Stakeholders appear to support using electronic health record data in clinical trials, but don’t think EHRs themselves are quite ready for prime time, according to comments on the Food and Drug Administration’s draft guidance on the topic.
The FDA had released draft guidance regarding the collection and use of EHR data in clinical trials on May 17. In response, Health IT Now (HITN) noted in its comments that it is in favor of including real-world evidence in clinical trials, but “more work needs to be done.” For instance, the lack of interoperability is a barrier to sourcing the data.
The draft guidance itself also needs clarity, HIT Now says. It doesn’t include data from registries, wearables and other forms of health IT. The organization also expressed concern about how the FDA would handle informed consent; the draft guidance was not clear if the FDA was referring to areas where informed consent would otherwise be required or if additional consent requirements were being added, which HIT Now does not support.
The American Medical Informatics Association (AMIA) took an even stronger stance that using EHRs for clinical trials was premature. While it “fully supports” FDA’s willingness to consider using EHR data, “we strongly caution FDA from assuming that most EHRs are readily configurable for clinical investigations, even among more advanced institutions.” AMIA specifically pointed to the fact that EHR data is intended to support patient care, and may not be of research quality, due in part to the lack of standardization. It also doubted that current EHRs can provide some of the functions that the FDA is looking for, such as audit trails and data modifications.
AMIA additionally noted that the FDA may be overly confident about EHR certification. “We perceive an overreliance on the assurances resulting from ONC’s Health IT Certification Program related to data reliability and interoperability ... ONC’s Health IT Certification Program is not designed to ensure data ‘reliability’ or integrity,” the organization warned.