Late last year, the FDA released a draft guidance on electronic source, or "e-source." It's not as exciting as rebelling Libyans or Charlie Sheen's rants, to be sure. But the ramifications are far-reaching, and could impact bread-and-butter monitoring work that is the lifeblood of major contract research organizations (CRO).
The agency's official goal is modest. It is merely seeking to augment its previous recommendations about software and clinical research. Specifically, it has been thinking about electronic data and documents in clinical trials. In brief, the FDA is attempting to guide the pharmaceutical industry and the medical research community toward something that might resemble modernity.
Modernity will not come easily. Many veteran clinical trial professionals belong to pagan support groups that worship paper—small, piled shrines of copier paper, according to most reports. That attachment to paper is a lamentable byproduct of all the problems that electronic systems for clinical research have caused.
Because any new FDA recommendations must coexist perfectly with all of its previous laws, regulations and unwritten practices, there are a few nuances to the FDA's perspective on e-source.
Chief among these: FDA is envisioning a single system that it calls an electronic case report form (eCRF). That linguistic choice (using "form" not "system") is confusing. It's unfortunate terminology.
In functional terms, though, the FDA's eCRF is a very sound idea. The FDA-approved eCRF would vacuum up every shred of clinical data and deposit it into a massive repository with an audit trail.
The only trouble is, many research companies use a patchwork of systems. Few pieces of software connect to their brother and sister applications from the same company, much less other vendors. Most clinical trials are dependent on an ever-changing mix of homegrown software, Oracle databases, personal spreadsheets, email messages, public calendars and properly maintained commercial systems. Combining all that into one federal "eCRF" will be no small job. Big vendors will have a major advantage over small ones.
The FDA, we hasten to add, is quite enthusiastic about the feasibility of its vision. This week the agency took the unusual step of holding a 90-minute web seminar to discuss the new guidance. The Drug Information Association (DIA) sponsored the online FDA presentation.
Leonard Sacks, for one, told the electronic attendees that using computers would be healthy for all participants in clinical research. Acting director of the Critical Path research agenda at FDA, Sacks sounded as glowing about the marvels of computers as any electronic data capture (EDC) vendor circa 2003.
"We think it will make trials easier for sponsors," Sacks said. "We think it will reduce costs. We think it will make the performance of clinical trials more efficient and quick. We're hoping that the rest of the medical community will join us in shepherding this ahead. We're looking forward to a new realm of clinical trials."
Sacks spearheaded the writing of the new guidance document, and said that its genesis dates back a few years. "We at FDA felt it was important to promote the usage of electronic media," he said, citing the prevalence of electronic systems in other industries. "There are obviously numerous burdens of paper-based studies. I think everyone is aware of this."
It's impossible to say whether an FDA guidance constitutes anything close to the agency's official policy direction. But if it is a harbinger, the FDA is clearly comfortable with central, computer-based monitoring that has nothing to do with paper source documents.
Granted, the focus of the webinar was not specifically on site monitoring. But Sacks and his colleagues made it clear that a well designed computer system could give sponsors and the FDA access to information in ways that equal or surpass a traditional eyeballing of actual paper forms with coffee stains. "Electronic platforms for clinical investigations have enormous potential advantages," Sacks said.
The pace of the webinar quickened considerably when another FDA official began taking questions from the virtual attendees. Was the FDA imagining the federally defined eCRF as a single system integrating multiple feeds of data? "To a large extent, yes," Sacks said.
But as he had outlined in detail, there are several large, important categories of data from diagnostic and radiology equipment, as well as other sources, that might not reside in the agency's eCRF, but which would assist in an inspection of a site all the same.
There was also considerable uncertainty around whether a new, improved eCRF would be a so-called source document—a primordial, sacred original. "An eCRF may serve as an electronic source document if specified by the protocol," said Jonathan Helfgott, another e-source expert at FDA, where he works in the Office of Scientific Investigations. He and other presenters stressed the value of having a data management plan and determining how e-source data will be handled. That plan can be part of the trial protocol, or a separate document.
Helfgott said that with transcription errors comprising the bulk of changes to CRFs, it is fine and appropriate for designers of EDC systems to have a pull-down menu indicating "transcription error" as the reason for a change to a document.
Helfgott also approves of empty text boxes for additional notes from clinical staff. Such text boxes, it turns out, are a feature that some EDC system providers advise against, simply because they can become stenographic garbage disposals in which unexpected observations may get stuck.
Another looming issue for the FDA, of course, is the degree to which the industry will be leaning on electronic health records (EHR). (Until 2014, the purchase of physician and hospital computer systems will be heavily subsidized by taxpayers; Congress believes doctors cannot afford to buy their own computers.) In the past, the vast majority of physicians only used computers to get paid. Now many are using them instead of paper medical charts.
On the webinar, the FDA said that it was trying to prepare for a day in which such EHR systems were more important in industry-sponsored research. (It sounded as if the agency might be working on guidance for EHRs, but that was never stated explicitly.) Helfgott did say: "We have seen [EHRs] used in a compliant fashion. As more processes get refined, we will respond in a timely fashion."
When asked about the industry's main misconception about e-source, Helfgott had an equally succinct answer: "That FDA needs a paper record on site to maintain compliance." So for all you paperophiles out there, be advised that the FDA is on record that a computer can do just as good a job.
The devil, of course, is in the fine print. And one complication there lies in the historically divisive question about where the electronic source documents should be stored after a trial finishes. Originals? Copies? Certified copies? It can get muddy.
Yes, most industry professionals know that clinical investigators have hallowed scientific authority and legal rights to control the data about their own patients. In practice, however, trial doctors are also contract workers. They may or may not receive a CD of their data at the end of a project, just because sponsors are such large and bureaucratic organizations.
Mathew Thomas, of the FDA's Office of the Commissioner, explained that one recent audit identified a sponsor that was diverging from the letter of the law. The site in question had no copy of its data, having previously shipped it to the sponsor. "This is not an acceptable practice," Thomas said. "The data collected at the site should be maintained by the clinical investigator."
No one will accuse the FDA of rushing into the e-source landscape. Early versions of EDC were tested two or even three decades ago. The current top five systems have all been in existence for a decade or more. Even five years ago, the use of an EDC system was a default or noncontroversial choice at most large sponsors. So there is a large body of practice around e-source and EDC. Whether any of that is going to collide with the FDA's current vision remains to be seen.
On the webinar, the FDA declined to predict how long it would take to issue a final guidance. By law, the agency must accept comments and questions from the industry and random members of the public. That docket can be found on regulations.gov and will be open until early April, 2011. The question now is how the industry can respond to the FDA guidance in a manner that moves the e-source discussion forward, and does so without enshrining Dickensian, paper-based inefficiencies in the computer-driven processes of the 21st century.
One of the biggest changes would be a clear declaration by the FDA that it is not necessary to send thousands of human beings to look at millions of pieces of paper at clinical sites when good eCRF systems are in place. Such a declaration, however unlikely, would be initially disruptive to elements of the research community that have major revenue centers built around paper and inefficiency.
But remote access to electronic clinical trial data could help the FDA and sponsors. It could yield massive long-term savings and operational advantages for sponsors. Information technology on its own, for many reasons, has not generated many financial or time savings. So it would be ironic and welcome if the FDA could finally produce those elusive byproducts of modernity by using its own word-processing software.
Here's an earlier ClinPage article on the guidance.d9A2t49mkex