When Speed Means Slow
August 29, 2007
The Clinical Data Interchange Standards Consortium (CDISC) does not want for savvy high-energy volunteers or high-profile member organizations. For evidence of that, just hang out at CDISC’s user meeting, which is coming up in October.
There’s just one problem. Using CDISC data standards, and doing so consistently throughout an organization, remains an unrealized goal in the industry. It’s a mountain peak that no company has bagged. Gartner VP and distinguished analyst Carol Rozwell, to our way of thinking, is a guide for that journey.
Rozwell serves pharma in a capacity similar to what Guy Kawasaki used to do for Apple, in a dimly remembered era before Apple specialized in disposable plastic toys. With “evangelist” in his job title, Kawasaki could make Apple seem like something more than a cult for artistes—a company with products relevant beyond a small band of graphic designers and anti-Microsoft jihadists.
Like Kawasaki, Rozwell can make the thorny realities of using clinical data standards throughout an organization seem clear, not like one of those vast diagrams of the human genome. In talking to Rozwell, we always feel like we might understand the business case for standards.
Not long ago, when we reached her, she was hard at work at her forthcoming Clinical MarketScope report, due later this fall. It will be an analysis of the clinical trial technology landscape. In our call, Rozwell was reflecting on whether the market for electronic data capture and other applications may finally be juicy and sizable enough for Fortune 100-sized firms to be scouting for acquisition candidates. “That’s a higher probability than it was 18 months ago,” she muses. The growth rates might be enticing to a goliath firm, she says.
No one in the industry really likes to talk about drug safety. There is a hope, as with a tooth ache, that it will pass. Rozwell jumps right in. She believes clinical data standards, properly used, could illuminate important scientific signals already lurking in the data. Rozwell says the potential of standards to facilitate prospective data mining is real, especially if an organization wants to look at more than one study.
“These are the kinds of capabilities that an organization gains when they have the opportunity to look at cross study data,” says Rozwell. “This is certainly one of the things that the FDA is thinking about as they talk about being able to use information post-study in order to protect public health.”
On a more immediate level, Rozwell says the CDISC Clinical Data Acquisition Standards Harmonization Initiative (CDASH) is progressing well. Its goal is to standardize case report form (CRF) structure and integrate that with the existing body of CDISC standards. “We’re standardizing the way that the data is collected right from the beginning,” Rozwell explains.
She’s impressed by the breadth of participation that CDASH has enlisted, which includes representatives from FDA, a silent partner in many CDISC discussions. The effort also has help from AMIA, ACRP, ACRO, BIO, Baylor, Duke Clinical Research Institute, PhRMA and both NCI and NIH. That is a pretty eminent list to agree on anything, and (we would guess) a tribute to the depth of support CDISC enjoys across government, academia and industry. Having Becky Kush running everything can’t hurt.
We asked Rozwell how the CDASH effort will accommodate the need for unique data elements in a CRF. It won’t be a problem, she says, with the people on the project understanding the need to balance one general approach for the industry and a bit of plasticity for special situations: “They know the standard has to be flexible enough to allow customization when it’s necessary.”
Given the maturity of the CDISC standards, we always have one question. Why aren’t more organizations drinking the CDISC Kool-Aid, and forbidding trial-design idiosyncracies that are demanded by one division or therapeutic area (TA) within a big company?
Another factor in standards adoption is more interesting. Rozwell walked us through a simple paradox. On the one hand, she points out, getting to the end of a trial is a reasonable goal for an individual manager or clin ops team. On the other hand, an organization’s larger, longer-term goal may be to have one trial data template, not eleven or an infinite number based on the preferences of a TA division head.
Rozwell is clear-eyed about this. “Speed really seems to be the biggest factor of resistance that we’ve encountered. If you’re really going to complete your study quickly, the way you’re going to do it is ignore everybody else and just go full speed ahead with what you think is the right approach. That’s going to be fastest, right? Standards, by definition, imply that we have to have consensus among at least two parties. And getting consensus takes time.”
But to her and Gartner there may be higher priorities—at least inside the most forward-thinking concerns. “If companies are going to be able to do the cross-study analysis, if the FDA is really going to be able to create that database of information that can be perused post-study to protect patients’ safety, on an ongoing basis, we have to have standards in order to do that,” Rozwell notes.
Let’s boil that down. Worrying about speed on a trial by trial basis, paradoxically, slows down your organization over the long haul. And going slow (i.e. using data standards broadly) will have a short term penalty. But it will lead an organization to faster decisions and deeper insights into your clinical data later.
Rozwell, we hasten to add, was gracious enough to be the first subject in a small but risky clinical trial—the podcast project here at ClinPage. In that particular study, there was no endpoint, no safe or effective treatment. Rozwell, a trooper, agreed to participate anyway.