Healthcare Analytics, Population Health Management, Healthcare Big Data

Precision Medicine News

FDA: Real-World Data, Machine Learning Critical for Clinical Trials

Real-world data, machine learning, and artificial intelligence will help to support effective clinical trials, says the FDA's Scott Gottlieb.

Clinical trials and machine learning

Source: FDA

By Jennifer Bresnick

- Real-world data gathered directly from EHRs and other data sources, paired with advances in machine learning, will be crucial for architecting the next generation of successful clinical trials, says FDA Commissioner Dr. Scott Gottlieb.

In a speech to the Bipartisan Policy Center on January 28, Gottlieb stressed the importance of modernizing the clinical trial process to take advantage of EHR data, Internet of Things (IoT) devices, claims data, and other non-traditional sources to support scientific investigations.

“Digital technologies are one of the most promising tools we have for making health care more efficient and more patient-focused,” Gottlieb said.  “This isn’t an indictment of the randomized controlled trial. Far from it.”

“It’s a recognition that new approaches and new technologies can help expand the sources of evidence that we can use to make more reliable treatment decisions. And it’s a recognition that this evidence base can continue to build and improve throughout the therapeutic life of an FDA approved drug or medical device.”

“Hybrid” clinical trials that combine traditional research approaches with real-world evidence (RWE) can support agile discovery and greater efficiency, speeding up the process of bringing new therapies and devices to market, he continued.

READ MORE: FDA Tackles Artificial Intelligence with New Software Review Plan

RWE is important for ensuring that clinical trials capture accurate and comprehensive data about diverse participants.

“We believe that more accessible clinical trials can facilitate participation by more diverse patient populations within diverse community settings where patient care is delivered, and in the process can generate information that’s more representative of the real world and may help providers and patients make more informed treatment decisions,” he said.

“This approach, called decentralized clinical trials, can help move prospective collection of data from the real-world—including randomization—outside of the brick and mortar boundaries of traditional clinical research facilities, tapping into not only EHRs but additional digital health tools like wearable devices.”

Some of these strategies are already being deployed in the research and clinical settings.

“Providers and other stakeholders are already exploring effective ways to leverage electronic tools to gather vast amounts of health-related data from EHRs and other sources,” Gottlieb noted. “And they’re working on ways to use advanced analytics, including machine learning algorithms, to transform data into evidence that can be used to help guide clinical decision-making or inform innovators during the development of medical products.”

READ MORE: FDA Sets Goals for Big Data, Clinical Trials, Artificial Intelligence

“At FDA, we’re committed to advancing ways that data leveraged from these streams – typically called real-world data (RWD) – is transformed into evidence using transparent data standards that can give all stakeholders confidence in the data’s provenance, so that more of this data can be used to generate the evidence that the agency needs to improve regulatory decision making.  Advancing RWD into regulatory-quality real world evidence is a key strategic priority for the FDA.”

The FDA is incorporating real-world evidence into the post-market safety and surveillance process through its Sentinel system, explained Gottlieb. 

Use of RWD gathered through Sentinel has already eliminated the need for post-marketing studies on nine safety issues across five different products, he said. 

The agency is planning to continue to explore the role of RWD in reducing costs and improving efficiencies as laid out in the FDA’s Framework for Real-World Evidence Program, released in December of 2018.

“Traditional post-market studies typically require years to design and complete and cost millions of dollars. By encouraging the use of RWD and RWE, we may be able to provide patients and providers with important answers much sooner by potentially identifying a broader range of safety signals more quickly,” he said.

READ MORE: FDA Seeks Patient Engagement for Clinical Trials, Precision Medicine

“The use of Sentinel and similar tools will increasingly let us shift some studies and data collection to the point of care, making collection of data and development of actionable evidence more efficient.”

The research industry will need to work closely with providers, patients, and manufacturers to develop standards and methodologies for incorporating RWE and RWD into clinical trials and regulatory oversight.

Industry standards and a collaborative approach will also be key for integrating machine learning and artificial intelligence into the innovation process, he continued.

“As the volume, velocity, and variety of real world data reaching the agency increases, we have an opportunity to use new software-based machine learning algorithms – like natural language processing or deep learning – to help develop regulatory science tools like surrogate endpoints or digital biomarkers that can be used to guide more efficient development programs,” Gottlieb stated.

The FDA’s Information Exchange and Data Transformation (INFORMED) initiative, a collaboration with the HHS IDEA Lab, will be working with additional internal and academic partners to develop an FDA curriculum on machine learning and AI, Gottlieb announced.

“The aim of this program is to improve the ability of FDA reviewers and managers to evaluate products that incorporate advanced algorithms and facilitate the FDA’s capacity to develop novel regulatory science tools harnessing these approaches,” he said.

“The agency will also pilot a competitive fellowship program under INFORMED in Artificial Intelligence and Machine Learning that allows post-doctoral fellows from leading academic centers to join the FDA for two-year fellowships to develop high-impact AI-based regulatory science tools by working closely with mentors from the agency’s medical product centers.”

Gottlieb also unveiled several other initiatives, including a collaboration between the FDA’s Oncology Center of Excellence (OCE), Friends of Cancer Research, the National Cancer Institute, and other stakeholders to create standards for using tumor data to understand patient responses to immunotherapy.

“Harmonizing the measurement of tumor mutational burden (TMB) across commercial assays used in routine oncology care can help reduce treatment variability and improve the utility of TMB as a potential biomarker for enriching clinical trials testing immunotherapies,” Gottlieb said.

Collaboration across public and private stakeholders will an ongoing part of the FDA’s strategy to modernize, streamline, and fine-tune the innovation process, the Commissioner said.

“By engaging with multiple stakeholders through collaborative forums and working closely with our agency partners at CMS and the National Institutes of Health, the agency can help promote more transparent standards for curating data, interoperability, and RWE generation that can help ensure that every American patient – no matter where they live - benefits from the full potential of these technologies to make our healthcare system safer, smarter, and move patient focused,” Gottlieb concluded.


Join 25,000 of your peers

Register for free to get access to all our articles, webcasts, white papers and exclusive interviews.

Our privacy policy

no, thanks

Continue to site...