012, the FDA issued a rule requiring medical devices and packaging to have Unique Device Identifiers: However, there has been limited benefit because the UDI is unavailable in administrative claims data and EHR data, Dhruva said. “The UDI has a lot of potential to track devices and outcomes, but while we push for it to be better integrated in EHR and claims data, we can work on getting more real world data for medical device surveillance.”
The pilot project allows patients to provide their own outcomes (through short questionnaires and through synchronizing data from mobile health trackers) after they have received a procedure that utilizes medical devices. The Hugo personal health record aggregates data from four sources: EHRs, pharmacy portals, wearable devices and patient answers to e-mailed questionnaires that become patient-reported outcomes.
(Hugo’s founder is Harlan Krumholz, M.D., a professor of medicine and epidemiology and public health at Yale School of Medicine.)
The study is following 60 patients for eight weeks after they undergo two procedures that use medical devices: the multiple devices (including sutures and stapler) used to perform bariatric surgeries (either sleeve gastrectomy or gastric bypass) in patients seeking weight loss and an ablation catheter when used in patients with atrial fibrillation seeking a return to sinus rhythm.
One goal is to test if the patients’ EHR data from multiple health systems can be synchronized into a research-ready database. (The Hugo PHR currently works with 600 health system portals. If patients have data with other providers’ portals, they can download Continuity of Care Documents and then upload them to Hugo, with research assistant help. Patients are being provided with syncable devices such as Fitbits to provide additional insights into their health and health outcomes. The study is also testing the feasibility of obtaining medication data from pharmacies. Patients who use Walgreen’s or CVS can log into their portals and download data.
Overall, the study is trying to answer the question of whether such a patient-powered mHealth platform enable more robust and thorough post-marketing surveillance.
“Getting comprehensive longitudinal data “can only be obtained if the patients link and upload data from different health systems into the application,” Dhruva said. “We hope this is going to become easier through implementation of FHIR and Blue Button 2.0, but it is important that patients link or upload data from all the health systems from which they receive care.” He said combining that with patient-generated data should give clinicians a better understand of how they are doing in the post-procedure setting.