At a recent gathering of pharmaceutical IT executives in New York, presentations highlighted the fact that many of the IT challenges across pharma’s value chain — from R&D to manufacturing, distribution, and sales and marketing — are similar, and mirror those of pretty much any other business. But, as any informatics director knows, there are some very crucial exceptions that set R&D’s IT needs apart from those of the rest of the company. The challenge for informatics managers, presenters said, is keeping researchers supplied with the cutting-edge methods they need in an environment that demands that all IT projects align with the company’s broader business goals.
Attendees at the Interphex Pharma-IT Summit in New York last week came from all walks of IT life, but shared very similar stories. For example, data standardization is a hot topic in bio- and cheminformatics, but it turns out to be a challenge for sales and marketing IT as well. As one executive noted during the meeting, it took his company — a global pharmaceutical firm — two years to develop and implement a data standard for compounds with different trademarks and brand names in different countries, just so that the company could calculate how much aspirin it actually sold. The delay — enough to give any IT professional a headache — wasn’t due to technical limitations, he noted. Rather, it was “the culture of independence in multiple global locations,” and the difficulty of getting buy-in from more than 100 local product managers scattered around the globe.
The need to align local goals and global goals pervades the pharmaceutical industry and, subsequently, the IT infrastructure that supports it. Meeting participants echoed the same concerns about disparate “silos” or “islands” of data, applications, and processes — whether those groups were split by discipline, by geography, or by organizational structure. In that sense, informatics managers trying to integrate silos of biological and chemical data have much in common with their colleagues implementing supply-chain-management systems.
The Bottom Line
The twin mantras of conference speakers, regardless of whether they were discussing ERP or R&D, were “return on investment” and “IT governance” — in short, quantifiable metrics that enable IT managers to prove that their projects are having a positive impact on the company’s bottom line. Whether the improvement comes in the form of more drugs in the pipeline, shorter clinical trials, reduced failure rates, or even a reduction in shipping costs, the IT department must provide measurable benefits.
One industry-wide trend, according to conference speakers, is that pharmaceutical firms are “beginning to standardize on a fewer number of IT systems,” according to Andrew Grygiel, director of life sciences for EMC’s Documentum subsidiary. This can lower costs in the long run, he noted, but in order to choose the right systems that fit “at the intersection of common elements,” IT managers must learn more about the IT requirements of the company as a whole.
Carol Ptak, vice president of manufacturing and distribution industries for PeopleSoft, said that her firm is also “seeing companies consolidate the number of systems they have so that they can buy from a single vendor.” Of course, she added, “One size doesn’t always fit all, so you still need some flexibility at the local level.” The tricky part, she noted, is achieving standardization and integration without losing flexibility. Ptak dismissed a common gripe among pharma IT managers that the “soft” benefits of new systems are difficult to quantify, making it hard to calculate the return on IT investments. “That is one of the great fallacies,” Ptak said. “If it’s not about the bottom line, then what’s it about?”
But PeopleSoft may have a different perspective on pharma IT than other firms, since the company’s involvement in the life science market is limited to marketing and distribution. In R&D, it’s very easy to calculate the “I” in ROI, but with drugs still taking 12-15 years to make it through the pipeline, “R” can be very difficult to determine.
Some informatics groups, however, are making headway on this problem. During the R&D track of the conference, a number of pharmaceutical informatics executives shared their tips on aligning the often-chaotic world of research computing with the demands of corporate-level bean counters.
Alan Wickman, director of the research informatics lab automation group at Pfizer, identified “the first step to ROI” for informatics projects — a “global workflow analysis” of the research processes that are already in place to “determine where the bottlenecks are and whether they can be addressed.” If the project is able to eliminate those bottlenecks, it can be considered a success, he said.
Barbara Tardiff, vice president of clinical informatics at Regeneron Pharmaceuticals, shared a similar strategy. She said that it was useful for her team to first identify process “triggers” that acted as criteria for handing off data from one step in the pipeline to the next in order to prioritize data processing tasks. In one case, Tardiff said, her team was building a database for animal studies on a particular compound, and tracked very carefully the types of questions that researchers were asking about the data. The next time they implemented a database for a compound, they were able to use those questions to “prospectively” develop a data structure and format that prioritized the information so that researchers could get answers to their questions in real time.
David Eiben, associate director of collaborative technology and research data management systems at Boehringer Ingelheim, said that his company is already seeing measurable ROI from a web-based collaborative software system it began rolling out in 2000. The system, based on the eRoom collaborative environment from Documentum, allows researchers working on the same project to set up an electronic “room” where data for the project is stored and shared. So far, he said around 100 eRooms have been set up across the company, and around 40 GB of data is being shared via the software. “If we didn’t have the software, that data would be on network drives across the organization,” Eiben noted. The redundancy of multiple researchers e-mailing and storing multiple versions of the same 40 GB of data would have required a 10-fold to 20-fold increase in storage requirements, he said.
Joseph Cerro, head of informatics research at Bayer, summed up the experiences of many at the conference who have tried to sell upper management on the value of their IT projects. “I’ve seen executives’ brains start to turn off when you say ‘data integration,’ but we all know it’s important,” he said. The difficulty is in translating those IT benefits into a language that management can understand and appreciate. At Bayer, that translation process has been formalized into the company’s organization, via an “R&D IT governance committee” for each of its global sites that includes representatives from the informatics team as well as at least one representative from each therapeutic area and technical unit. The mission of the governance committee is to “balance the local life sciences IT portfolio,” while the chair of each local group liaises with IT governance committees on the corporate level to ensure that its activities are in line with those of global IT — a policy that “sometimes gets in the way of what I want to do today,” but an effective one nevertheless, Cerro said.
Cerro said that bringing users, developers, and managers together to hash out IT priorities can overcome some of the challenges of calculating ROI. In one example, potential projects were plotted on a chart with one axis representing “necessity” and the other “feasibility.” Cerro conceded that “those look like qualitative terms,” but added that once you bring all the stakeholders together to determine what the appropriate metrics are to quantify those values, “you really get the conversation going.”