By Mark D. Uehling
May 15, 2004 | This year's sessions on e-clinical research repeatedly circled back to one theme: the dance between new technology and data standards that could pave a seamless electronic path from the bedside to the regulators who analyze the data. For the moment, the destinies of software and standards are intertwined, with chicken-or-egg uncertainty about which impedes the adoption of the other.
"We’ve had a fair amount of time to embrace these technologies and these standards," said Ken Getz, co-organizer of the two-day clinical track and chairman of the Center for Information and Study on Clinical Research Participation. Getz noted that 80 percent of clinical trials still collect data using paper, not e-clinical software. "There are many barriers to adoption. There are internal issues, cultural and philosophical, that industry must deal with." One ominous indicator: Companies are now spending more money internally to develop their own tools than they are with commercial vendors.
Picking up on the theme of hesitation to adopt technology, the other co-organizer of the clinical sessions, Becky Kush, president of the Clinical Data Interchange Standards Consortium (CDISC), presented data about EDC adoption internationally. The reasons EDC is faltering overseas turn out to be identical to why it is tentatively used in the United States. Among other things, organizations sponsoring clinical trials fear changing their business processes (76 percent) or see no return on investment (71 percent).
Ironically, given the historical timidity of the pharmaceutical industry, some of the largest acts of bravery described at this year's meeting were undertaken by big, bureaucracy-laden sponsors of trials, not intrepid smaller companies. The big companies seem to be trying to conform to CDISC standards on their own, having given up on the major EDC vendors to help them.
The Miracle at Merck
The most surprising numbers and upbeat testimonial of the week were served up by Jim Karkanias, executive director of clinical data processing in the Human Health Division at Merck. Karkanias outlined a tight and friendly partnership with two vendors, Microsoft and DataLabs.
Karkanias said the use of garden-variety Microsoft collaboration tools (.Net, InfoPath) and research-specific EDC software from DataLabs appears to reduce the cost of managing clinical data by 22 percent. A full 30 days can be shaved off the paper-bound process of closing a clinical trial database, Karkanias estimated.
The savings per trial could amount to $400,000. Probably the most eye-catching figures were from a hypothetical Merck trial taking 100 arbitrary units of time. The Microsoft-DataLabs electronification of that trial could complete the same amount of work in just 30 time units.
Specifically, Karkanias said, there was a 70 percent drop in data review time and a 34 percent drop in study design time. He believes the improvement will continue. "This process architecture we’ve come up with has a lot of headroom for improvement," he said.
Karkanias even showed a picture of an Xbox, Microsoft's gaming console, and said Merck is considering the device for clinical applications in which an easy user interface might encourage patient participation in a clinical trial.
There was equally good news from the University of Florida. Ron Marks, professor of biostatistics, reviewed familiar numbers about physicians doing a single clinical trial -- and then leaving the business due to paperwork or financial burdens that trials bring. Citing modern industries such as airlines and banks, Marks said greater efficiencies are inevitable. He pointed to a large hypertension trial at his own university as an example.
The INVEST trial, as it's called, is running in 14 countries with more than 22,000 patients and 860 physicians. Just about everything is happening electronically. "The investigator training was online," Marks noted. "The institutional review board was online. We did medication dispensing, adverse event processing."
Wyeth Builds its Own
But at Wyeth, Jerald Schindler, assistant vice president of biostatistics and clinical technology, had a more sobering story. As long as four years ago, Schindler said, Wyeth wanted to explore standards. No vendor was interested, he said.
Schindler said Wyeth has not put EDC on a pedestal by itself. Rather, it is trying to reconceive the entire clinical trial process in digital terms. "We wanted to make all our clinical trial information accessible through a common interface," Schindler said. "We didn’t want to have this issue of duplicate data."
Even an apparently mundane question such as "How many patients have enrolled in this trial?" can have several answers. To avoid that, Wyeth built a real-time database with administrative data, investigator names and addresses, quantities of medicine and placebos, and a host of other information that commercial software vendors typically lump into two or three different applications that may or may not access the same database.
Schindler sounded disappointed in what the vendors offered. "There isn’t any company that can sell you the whole solution soup to nuts," he said. "We looked at all the systems out there. For one reason or other, we rejected all of them." With major companies such as Merck and Wyeth still energetically developing their own software, of course, a smart vendor's only chance for survival may be to develop clinical tools individually, and not try to supply an entire workbench.