Health Apps, Bar Codes, and Big Data: The “Brave New World” of Medical Device Innovation
WASHINGTON, DC—The US Food and Drug Administration says it is making headway with plans to wade into the world of Big Data and streamline the path from bench to bedside for innovative new products, while at the same time boosting its ability to swiftly analyze safety signals.
Those were some of the key messages from an FDA Town Hall held here at CRT 2016 in a session dedicated to reviewing the agency’s progress in launching a nationwide medical device surveillance and evaluation program.
William Maisel, MD, deputy director for science and chief scientist at the Center for Devices and Radiological Health (CDRH), gave a “state of the union” talk about strides made in the clinical trial and device evaluation arena. In 2015, Maisel noted, the CDRH managed to reduce the number of days to full approval for new IDE trials to 30, down from 442 days in 2011. In 2015 alone, the agency approved 61 high-risk devices (the highest number since 2001), granted 18 de novo requests, and approved an additional 79 novel devices.
In 2016 and 2017, the CDRH will focus on its National Evaluation System for Medical Devices, a plan that’s been in the works for several years with the aim of tapping into already existing registries and databases for the purposes of verifying the safety and efficacy of devices.
“We’re looking at how to better harness evidence from clinical experience so that we’re not starting from scratch, not running out to do a novel, randomized, blinded, clinical trial in every case,” Maisel said. “We have a lot of patient experience in clinical care in this country, so [the aim is] better harnessing and leveraging that information and better building and using the infrastructure that exists in this country.”
The FDA’s approach also seeks to incorporate consumer-generated data, pointing to the explosion of medical apps and the rich source of data these could provide. A full 85% of US adults owns a mobile phone, roughly half of which are smartphones, he noted, and 1 in 5 smartphone users has a healthcare app on their phone. In this “Brave New World,” Maisel said, patients themselves will be key partners in collecting, connecting, and communicating health data.
Another central push by the FDA is implementation of a unique device identifier (UDI) system. In the future, all FDA-approved medical devices will need to carry a label that can be read by both humans and machines. This label is like “the barcode on your food when you go to the supermarket,” Maisel explained, and includes information that can be incorporated into electronic health information and insurance claims data and provide the infrastructure for different types of analysis.
Codes were added to newly approved Class III devices in September 2014 and implantable life-supporting/life-sustaining devices, like LVADs, in September 2015, Maisel said. By next fall, Class II, “moderate-risk” devices will all need to carry the mandatory coding, and by 2018, Class I devices and unclassified devices will also be labelled.
“Obviously FDA doesn’t control what needs to be incorporated into electronic medical records or insurance claims data, but certainly work is being done in many sectors to leverage this type of information” using the UDIs, Maisel said.
Gregory Pappas, MD, associate director for device evaluation at the CDRH, who gave an overview of some of the specific goals for the National Evaluation System, pointed out that a similar linkage has been in place for drugs using the National Drug Code since 1972 through the Sentinel program.
In his talk, Pappas walked through the “strategic priorities” envisaged for accessing and linking different types of datasets. By the end of 2016, he said, the FDA hopes have gained access to 25 million electronic patient records, a number that will rise to 100 million by the end of 2017. Tied to that, the agency aims to increase by 40% the number of premarket and postmarket regulatory decisions leveraging this kind of real-world evidence, upping that to a 100% increase by the end of 2017.
There are already more than 25 studies underway or being launched looking at device safety and performance that are linking registries to longitudinal data systems that include patient outcomes from medical records and insurance claims, Pappas noted. Cardiologists have taken the lead here, he added, pointing to the STS/ACC TVT Registry, which is providing real-time patient data from hospitals across the United States.
Immediate next steps, said Pappas, will be to push for coordinated networks of registries and also bring existing registries up to “regulatory grade.” Beyond that, the FDA is continuing to work with a planning board, involving multiple stakeholders—including the National Medical Device Registries Task Force, to create the framework that will support the system.
Fixing a Fragmented and Distrustful System
Mitchell Krucoff, MD, of Duke Clinical Research Institute (Durham, NC), co-authored a report from the National Medical Device Registries Task Force published in the Journal of the American Medical Association last year. He outlined the opportunities and challenges for the national system, noting that most cardiologists tend to think of registries primarily as sources for “retrospective muckraking” or as databases for safety surveillance.
The proposed plan goes far beyond this, he said, and will have the same quality and opportunities of randomized clinical trials, but with the opportunity for far richer and complete follow-up.
The biggest barrier, however, will be convincing different stakeholders of the benefits of data-sharing. “What it will really take to make this work is more than just money and more than just time,” he said. “It will require a real transformation from what has traditionally been a fragmented and distrustful ecosystem—which we call medical device innovation—where stakeholders have largely thought about themselves and competitors have largely thought about how to do things without sharing anything with anybody else.”
This will take a paradigm shift, he continued. “Frankly, you can get US Congress to throw all the money in the world at this and it won’t work, unless we can . . . recognize that the vast majority of barriers to this kind of efficiency and quality are precompetitive barriers. They have nothing to do with individual device manufacturing success or any one stakeholder and everything to do with problems we all wrestle with all the time in collecting accurate reliable information—maintaining privacy while having the ability to collect what we’re doing in clinical practice as a nation on a routine basis, to understand if devices are doing well or having problems.”
Launching a National Medical Device Evaluation System, FDA Town Hall, Technology & Innovation [session]. Presented at: CRT; February 23, 2016; Washington, DC.
- The Gift of Data: Medical Journals Back Public Data Disclosure as Prerequisite for Publishing
- Experts Say ‘Big Data’ Represents Future of Health Care Research
- Keynote: Simplified Trials, Community Engagement Key to Reinvention of Clinical Research