By Mark D. Uehling
July 14, 2004 | Must drug companies buy new technology? Not necessarily. The importance of making existing technologies hum was the refrain at the Drug Information Association’s 40th annual conference*.
“We really don’t like new technology,” confessed Mark Carroll of Merck, estimating that perhaps a sixth of the company’s clinical trials would use electronic data capture (EDC). “We like it to be tried and true. We are not going after the new technology until it has been proven.”
Carroll said something that very few people in the pharmaceutical industry have had the courage to utter: Just getting existing technology to work -- whether from vendors or homegrown applications -- is a major undertaking. “We might need more people who know less about the business and more about the tools, because they are so complicated,” Carroll said.
Jonathan Engler, CEO and founder of ClinPhone, put it this way: “The software itself is becoming commoditized. It’s less about the technology and more about the service.” Mentioning one leading EDC vendor in particular, Engler said, “I don’t think the software printed on their CDs is worth what they are charging for it.” Having said that, Engler said ClinPhone could not rule out acquiring another major clinical application to complement its industry-leading system to use the telephone to enroll and randomize patients and manage drug supplies.
There was other news. The North Carolina EDC company etrials announced an easier way to go directly from its systems into SAS datasets, the preferred format for statisticians who analyze clinical results. The SAS program Drug Development has been integrated with etrials’ QuickStudy offering, which includes a patient diary and interactive voice response system. The goal appears to be to make it easier for smaller sponsors of clinical trials to begin working with clinical data. “What we’ve heard, over and over, is the importance of data integration,” said Craig Pyne, etrials’ director of marketing.
At SAS, the company clearly understands how difficult it has proved to integrate the front end of clinical IT -- gathering data in the clinic -- with the databases at the back end. Big companies routinely spend tens of millions of dollars to integrate the two, only to export the data into another system for analysis. “We are offering the ability to eliminate the CDMS (central data management system) step,” said Chris Glass of SAS, where he is channel development manager for SAS Health & Life Sciences. “This does not require a lot of infrastructure, and it's not going to require a lot of extra personnel, in terms of database administrators.”
Integration was another theme of the meeting, with many presentations and vendors proclaiming their proficiency. But “integration,” like any other phrase, has a large number of meanings. Integic, a software company with $160 million in annual revenue based in Chantilly, Va., hopes to collate the spreadsheets, Post-it notes, and other stray documents that are used to run clinical trials. Integic announced Global Trials Manager, a clinical trial management system, handling administrative tasks, payments, monitoring visits, timelines, and project management across one or many studies.
The new Integic package was developed for Icon Clinical Research, a contract research organization (CRO), and should be ready later in the year. But the application will be sold to other CROs, and Integic said it could be used by pharmaceutical companies struggling to keep tabs on which trials are farthest behind, say, in the recruiting of patients.
Jane Light, principal consultant at Integic, said the software is designed to work with the always-behind, eternally optimistic mindset of the rank-and-file clinical project manager. As the details pile into vast snowdrifts scattered across dozens of spreadsheets and Access databases, it can be hard for executive-level overseers to know when to intervene. “It’s difficult to get accurate information across a study,” she noted. “So much of the tracking is done after the fact. It’s hard to be proactive.”
In the conference sessions, there was a rueful awareness that the adoption of EDC was more or less stagnant or growing slowly. Richard Nelson, executive director of clinical operations at PharmaNet, said training for EDC could not be overemphasized, and he walked his audience through the various options: in person, Internet, or CD-based. “Every EDC vendor will tell you their tool is easy to use and requires no training,” Nelson said. “As intuitive as you can make it, you still need training.”
In many a session at this year’s meeting, there was a growing recognition of the limited role of EDC in general -- or put another way, the critical nature of all the other computer systems that drug companies need to connect to EDC. In that context, the cleverest presentation of the week was from John Snapp, manager of system projects at the Duke Clinical Research Institute. “EDC systems are most commonly used in a standalone process that has limited interfaces with other necessary trial management systems,” Snapp said.
But then he described an ingenious adaptation of an EDC system to do an entirely different and equally specialized task: the management of a handful of academic centers all participating in the same trial. Once again, available technology was used to integrate disparate data that don’t always flow easily up the chain of command.
Duke’s trick was to tweak an unidentified EDC system so that entire sites were treated as “patients.” “You’ll say, ‘This will confuse people.’ It didn’t. It said ‘patient,’ but it really means site. It’s using the power built into the EDC system,” Duke said. As a result, managers and collaborating sites could see at a glance where the fastest problems -- and greatest delays -- were lurking. Snapp noted that administrators seldom have any idea how difficult it is to pull together what may sound like an innocuous report about patient enrollment throughout the trial. “We do clinical trials. We know reports. I have not seen vendors with useful reports. Usually they are specific to a protocol. If they’re generic, they provide little day-to-day value. You need a worklist of who do I need to call today to get their work in.”
When it came to electronic diaries, the news was not that paper diaries contain grotesque (if well-intentioned) fabrications by patients. That’s been in the journals for years. What changed is there now appears to be no attempt to justify a return on investment for handhelds.
Jay Pearson of Merck said a rough estimate of what diaries cost on a per-patient basis might be $800 to $1,200. No matter how much cleaner the data from such equipment might be, he said, it would be difficult to realize operational efficiencies equal to that cost. “We are focusing on the science driving it and whether it gives us a better chance at getting the truth, which is really what we want anyway,” Pearson said.
Using PHT as its vendor, Merck achieved high amounts of compliance in both paper and electronic diaries. But some users of paper diaries reported suspiciously predictable time intervals in the trial: 30, 60, 90 minutes. The electronic patients had a much more even distribution of erratic reporting, falling in a predictably random (i.e., Gaussian) manner into a bell-shaped curve. Some patients’ own comments illustrate unintentional admissions of departures from promptly and honestly reporting their experiences in real time. Two examples: “I only had to play catch-up 60 percent of the time” and “I had to go back to my old answers to see what I should be filling out.” Said Pearson: “The patients don’t view this as cheating. They are ‘helping’ you.”
At the PHT booth, marketing director Sheila Rocchio reviewed some of the company’s latest technology, most of it having premiered recently. The Palm Tungsten T3 was perhaps the slickest device, with a super-clear transreflective screen. But there was also a Web-enabled Nokia 600. Rocchio said some sponsors of clinical trials are interested in sending an SMS message to clinical trial participants. Wireless diaries are hot, and, in general, “business is up 60 percent,” Rocchio said. “We’re sensing uptake.”
So is Mahesh Singh, an India expert at Pittglio, Rabin, Todd & McGrath, an IT consultancy. He and other presenters focusing on India said that as the country brings its patent law into tighter harmony with western countries, local companies and foreign companies will expand operations there. “There is more work going on in data management and now trial management,” Singh said. “There is going to be more work with high-throughput screening.” The DIA crowds heard that growth in India could be as fast as 30 percent a year, partly because of large, untreated populations of patients with the same diseases that afflict more prosperous countries. It doesn’t hurt that big U.S. companies such as Pfizer, Lilly, and Quintiles have been in India for years, both training investigators and learning cultural subtleties there.
Perhaps the most galvanizing presentation was late in the week, after most of the attendees had departed. Jerald Schindler, assistant vice president, global biostatistics and clinical technology at Wyeth Research, discussed an unabashedly audacious project to tie together every last scrap of clinical data. “What we’ve done at Wyeth is create one system so that everyone looks at the same system to get the same information,” Schindler said. “Everybody has the same number all the time. That’s not just true of enrollment data, but for all the other information in the system.”
Some of the schematic charts listing the component applications and databases were dizzying in their complexity. But to the end-user, all of the Wyeth data are available via the Web. Said Schindler: “We’ve built a system where we can plug and play the best in class applications. The advantage here is that we can shop around. We can look at all the systems. We are not limited to one system from a particular vendor.”
Douglas May, a consultant who had worked with a number of companies similar in size to Wyeth, was impressed not only by the technical details, which are tough enough, but also by the diplomatic exercises that required different fiefdoms to standardize around key forms, terms, and processes. “Wyeth went through a painful exercise, culturally, to standardize,” May said. “They did the hard lifting of getting all the groups to sing from the same piece of music. That may be the biggest obstacle in other companies. That is a real accomplishment. In other projects, that is the downfall of many great visions.”
Michael Barrett, a former Forrester Research analyst now working solo, said he too believes Wyeth’s progress to be an unusually broad integration of clinical data throughout a multibillion-dollar company. “This is relatively rare. I’ve worked with a number of large top 10 drug companies. The conversation in those companies focus on integration around a single drug or trial.”
The bad news from Barrett was that some of the companies selling proprietary software to collect or manage clinical data might fall by the wayside. “I see competitive advantage moving quite suddenly away from the integrated front-end or back-end solution to the company willing to do away with proprietary solutions in either space. Yesterday’s frontrunner can’t afford to take the risk of interoperability. They’ve got the investment sunk into yesterday’s proprietary, marketing-leading tool.”
*DIA: Washington, D.C., June 13-17.