Along with stock options and online pet supplies, the mid-to-late nineties brought an unprecedented level of hype to the concept of “knowledge management” or KM — a technology-driven approach to sharing information across an organization. Like other fads of the dot-com era, knowledge management quickly fell out of favor as implementations proved costly and ineffective. However, in the life sciences arena, it seems the concept may not be destined to go the way of the Macarena just yet. The once-passé phrase is increasingly popping up at industry trade shows and conferences, and not just in the marketing materials of vendors and consultants: Discovery informatics teams at biotech and pharmaceutical companies are resurrecting KM strategies with an eye to making them work this time around.
Why the resurgence of interest? According to some pharmaceutical executives, it’s the next logical phase for many of the data integration projects that have sprouted across the sector over the past few years. “Data integration is the flavor of the month,” said one pharmaceutical IT manager. “And as a consequence, knowledge management is on the ascendancy because once you’ve integrated the data, the next question is, “How do you make use of it?”
Others credit the downturn in the economy for increased interest in the approach. Biotech and pharmaceutical firms are being forced to do more with less under increasing competitive pressure, so getting decision-making information to the people who need it as quickly as possible is becoming ever more important. And as downsizing forces employees and their knowledge out the door, companies are looking for ways to capture that experience beforehand, preferably in digital media that can be shared, downloaded, and updated.
Finally, according to some knowledge management proponents, the technology is simply much better now than it was several years ago. Arjun Bedi, a partner in the life sciences group of Accenture, said that KM technology has “evolved” over the past few years to better merge “explicit” data, such as raw datasets, lists, and archives, with softer, “tacit” data — which he described as “what you learn as the result of an experience, or judgments that reside in people’s heads.” In addition, Bedi said, people have changed their approach to KM implementations to ensure that KM is effective. “Five or so years ago, there was a lot of hype about knowledge management, and one of the issues was that the kind of solutions people were talking about were all-encompassing, big, organization-wide, and that was really the wrong way of looking at it because you could never quantify the business value of doing that. The approaches that we’re taking now are much more practical, pragmatic, and holistic.”
The tide does seem to be turning in the life sciences, especially among large pharmaceutical firms. A glance at some recent conference presentations indicates that Novartis, Bristol-Myers Squibb, Aventis, and Lilly have all embarked on discovery-related KM projects.
The head of discovery IT at a large pharma recently told BioInform that his team was evaluating KM solutions from Accenture as well as several other consultants and vendors. “Like most large pharmas, we’ve got a lot of information in certain repositories — highly centralized, very effective. However, our documents and our reporting tend to be a lot less structured, so that information is stored in a number of places,” said the executive, who spoke under the condition of anonymity. “Our key issue is really first of all knowing what data is out there, and then getting to it, and then applying information standards and curating the data so it’s meaningful to a person who might want to look at it across with other data types, and then actually having a front end on it that brings it together so that you can browse it, store relationships, and extract knowledge from it.”
Riding the Wave
Consulting firms like Accenture, as well as bioinformatics vendors looking to expand their reach a bit, have been quick to take advantage of the KM renaissance. For example, Rescentris, a new company launched by LabBook’s former executive team, is positioning itself as more of a knowledge management provider than a bioinformatics vendor in its most recent incarnation. “We include bioinformatics, if necessary, in the knowledge management component,” said CEO Adel Mikhail in a recent interview, adding, “We are on the verge of an emerging market for knowledge management in life sciences.” Other companies, like Spotfire, are adding collaborative features to their existing software solutions, while still others, such as Ingenuity, are marketing pre-packaged “knowledgebases” of curated and annotated data that fit into a customer’s broader KM infrastructure.
By all accounts, discovery is a particularly ripe area for KM solutions right now. “One of the major issues has been this explosion of information over the past five years, and that’s been felt more so in discovery, in the front end of the R&D pipeline,” Bedi said. “And when you combine that external mass of information that’s increasing every day with a lot of the internal information that gets generated, especially on the back of the genomic revolution, you’re now in a situation where there are landfills of data but no knowledge.”
According to Michael Breggar of Deloitte Consulting, the wash of data within discovery coincided with the rise of web-based tools, leading many pharma and biotech IT groups to the mistaken assumption that as long as data was moving quickly, it was speeding time-to-market. Not so, Breggar pointed out. “After years of taking web-enabled technologies for granted, we’ve come to the rationalization that the only thing the web enables us to do is move data from point A to point B faster. It does not do the heavy lifting for us — that’s still our responsibility…Whatever time is gained by the data transfer is lost if people have to look for stuff,” he said. This is where KM comes in —or would ideally save the day.
Getting such a system in place may be more likely today than it would have been five years ago, but even the strongest KM advocates warn that the vision of “one-touch access to all the data and knowledge within an organization,” as Manuel Peitsch described it at a recent industry conference, may be a decade or more away. Peitsch, head of informatics and knowledge management at Novartis, is spearheading a global KM project that he acknowledged as ambitious, but necessary. “It may be a dream to have all in silico drug design by 2020, but the pieces will be there,” he said. The project counts among its goals nothing less than “transforming the way we discover drugs,” according to Peitsch.
The company has begun implementing several phases of the project already, starting with data integration and data management, moving up through data mining and visualization, and finishing with an access and navigation layer that puts all the company’s knowledge “no farther than three clicks away from a user,” according to Peitsch. But while the technology aspects of the project are pretty much under control, cultural barriers are proving much more difficult to overcome. Researchers were reluctant to add a few lines of metadata to assays, for example, because they had to look up the right codes for each experiment. Upon closer inspection, even those who appeared to be complying with the system turned out to be entering the same codes over and over, Peitsch said.
According to Breggar, it’s very common for cultural and social issues to be overlooked in implementation of KM systems. “Everybody tries to focus so much on the technology aspects of knowledge management, when the real issue is how do you get people to actually do it and want to do it,” he said. Breggar advocated using “peer pressure,” along with careful consideration of existing work patterns when planning the system, as the best route toward user compliance. “The real challenge is the people component, and I don’t see it done well most of the time,” he said.
The anonymous pharma discovery IT manager, who is still evaluating KM technologies, admitted that the company-wide project has so far been “a lot more difficult than anybody thought it would be,” but noted that the biggest hurdles may still lay ahead. “I think there’s quite a big change management issue coming up here that we’ve haven’t really started addressing yet: How is it really going to change people’s lives, and how do we actually get people to use the tools, no matter how great they are? Or will they just feel so swamped with information that they’re frozen, and can’t decide because there’s so much out there?”
Accenture, which has already helped implement Bristol-Myers Squibb’s SMART-IDEA knowledge management platform [BioInform 11-04-02], is counting on its experience to give it an edge as it seeks to attract large pharma KM clients. “One of our strengths is our holistic approach of not just looking at technology, but also process and people issues and performance,” Bedi said.
But the consulting firm still sees room for better technology in the area, and is developing its own “Knowledge Discovery Toolkit” that will be tailored to the needs of pharmaceutical discovery. Bedi said Accenture is currently in discussions with a number of potential clients about using the technology as the basis for their KM implementations.
Apparently having learned a bit from first-generation knowledge management projects, Accenture is taking a low-key approach to client evaluations and pilot projects, Bedi explained, by starting with specific research areas where it can easily prove the business value of its technology. “I think that’s more important than just hyping something,” he said. “Unless you can show a bottom-line impact, it becomes difficult to justify an investment in any area.” However, despite the evolution of the field over the last half decade, Accenture — along with other companies hoping to ride the swelling wave of interest in KM — still has to contend with the ghosts of yesterday’s hype.
Bedi said he’s considering coining a new term to describe the concept, “because clearly when you say ‘knowledge management,’ people say, ‘Oh, that’s the big thing with no business case.’” Any ideas for a better way of describing the new and improved KM? “I’ve been using ‘knowledge integration,’ but I’m not happy with that either,” Bedi admitted.