Is electronic data capture (EDC) creating inefficiencies? That's one conclusion to be drawn from the interim results of an eClinical Forum survey.
Surprisingly, the survey suggests that the use of paper case report forms (CRF) is rising even amid growing adoption of EDC. While most sponsors are committed to EDC, site staff believe it is forcing them to spend more time working on trial data, not less.
“For sites, there's an increased workload,” says Richard Perkins, president of eClinical Forum, a Strasbourg, France-based nonprofit group made up of 60 members from pharma companies and other businesses in drug development that use EDC. “Forty-six percent of respondents said they have to do more work because of the presence of EDC. That's double what it was eight years ago,” when the survey was last undertaken.
Part of the reason for that decline in productivity is overlapping regulatory burdens on sites and sponsors. “Many sites have SOPs [standard operating procedures] that require them to create both a paper and an electronic version of a CRF. Some even have to create three versions—another for the patient record,” says Perkins, founder of Friesenheim, France-based ConSept Consulting.
Can the blame be deflected to an entirely different industry? Maybe. Dimitry De Schepper, life sciences product manager for TechTeam, a tech support company and eClinical forum member that worked on the survey, says part of the problem is a slow adoption of another type of huge software system—electronic health records (EHR). “I think the evolution of EHR in the marketplace is the cause. I sense that people are afraid of putting the data right into the EHR, so they put it on paper first,” he says. (Here's a story we previously published on TechTeam.)
The survey, called “Investigational Site Perspectives on Clinical Trial Information Systems,” is still collecting answers. The survey suggests that EDC systems typically don't have the trial data until 48 hours to a week after it is written down by hand or collected in some other system. (Some sites and sponsors report much longer lag times.) “You would think it would be immediate, but it isn't,” De Schepper said.
In 2001, the last time the survey was conducted, 16 percent of respondents said study data was initially entered from a paper CRF; in 2009, that number rose to 26 percent. Also in 2001, 66 percent of the data was initially entered into a paper patient file. Eight years later, that number had increased to a bit more than 71 percent.
Some of the barriers may be cultural. Almost 60 percent of respondents said they thought it was never appropriate to enter data into an EDC system in front of a patient, and 17 percent said it was appropriate only in rare circumstances. Perkins says these answers could be traced to the fact that about half of the survey's respondents were from Asian countries. He said those at eClinical Forum may soon be exploring that wrinkle in the results.
De Schepper wasn't surprised by the disproportionate number of survey respondents from Asia. "Japan is the second-biggest pharmaceutical market after the U.S., with major pharmaceutical players like Takeda," he said. "The fact that we used an online survey and provided translation in Japanese made the study more accessible for that region."
After the closure of the survey, the Asian survey respondents will be treated as a subgroup, said De Schepper. "We will investigate whether their answering pattern is statistically different from the rest of the population," he says.
Personnel at sites, alas, seem to see EDC as a time waster—not a time saver. And that perception may have profoundly shifted just who fills out the survey. In 2001, 69 percent of respondents were investigators themselves, and only 12 percent were study site coordinators. In 2009, that flipped: 71 percent of respondents were study coordinators or managers and just 14 percent were investigators. So the Forum survey answers are now coming from those in the trenches, with less education but more intimate knowledge of the EDC systems.
Most of the current survey's questions were the same as those asked in 2001. Paul Gemmel, professor of health care and service management at the University of Ghent in Belgium provided statistical analysis.
Back in 2001, 52 percent of the respondents (there were 830 of them) said they had had some experience with EDC. Today, 95 percent of the 629 respondents had EDC experience. So awareness of the technology is growing. What was something of a shock, says De Schepper, was that just 58 percent of the respondents said that current trials they are involved in are using EDC. That's risen from 13 percent in 2001, but it's not the increase that eClinical Forum expected.
“Clinical trials are becoming more and more complex,” said De Schepper. “We thought electronic measures would be much more widespread by now.”
Pockets of Happiness
There's been an explosion in global trials seen since the last survey eight years ago. De Schepper said this year's survey showed an interesting geographical split among who's rapidly adopting EDC and who's not. “In the U.S. and Europe—the traditional markets—you see them go very fast in terms of the evolution of a technology,” he says. “Then you have the more emerging markets running behind and really needing support in setting up trials, in getting internet connectivity and sometimes even needing computers.”
Perkins adds that the emerging markets in Asia may be the exception, as they are known for rapid adoption of new technologies. Other interesting results: 27 percent of respondents said that EDC is used in all of their trials. In 2001, that number was less than 1 percent. In those trials, satisfaction levels are impressive; fifty-six percent said they were satisfied with EDC, and another 11 percent said they were very satisfied. So not all the news is bad.
Can EDC be improved? The respondents had suggestions: adopting clearer governmental regulations around EDC; implementing better edit checks at the time of data entry; implementing faster upgrades; making systems more user friendly for sites; better harmonizing of modules used by project role (investigator, monitor); speeding up the systems; allowing textual notes field for unusual situations; standardizing eCRFs; and, in particular, more training and help-desk services.
“Speed of communication was the biggest problem,” says De Schepper. “Training, support, ease of access and lack of standardization—those are the real barriers of using the tool.”
One conundrum is that as EDC systems evolve, they're getting more complex, which is making them harder to use. “It may be very intuitive to people using it daily, but there are doctors all over the world who are not used to working with it,” says De Schepper.
He says he hopes the report will spur more sponsors and contract research organizations to make sure all the people working within their EDC systems have lots of training and robust help-desk services. Otherwise, the natural bent is to not use the systems in the way they were intended, which in turn slows studies down.
Do people who work on trials think the application of EDC to all trials is a good thing? Sixty-three percent of respondents said yes, while 21 percent said they thought EDC should only be used for “simple trials.”
The survey results are based on the responses eClinical Forum and TechTeam garnered by Sept. 23, but the organization is keeping the survey open so that more responses can come in. Much of the survey is, as of today, available on the eClinical Forum website. In January, the full study becomes available.
—by Suz Redfearnd9A2t49mkex