The FDA is about to close its comment period for its new draft guidance on electronic "source," or master, versions of data in clinical trials. Should the agency implement and finalize that document in the coming years, there could be profound changes in the way that clinical data is collected by the pharmaceutical industry.

The biggest potential change? Paper-based facts now filed away in boxes might become electronically accessible to FDA inspectors. With that prospect in mind, Clinical Ink is an interesting company at an interesting juncture in time.

The firm is based in North Carolina and makes a $1,500 tablet for clinical trials. The device is being piloted in several small studies. It comes with preloaded medical dictionaries and proprietary software that decodes a user's handwriting in ways custom designed for industry-sponsored research.

New Tool

Clinical Ink president Ed Seguine formerly toiled in the venture capital and technology branches of Lilly, before creating a Lilly-backed startup, Fast Track Systems. Medidata Solutions, the electronic data capture (EDC) supplier, later bought Fast Track.

Seguine says the guidance could be a turning point in the acceptance of clinical trial data not initially recorded on a piece of dried tree cellulose—i.e., paper. As with any new communication from the FDA, there is always the potential for a period of additional ambiguity and confusion. "This is a really important step for the industry to be clear on this," he says.

If the new guidance becomes policy, Seguine believes, the FDA will expect a much larger amount of data than is now collected and verified by industry. Some vendors may struggle to handle everything that the FDA wants to audit and verify in an inspection.

Says Seguine: "The FDA is asserting that the investigator has to have the control of all this information. It's reiterating that the power and responsibility is at the investigator site."

FDA Guidance

One new wrinkle in the FDA guidance, he says, is that the traditional concept of a case report form (CRF) is being enlarged. At the FDA, it seems, a CRF is no longer just a simple form. It is a large collection of data.

Doctors have always gathered that compendium of data. But it's not necessarily included or managed in industry's CRFs today. Says Seguine: "The physician needs to collect a more encompassing set of information. Not just the discrete data you collected traditionally in the CRF."

As an example, Seguine supposes that a protocol requires three blood pressure readings spaced at five-minute intervals. What should be recorded? An average value? Or all three numbers?

Offering More Data

While industry would have typically just supplied a single number in the past, Seguine suggests, a new era of sending the FDA all three numbers may be approaching. Traditional EDC systems, needless to say, would probably just have the single value.

How to create a more comprehensive record? Clinical Ink's system can manage all the values. Indeed, it contains electronic representations of the paper forms that nurses or clinical research associates would expect. That means that there is no partition or split between what gets saved on paper and what gets saved electronically—every bit of data only exists in digital form.

The company's system also records both structured and unstructured data in free text notes that may be newly germane to FDA inspectors should the recent guidance be adopted. In short, Clinical Ink is trying something the EDC companies have never dared to do: a) getting rid of the paper and b) respecting the way doctors and nurses generally work when not doing research.

Fear Factor

Such changes could be welcome for efficiency-minded firms. They could be terrifying for companies that have memorized (and perhaps misunderstood) earlier FDA commandments about paper as a sacrosanct material.

For clinical sites, Clinical Ink believes its approach might be more attractive than the present technologies thrust into the hands of doctors and nurses. The daily demands of recording information for clinical trials are onerous.

There is little argument that the age of EDC has created new work and tension headaches for the diligent clinical research associates (CRAs) who must constantly transfer information from paper to computer and float between those two worlds. The absurdity of that transfer is likely to grow in the coming years because basic physician computer systems, electronic health records (EHR), are becoming more sophisticated and widely used.

Paperless Approach

The main challenge with Clinical Ink's approach is ensuring that a few taps of the stylus or squiggles on the tablet's screen have been interpreted correctly. In a demonstration for a reporter, the process had impressive speed and accuracy. Our guess is that many CRAs would prefer working with the Clinical Ink tablet than endlessly moving between a paper form and computer keyboard.

For sponsor companies, of course, doing away with paper could save money. Lots of it. Cleaning up after EDC remains expensive, with monitoring budgets consuming perhaps a third of the cost of the trial.

"Nothing is being done to address that," Seguine notes. "For pharma companies to fundamentally address the biggest drivers of cost, you need to collect 100 percent of all that information electronically in the first place. If you do that, all that electronic information can be reviewed remotely."

Less Monitoring

No rational person would have proudly developed the present pharmaceutical work flow for source document verification (SDV). It's a Rube Goldberg approach. The question now is whether the new FDA guidance will persuade any companies to abandon what they've done for years. Contract research organizations are particularly invested in the paper- and people-intensive SDV process.

Seguine notes that some big firms have cautiously tested less-than-100-percent verification of paper source documents, only to be chastised by the FDA. Having been burned in that experiment, the sponsors naturally resort to Dickensian work processes and resume 100 percent visual inspection of the forms again.

In some ways, Clinical Ink is proposing to take EDC to the next level. "EDC is all about collecting the data," says Seguine. "But you've got this mass of additional paper documents at the site that you have to monitor. EDC has not reduced the cost of clinical trials. EDC fundamentally is disruptive to what goes on at a research site."

In contrast to many EDC systems, he says, the people using the Clinical Ink system tap in data and immediately decide if there are errors. That prevents mistakes from being created in the first place. "We run all our edit checks locally," says Seguine, referring to the computer in the hand of the user. "Our edit checks are super fast and nonintrusive. We don't show every edit check as soon as they enter the data. You can't be constantly interrupting them."

IT Magic

Just porting a clinical application to a tablet is easy, Seguine says. It's important to understand what's under the hood. One nuance is that Clinical Ink's tablet, made by Motion Computing of Austin, Texas, contains so-called "active" screens to decipher what the user scribbles. His system can analyze more pixels from the handwriting than the technology of just a few years ago. "We could run the same application on a passive screen and it wouldn't guess as well," Seguine says.

The Clinical Ink software is probably as significant as the hardware. In some cases, a grid that encloses each letter individually helps the tablet decode your jottings one character at a time. In other cases, the software can drop its own grid over a handwritten, cursive word.


Clinical Ink's system analyzes handwriting and makes changes easy (highlighted orange area)

Tellingly, the Clinical Ink user interface was adapted from an EHR system and uses a familiar metaphor of patient charts. That should be more conducive to the way people actually work in clinical settings than the database-driven forms of many EDC computer screens, which are typically divorced from the rest of the way patient care is managed. "You need an application that is designed to getting things done," Seguine says.

Sign Here

Handwriting recognition software, while initially maligned in the news media, has improved to the point where the tablet almost instantly knows what the user intended. The usage of check boxes and multiple choice questions helps improve its unspecified error rate. "You write your natural handwriting," Seguine says. "It converts everything to digital text."

Source document verification is wholly eliminated, Clinical Ink says. The numbers of queries are massively reduced. Seguine says that could shift the clinical investigator's perception of clinical trial work as being tedious and poorly compensated. "Our solution is also part of the solution to the declining level of participation of researchers in the U.S. if you actually make it easier for them," he says.

About now, some readers may be saying, well, all we really need is a solid linkage between EHR systems and those for industry-sponsored research. Not so fast, says Seguine.

EHR-EDC Complications

First, there is no way for any EHR today to understand a trial protocol and its sequence of exquisitely choreographed and contractually mandated events. Second, for all the fine work done by the Clinical Data Interchange Standards Consortium (CDISC), it has yet to issue a way for audit trail data to be passed around between different types of systems.


Clinical Ink: Deciphering cursive notes from physicians

Without those missing pieces, it won't be easy to really hook up an EDC system to an EHR system. That could mean trouble if the FDA arrived at a clinical site on a random Thursday morning and asked So, with your patient #A731HMQ, who changed this CRF? Why? Does this CRF match what's in this patient's chart?

Yes, there are a few companies claiming full EHR-EDC integration. But ordinary, routine connections between multiple clinical trial systems and multiple general-purpose medical systems remain years or even decades away, in Seguine's view. He notes that Britain has had a unified national EHR for 15 years, and it still hasn't been adapted for usage in clinical trials. (We've made a note in our calendar to call Seguine in 15 years, in 2026, and ask him if the EDC-EHR linkages have been resolved.)

EDC-EHR wiring connections could remain messy, he says, given the complexity and heterogeneity of the systems on both sides of the chasm between routine health care and industry-sponsored research. Clinical Ink's tablet, in contrast, is ready today.

Two previous ClinPage articles concern an FDA webinar on esource and an exploration of the new meaning of CRF.