By Mark D. Uehling
January 12, 2004 | Nextrials, an application service provider (ASP) specializing in electronic data capture (EDC), says it has developed the only fully integrated, Web-based method to manage lab data and clinical chores that traditionally require separate applications. Need patients randomly assigned to a drug or a placebo? Nextrials can do that. Need to generate MedWatch PDF documents describing adverse events? Done. On the fly.
Founded in 1999, the San Francisco company's recently released Prism 2.7 software could shave market share from better-known EDC competitors. Nextrials has 1,000 users at 200 clinical sites, with customers such as Novacea, Genentech, Baxter Healthcare, and Medtronic. The company has completed more than a dozen trials, with a similar number under way.
"This is novel," says David Shiple, senior analyst at Forrester Research, commenting on Nextrials' way of handling laboratory data. "Usually, lab results go back to back-end clinical data management systems and are not transacted through the EDC interface." Even with a small installed base, Shiple says, Forrester still ranks Nextrials among the top dozen EDC vendors.
Everything Nextrials does through its Web interface is Unix-based, running on an Apache server. The database is not Oracle but PostgreSQL, a free, open-source relational database. Other terabyte-scale users of that database include the American Chemical Society and Affymetrix.
Untarnishing the EDC Industry?
With everything in XML, and participation in the Clinical Data Interchange Standards Consortium (CDISC), Nextrials seems well positioned to grow. It's also ready for the backlash when other EDC solutions don't work. "We've been frustrated by all the bad experiences people have been having with other companies," says co-founder and president James Rogers. "You go to Drug Information Association meetings, and you hear horror stories." Services and consulting are an important part of the company's revenue mix. "We're getting direct feedback from the sites about what's working and what's not working," Rogers says.
Nextrials says it is able to maintain the integrity of a local lab's information. Each lab in a trial may use different ranges and units. One current study is collecting lab data on the Web from 400 different labs connected to more than 60 clinical sites. On the back end, the Nextrials software converts all the values into standard international units, affording sponsors a single view of the data.
Nextrials can also ensure that no additional results - data not part of the study protocol - get transmitted en masse to a trial sponsor's database, as happens with the most sophisticated labs that move batched data electronically. Merely taking possession of such reports may pose a regulatory risk: Getting data creates a legal obligation to inspect it. "There is a lot of pain associated with lab data," Rogers says. "Lord help you if there was something wrong on a lab test you didn't look at."
Nextrials also integrated a safety system to monitor adverse events that emerge during a clinical trial. The system can generate FDA-mandated MedWatch PDF forms on the fly by using two linked databases - one for clinical data, one just for the adverse event data.
And its product-tracking features, Nextrials says, allow medicines to be shipped to clinical sites as needed - with little human intervention. That reduces drug wastage, a welcome change for biotech companies with perishable or expensive-to-manufacture formulations. "You're able to log on and see who exactly allocated the drug at what time and for what reason," says Alan Arroyo, associate director of clinical operations at Novacea, a small Oregon biotech company that is using Nextrials.
Arroyo has tested EDC solutions installed on dedicated laptops. "You frequently have hardware problems," he says. "Data don't replicate or data are found on the laptop, but somehow they don't migrate to the central database." In contrast, he says, Nextrials has been highly reliable, with calls to technical support returned within 30 minutes.
There are well-regarded vendors specializing in each component of what Nextrials offers. But in the real world, Nextrials says, linking legacy and third-party products requires a large amount of manual intervention. In-house and commercial applications and their data must continually be compared, reconciled, and managed. "You have to take a hard look at how they have been integrated and supported," says co-founder Anthony Costello, Nextrials' vice president of data services and product development.
Formerly in charge of oncology data at Genentech, Costello notes that different contract research organizations (CROs) may all structure their lab, safety, and trial data differently. On a consulting basis, Nextrials can synthesize such data from a variety of CROs, allowing a sponsor to analyze a portfolio of compounds or trials.
Ron Waife, a clinical trials consultant and president of Waife & Associates, says that Nextrials' approach could be interesting to some sponsors, but cautions that proprietary or legacy applications to manage lab data may not be ripped out of the largest drug companies right away.
"These sponsors are unlikely to want to move to an integration platform based on a small, new company's idea," Waife warns. "But integration of diverse enterprise applications remains an important long-term efficiency and competitive goal for these companies."