In theory, anyone who abhors paper in clinical trials should love electronic source technologies, or e-source. That’s because e-source (the direct entry of data into a keyboard or Palm with no paper form) is a silver railroad spike through the Dracula heart of inefficiency. What’s not to like?
In practice, taking on paperophilic vampires is messy.
That was clear from a conversation with Assero Ltd.’s David Iberson-Hurst, a consultant and clinical toolmaker. He’s created a program for testing compliance with the Clinical Data Interchange Standards Consortium’s (CDISC) operational data model. Iberson-Hurst is a cheerful fellow – not the sort of chap to complain about bullets, arrows, or teeth now lodged in his thick British hide.
Here’s the back story.
In November of 2004, Iberson-Hurst and CDISC President Becky Kush found themselves making their umpteenth presentation of clinical data standards to the FDA. The FDA welcomed a document on patient-reported outcomes. Iberson-Hurst gamely volunteered to coordinate that document’s assembly by a corps of savvy CDISC volunteers.
As an example of the thicket into which they plunged, Iberson-Hurst cites a simple question. “What do we mean by ‘an investigator has to maintain source documents’? ” he asks. “The agency and the vendors have different feelings about what ‘maintain’ means. Some will interpret it loosely and some will interpret it strictly, depending on where they sit.”
Investigators, needless to say, keep paper source documents.
But e-source data is stored centrally, on a server arranged by a vendor. “That’s inconsistent with the regulations the way they are currently framed,” says Iberson-Hurst. “Vendors are being paid by the sponsor, but everything they’re doing is the responsibility of the investigator. The vendor is very much doing investigator responsibilities.” He speculates that vendors would never seek formal designation as investigators, which would subject their offices to FDA inspection.
The world of the electronic patient diary has leap-frogged beyond what was envisioned by Congress. The law covers a simple landscape of sponsors, contract research organizations and clinical investigators. Under the law, the onus of ensuring the accuracy of the data is on the investigator.
Today, however, a patient with a handheld computer like a Palm sends his most intimate biomedical experiences straight to a server in Newark or Helsinki. Only later will an investigator or clinical research associate glance at that data. It tortures standard English to say an investigator “maintains” such records.
For the vendors in the e-diary trade, this is a red herring: they don’t let sponsors touch the data, even if it does live far from an investigator’s clinic. Let us not forget how ambiguous and cumbersome paper diaries are. Electronic diaries eliminate the need for three well-compensated people to convene to interpret the jagged cursive of an 80-year-old nursing home resident.
Still, for Iberson-Hurst, cutting the investigator out of the record-maintenance loop creates other issues. In his view, there is nothing inherently inadequate about the current regulations. There are, however, regulatory tweaks that need to be sorted out. His CDISC position paper on patient-reported outcomes and e-source, now in its third iteration, might be the nucleus of an FDA guidance document to emerge in another year or two.
Murky Status of Tech Vendors Under FDA Regulations
In general, Iberson-Hurst likes technology. He’s just nervous that the law and the practices of the industry are not in alignment. Gaps might be exploited not by leading diary vendors (which have labored to solidify their good reputations) but by newcomers or fly-by-night operators intent upon mischief or profit.
Iberson-Hurst declined a plea to let his imagination wander.
But it’s easy to conjecture that unethical diary vendors holding pivotal diary data might be tempted to copy or change certain files at a sponsor’s request. (Sponsors routinely make such requests of electronic data capture vendors, in circumstances both innocent and suspicious.) Without paper records, tampering would be as hard to catch as election irregularities that arise at paperless voting booths. For lack of a better phrase, call this the Ohio factor, after the Midwestern state with minimal oversight of its polling places.
Iberson-Hurst can be coaxed to say that the stakes are high. “This is an industry where you’re not going to kill somebody today,” he says. “But you could potentially be doing harm somewhere down the line.”
Iberson-Hurst was kind enough to send along a few PowerPoint slides from the FDA, presented at a conference in April 2005. Among other things, the FDA appears to want the ability to ensure that source data cannot be modified without the knowledge or approval of the investigator.
That is key. Adding the investigator back into the record-maintenance loop for changes to patient diary data is something that might force changes to how selected diary companies operate. We’re just guessing. But it may account for why Iberson-Hurst (on behalf of many CDISC collaborators) has a few flesh wounds.
Iberson-Hurst is hardly pointing fingers. He doesn’t think any major diary vendor would cut corners. He knows such companies believe that their data is more secure than data in diaries using paper. Still, he thinks the rules should be unambiguous: “It’s mapping what they’re doing to the regulations in a clear way such that everybody’s obligations are very clear.”
He’s quick to add that technology alone is neither the solution nor the problem. Process, process, process. As many readers keep trying to teach us, it can’t be overemphasized. As an example, he imagines one clinical site that used a paper source document and another that used only e-source. One approach might be acceptable to the FDA. Another might be risky. Iberson-Hurst says the focus should be on “process plus technology.”
A Proposed 'Black Box' to Protect Diary and Other Data
Standards - the whole reason CDISC originally got involved – could be part of a solution. Iberson-Hurst notes that a data repository in a CDISC format might be a convenient way to archive the data over time.
A standards-based archive would be vendor-neutral. Read-write access to the data could be easily controlled. Sites and regulatory agencies would not be tied to complex or expensive tools. Even better, standards-based archiving might allow an investigator to verify that John Doe was his patient and that his asthma attacks (recorded in a Palm handheld in 2007) were faithfully reported to the FDA by the trial’s sponsor (in 2012).
Crucially, leveraging CDISC could provide an analog to the paper case report form. How? By creating what Iberson-Hurst dubs a digital black box, similar to those found in airliners. Just as paper forms can be checked to show how and when an investigator altered clinical data, his black box, code-named “Mercury,” could allow the FDA to detect electronic changes as easily as it can with coffee-stained paper forms. Each time-stamped, audit-trailed version of a digital case report form could be examined to satisfy the regulators. Quite ingenious! Such technology could preserve all of the evidentiary benefits of paper – and the central role of the investigator.
Indeed, as Iberson-Hurst explains, “Mercury” or something like it could protect the investigator community. Physicians may not be aware of the risks of having patients use electronic patient diaries. “There is nothing in the regulations that allows the investigator to delegate their responsibilities,” he says. “Anything that the vendor is doing is the investigator’s problem. Most investigators probably don’t know that.”
Because the previous iterations of the CDISC e-source document have been so fondly received – which is to say that the CDISC team got a few red-hot emails – the next one will remain semi-private and be read only within CDISC.
But shortly, Iberson-Hurst expects, it will be released into the wild for the entire planet to criticize. “I believe we have some good solutions that will keep everybody happy,” says Iberson-Hurst. “The intention is to boost electronic processes and e-clinical trials.”
Nominally, officially, CDISC’s document is just about electronic patient diaries. But it will eventually impact the electronic data capture (EDC) providers and large hospital and physician practice IT companies specializing in electronic medical record (EMR) systems.
Once the FDA and the standards-setting bodies (not just CDISC but HL-7 and others) figure out e-source for handheld diaries, they do not intend to reinvent the wheel for EDC and EMR systems. That, after all, was the original intent to those who put together 21 CFR Part 11. If all this could be connected, Iberson-Hurst believes, life for the sponsor community and the clinical sites could get a lot simpler.
“E-source is about several things,” Iberson-Hurst explains. “The main one is obviously removing the paper. We’re also speeding the process and getting better compliance with e-diaries. But it is also about allowing multiple systems to collect the data. If we have multiple systems, then we need to integrate the data sources. If we use a single standard, the CDISC operational data model, it becomes a lot easier.”
There are, to be sure, a few definitions of e-source floating around the industry. Strictly defined, however, it remains a rare bird that scares some vendors (who fear losing their proprietary data formats) and sponsors (who worry about losing the snuggly, fuzzy security blanket of paper forms). Only Medidata, working for Johnson & Johnson division Janssen Pharmaceutica, has used laptops to implement pure e-source in non-diary trials.
Editor’s note: comments from diary vendors and the FDA have been solicited but were not available at press time. We hope to report those as they are available in the autumn. If you’ve not already complained about something to Iberson-Hurst, by all means drop him a line.