Yesterday, in a separate story, we talked with consultant George Laszlo about clinical data repository (CDR) functionality and terminology.

This article is more forward-looking, and mentions some choices and long-term benefits for a CDR.

Laszlo thinks the industry should ponder where clean clinical trial data should live. How easily will ad hoc reports be run off that bucket? Laszlo is the first to admit every company in the industry will have its own preferences and solutions. But he’s not sure the existing landscape of clinical data management systems (CDMS) and electronic data capture (EDC) solutions will be adequate to the task of working with clinical trial and other types of data a) after it has been cleaned and b) in a validated manner that serves everyone at a sponsor company.

Where To Put It?

Many sponsors of clinical trials, he says, are slowly waking up to the fact that some data analysts asked to run ad hoc reports have confusing, hard-to-reproduce workflows that may or may not be occuring in a 21 CFR Part 11 environment. Statisticians may have to use umpteen different applications, cutting and pasting data between them.

“One overriding concern is that the manual, disjointed processes that they are now using by jumping from one application to another, like using Excel for one case and jumping to SAS for another and S-plus to get graphics out,” says Laszlo. “It’s not that they can’t do what they want to do, it’s that the way they’re doing it is inefficient.”

From a regulatory perspective, a compliant Microsoft Office solution is a rare bird indeed. But sales, medical affairs and other colleagues in some companies want to see reports relying on clean clinical data. Simply determining whether such reports are feasible can take time. Generating the reports using Spotfire or some other application is another matter.

Indeed, Laszlo says the process of finding, transforming, and analyzing trial data can take months at many companies. Laszlo doesn’t think it should take months, mind you. Organizations with the right kind of repository should be able to get the data and analyze it far more quickly. “They could do it in a week or two days. The data are there,” says Laszlo. If other colleagues (with the appropriate roles and permissions) had access to the data in a repository, it could relieve pressure on the statisticians.

Four Repositories

Here’s the rub. There is not a large selection of commercial clinical data repositories. Laszlo can think of four. (If he’s left one out, we’re sure we’ll be alerted by eagle-eyed readers.)

1. One of them is from a Boston-area company we wrote about a while back, Waban Software, which Laszlo is quite familiar with as a consultant. He has several ongoing projects with Waban.

With that disclaimer, he still thinks it’s the best. Says Laszlo: “I think Waban is the most mature. It is able to do most of this and has end user acceptance higher than the other products.” The company’s website is here. One novel feature is data “scoping.” It’s the ability to take a snapshot of clinical data at a certain time point, excluding prior and later data.

2. SAS Drug Development is one of the better-known data repositories. In the interest of full disclosure, Laszlo has done work for SAS as well. This earlier ClinPage story concerns a recent partnership between SAS and Medidata, the New York City electronic data capture vendor.

3. There’s a European entry, Entimo, with a lesser-known product called ICE Portal. Among other big pharmas, Schering is a customer.

Entimo touts its CDR’s ability to mend fences in the investigator community. “entimICE Portal technology can aid in addressing some of the most common investigator complaints, such as out of date information, difficulty in contacting project staff, and late grant payments. Improving communications between different groups within a sponsor organization, like clinical, sales, and marketing, also helps prepare the way for product launch and commercial relationships with former investigators,” the company website notes.

Last But Not Least

4. Finally, there’s Oracle. As this press release confirms, Laszlo points out that Oracle sold at least one clinical data repository to Boehringer Ingelheim in March, 2006.

Laszlo is surprised the world hasn’t heard more about the product, given that Oracle is not shy about marketing its products. Its repository has since been rechristened the “Oracle Life Sciences Data Hub,” as this PDF from Oracle explains.

Laszlo says Oracle may have concerns about cannibalizing sales of its CDMS, Oracle Clinical. “The need for Oracle Clinical pretty much drops dead,” he says. “People are going to migrate over a five year period. They are going to sunset Oracle Clinical and use it as a vehicle through which they pipe data to a repository, but only because they have legacy data there.”

He didn’t mention it, but it’s hard for us to see why the other major CDMS, Clintrial, from Phase Forward, would avoid a similar fate as EDC systems clean the majority of the data and repositories handle the rest.

Why It Matters

For companies not handcuffed to a CDMS and able to consider a clinical data repository, Laszlo says, the real payoff for a CDR will come in the future, with broader use of data standards like those promulgated by the Clinical Data Interchange Standards Consortium (CDISC) or HL7.

It is here that the return on investment in a CDR is readily apparent, with functionality that is both ready for personalized medicine and attuned to a less-paper centric process. With a repository in place, Laszlo says, “you are in fact enabling the ability to go across trials and data domains. You can mix the genomics and the trial data and the lab data and even animal data so that you can find other information that you could not have been able to do before.” We so seldom hear anyone even mention preclinical data, much less know what to do with it, that we had to sit up straighter after that.

The industry’s predominant information storage medium, he says, will fade. Paper will come to seem less quaint, less essential. “The speed with which the industry processes its data will be an issue from this point forward,” says Laszlo. “’We’ve always liked paper’ will not be a credible defense when someone asks why the results weren’t available sooner.”

Editor’s note: Day-to-day users of any of the above repositories are welcome to contact .(JavaScript must be enabled to view this email address); we would love to hear raves, pans or other observations about the right repository for your organization.