OK, he doesn't like pens. But he's got the right idea. "Paperless" is a metaphor for automating drudgery, cutting costs, and improving work quality at the same time. Depending upon whose estimate you prefer, 50 percent to 70 percent of operations in a Good X Practice (GxP)-regulated laboratory are consumed by compliance issues.

"In a regulated laboratory, this really means management of the data stream," Talbot says. "[Information about] approved methods gets collected in a combination of laboratory notebooks and folders. There are boxes and boxes of chromatograms and spectra-generated information from which the lab used to generate results, which are stored on assay sheets because the raw data are too hard to summarize."

"Not only do the labs have to manage all these data and all this paper, but they also have to manage the relationship between all this information. At any given time they have to [be able to] look at a data point, be able to access the assay sheet to understand how that result was generated, and even be able to drill down to the raw data," he says.

Most companies could cut this overhead in half, Talbot says, by making smart use of IT and automation tools. But progress has been slow, stymied by cultural and cost barriers, and a fragmented supplier market in which tools were too broad, too narrow, or unable to integrate.

"That's changing," says Talbot, a longtime IT exec (18 years) in the pharmaceutical industry with stints at Sandoz (now Novartis), Berlex Labs, and Purdue Pharmaceuticals. As executive director of research systems for Purdue, Talbot spearheaded the company's successful efforts to modernize its R&D labs.

Over the course of four years, Purdue implemented the Paperless Platform at its R&D facility in Ardsley, N.Y. Last summer, the finished system rolled out, and plans are to roll it out globally to R&D in the United Kingdom and Germany. The early results look promising, but will take time to quantify.

Kicking the paper addiction isn't easy, Talbot concedes. Upgrading usually involves knitting together five separate systems: chromatography data collection, electronic lab notebooks, scientific data management systems, traditional LIMS, and links to an enterprisewide document management system.

"There are several major barriers," Talbot says. "The first is that the scope of project is overwhelming, especially if you have any experience in the lab with any one of these systems projects. Multiply by five, and you get scared pretty quickly. The total set of features you're talking about can be really terrifying. You need some kind of way to focus down on what set of features you want."

Cost is another major hurdle. So is systems validation. The effort to validate each system can last 12 to 18 months. "Multiply that by five, and people really get nervous," he says. Dealing with multiple vendors and tricky integration can present a long-term challenge.

"Ten years from now, if you want to see how particular things went from a study, you've got to bring back data. Keeping those systems synchronized to tell a whole story 10 years from now is very difficult. Every type of software comes with an upgrade twice a year, and keeping both current and in sync with the other four systems is a challenge," Talbot says.

Finally, cultural barriers remain, though Talbot says, "I thought bench scientists would resist the change and lab managers would support it." Instead, bench folks love losing the worst parts of their jobs, while managers fear productivity will take a painful hit during lengthy systems implementation.

"So," you may be asking, "why bother?" The payoff is in three broad areas. First, it is possible to cut costs as select jobs go away. "Primary data reviewers' jobs become much different, if not unnecessary," Talbot says. Second, lab throughput and work quality are improved. Third, ensuring regulatory compliance becomes easier and automatic. 


ILLUSTRATION BY MARK GABRENYA