IT and Pharmaceutical Data: Finding Needles in Haystacks

Knowledge management is still an elusive concept for many CIOs, but not in the pharmaceutical industry, where collaboration and high-speed data discovery is key to future profit. Daniel E. Klingler, Ph.D., as CIO and senior vice president of Information and Knowledge Management for the Worldwide Medicines Group at Bristol-Myers Squibb, oversees all of the business processes that are part of the company’s pharmaceutical business. Klingler has responsibility for developing, delivering and supporting all of the applications that are used within the pharmaceutical area, including scientific R&D. CIO Insight spoke with him recently about the role of IT in the drug business—and how good project management rules can help speed results.

In the pharmaceutical industry, time is money—and the race to future profit will be won by companies that can use IT to sift through data faster, and shorten time to market. In that sense, pharma is leading the way when it comes to thinking up new ways to use KM to bolster the bottom line. How do you see your role in this arena?

The businesses challenges we face in pharmaceuticals are becoming more numerous, more complex and much tougher. When I came to Bristol-Myers about four and a half years ago, I kept hammering home the point that the pharma business today demanded three things: changes in the way people worked, changes in the processes that they used, and changes in the technology needed to support it all.

As a result, we’ve launched a fairly large number of large and complex business change initiatives. I don’t want to call them information technology initiatives because they’re really business change initiatives, integrating people, integrating data, integrating processes across geographies and assumptions.

Historically, most people in IT—unless they came out of formal engineering backgrounds—tend to learn their project and program management skills on the job. There’s still a fair amount of management-by-the-seat-of-your-pants going on in that regard, when people get thrown into big projects.

When I first came here, I initiated formal, quarterly portfolio reviews of all of the major projects that are in our IT portfolio. In reviewing these, we spent an awful lot of time assessing how our projects were being managed, what risks are involved and the challenges that any project faces. We also ask ourselves what are we doing to minimize those risks, and so forth.

About two years ago during one of these portfolio reviews, there was a groundswell of complaint from the staff, that they were being constantly moved around on new projects. We’d have large groups of people being brought together to work on projects, and we lacked a common vocabulary, we lacked a common framework for working together. We somehow needed a grounding in project management that we all could share and identify with, and that we could all put into place.

So there was an appeal, in essence, from the staff. There was a recognition on the part of my management team that if we were going to deliver some of these large and influential initiatives, whether in R&D, manufacturing, or sales and marketing, that we needed to have more sophisticated and standardized IT project management process tools.

Consider the drug discovery area, where the ideas for potential new drugs come forward. We have a large number of biologists and chemists who work to take ideas for new drugs from the idea stage to at least the point where we can evaluate whether we think those potential molecules are going to be safe and effective in humans and, thus, could at some point go into human testing.

Historically, drug discovery has been kind of a cottage industry. Lots and lots of people and lots and lots of labs are doing lots and lots of different things, generating a lot of information in notebooks. But there are so many different kinds of information that are used to develop an effective drug, that nobody has ever really had a view to any of that.

The biologists see their own data, the chemists see their own chemical tests. People who are involved in toxicity testing see their own data. People who are involved in the pharmacokinetic testing see their own data. And there hasn’t, in essence, been any way of integrating or aggregating that information so that effective decisions could be made about which compounds to bring forward and which not.

Well, that kind of cottage-industry approach was okay when we had people working in labs developing maybe one new compound a week. But with the massive sort of high-throughput chemical generation and screening technologies that are in place right now, we can craft 10,000 new molecules a week, 100,000 new molecules a month, and the volume of data is just overwhelming. So we put a set of technologies in place working very, very closely with our colleagues in drug discovery to standardize a number of aspects of the drug discovery process. We wanted to make sure that the people who are working on that process understand their roles, and then integrate all of that data. Anyone who is involved in the process has full visibility to all of the data necessary to support the development of these potential drugs.

?So whether it’s a chemist, whether it’s a biologist, whether it’s a laboratory manager, whether it’s a clinician, they all have access to all of this data fully integrated. So when someone says, “I need to know about the performance of a particular molecule,” all of that information is aggregated and there are decision-making tools that support decisions about whether to take that potential drug into further testing.

You could call this stuff knowledge management. In essence, it supports the way that drug discovery as a discipline is done, which is about managing knowledge. It’s about having a full set of analytical capabilities in place so that we can look at this information appropriately.

Increasingly, IT is becoming mission-critical in drug discovery.

Well, that’s certainly true in R&D. At the end of a research and development project, the outcome of any project is really information, and what we do in R&D is process and sift through and analyze massive amounts of information.

Also, the genomics revolution has been on the one hand a real boon to our business but, on the other hand, it has been a tremendous challenge because what used to be a kind of cottage industry has become highly systematized simply because of the volumes of data that we are generating. One of our challenges is to make sure we’re building data warehouses and not data landfills. Given the volume of information that we’re having to deal with right now, you could not do research and development in this industry without massive investments in information and technology.

Latest Articles