FDA, e-Source and Clinical Data – Is the Needle Moving?

On June 26, 2015, FDA CDER posted a notice in the Federal Register entitled “Source Data Capture from Electronic Health Records: Using Standardized Clinical Research Data.”  I thought it might be worth a quick look at the responses posted to the associated Docket by the August 10 deadline to see what came of it.

The notice indicated an interest in “supporting demonstration projects to test the capability and evaluate performance of using an end-to-end Electronic Health Record (EHR)-to-Electronic Data Capture (EDC) single-point data capture approach, using established data and implementation standards in a regulated clinical research environment,” soliciting “written requests for participation in the demonstration project.” It listed some potential benefits of this approach, and posed six specific questions. There were no details on what support was available, what number or types of demonstration projects were expected, or how the response should be organized, inviting “comments, proposed approaches, interest to participate and responses to the question.” So it’s not surprising that there was a great deal of variability in the level and types of responses ranging from a few brief questions to comprehensive project proposals:

• 4 provided comprehensive project proposals including ability to identify an actual study protocol.

• ~13 offered potential solutions to run a demo study or a Proof of Concept (POC) consistent with the notice but no indication of using a live study.

• ~17 promoted a specific solution not entirely consistent with the Notice, or just offered some general help.  Some of these were intriguing (POCs for recruitment, protocol feasibility, devices or drug safety) but not really using EHR to EDC.

• ~8 seemed to be proposing specific solutions that appeared inconsistent with the notice.

• 5 offered comments or questions only.

The type of respondent was also rather diverse (some including more than one category):

• 18 technology vendors or CROs

• 10 healthcare providers

• 6 pharmaceutical manufacturers

• 5 academic or professional organizations

• 2 standards development organizations.

In terms of standards proposed (by my count):

• 14 referenced the emerging HL7 FHIR standards

• 9 - the IHE/CDISC profiles

• 4 - other CDISC standards (ODM, SDTM, BRIDG) but not the profiles

• 2 - the Structured Data Capture (SDC) initiative of the US Office of the National Coordinator

• Several mentioned clinical vocabularies.

• 15 did not involve standards, just mentioned some or commented on what was wrong with those. 

• 11 were based on what appear to be proprietary products.

So what can we make of this so far?

• Interest in exploring this area seems to be finally starting to build in the research community, even though there was clearly confusion on exactly what was being requested. This interest might be attributed to evidence of FDA interest in supporting it.

• It was interesting that only 6 manufacturers responded, compared to 10+ healthcare providers.

• There was a surprising degree of interest in the emerging HL7 FHIR standard in addition to the IHE/CDISC profiles that have been available for some time.

• There seems to be a lot of different interpretations of what it means to use standards.

• The CMS Meaningful Use Final Rule was released after this notice, but offers an exciting opportunity to explore the use of FHIR-based APIs in the USA to directly access EHR data in the future for research purposes.

So CDER has thrown down the gauntlet to industry on eSource, and now has their work cut out for them on next steps. Here’s hoping a variety of standards-based solutions (both CDISC and HL7) will be tested, and that sufficient metrics will be gathered to help determine whether the state of the art now supports doing this in a cost-effective manner that minimizes risk and has a measurable return on investment. Let’s hope that similar objective data will soon be available from another sponsor-led project testing CDISC-IHE standard that is also getting underway.

And while it will be invaluable to have actual data on this EHR to EDC approach, seems we should also be exploring other use cases such as feasibility, recruitment, and drug safety.  Meanwhile, for those of us who are reimagining research, shouldn’t we also be looking at ways that we can access and use healthcare data without having to go through an eCRF in the middle? Now that’s a potential game changer worth exploring in the future.

Wayne Kubick is an expert consultant and advisor on clinical research through technology and data standards. He is also an Editorial Advisory Board member for Applied Clinical Trials. He can be contacted at wayne@waynekubick.com. Follow him on Twitter @WayneKubick.