Gil Roth05.03.07
CTM Supply Chain Management
How to bridge planning and execution
By Gil Roth
In February, 2007, Perceptive Informatics and Tourtellotte Solutions announced an alliance to "enhance supply chain forecasting and management for bio/pharmaceutical companies." I was interested in the integration of the former’s IVRS (Interactive Voice Response System) and the latter’s tcVisualize software for supply chain modeling. To find out more about their alliance and the clinical supply chain issues they hope to solve, I spoke with Ed Tourtellotte, president, Tourtellotte Solutions, and JeanRemy (JR) Behaeghel, director of product management, Perceptive Informatics, a Parexel company. We ran an excerpted version of the article in the print edition. Here's the full-length version, which includes a Will Ferrell moment from your editor.--GYR
Contract Pharma: What are the biggest problems you see within the clinical trial supply chain?
JR Behaeghel: When you look at clinical supply logistics and supply issues, you need to connect two words together: "Planning" and "Execution."
With planning, you're trying to establish what your study will be, what the parameters are, etc. With execution, you're trying to gather the available supplies and distribute them to the sites. And what we've tried to do in this alliance is to bring those two words together. Ed's company provided the planning side, and Peceptive, with its IVRS, CTM supply, and integration technologies experience, bring the execution side.
So we're trying to bridge the gap between planning and execution. This way, you can have accurate planning and you can take information during the study and feed it back into the plan to adjust as the trial changes over time.
Ed Tourtellotte: There's a variety of pretty nasty things that occur when clinical trial supply isn't properly planned. The most basic one is that you package too much and it gets wasted. Big Pharma can count up how much of that they waste every year. It's a substantial sum.
On the other side, if you don't have enough supply, you risk losing patients during the trial. It's a risk to the patients' health, it reduces the statistical power of the sample, and it causes all kinds of mayhem with the supply team, which ends up having to perform site-to-site shipments.
If you suffer delay due to lack of supply, there's that standard take on how much a day of patent protection is worth, and how much a delay costs.
Now, what we've done is get pretty good at the planning stage. Then we took it a step further and said, "We'll take the actual data and plug it back in and marry those two words together." Now we've got a constantly changing plan, that's got constant reforecast ability, which no one's had before.
I think in the last few years there's been a push toward simulation as a strategy for supply chain management. We've taken it a step further and integrated the plan to the actual data, allowing people to reforecast every day.
CP: Do you think the globalization of trials, and the complexity attendant with that, made this level of reiterative planning a necessity?
ET: Complexity is increasing, and it's because of pressures on cost, ironically enough. Complexity usually implies more cost, but sponsors really don't have the resources to spend more.
JR: Trials never roll out the way they're supposed to. There's always something coming up. "Do we need to add a site? Do we need to add a country?" Then there are changes in how things are done in different areas of the world.
ET: If you had unlimited capacity to make, store and move drug product around, then this wouldn't be a problem. But in the real world, these things are costly, and they're really constrained. At some point, the necessity is called out by the costs, and the hard choices you have to make. If you produce and package one lot rather than another, and you have to delay another trial.
CP: What sort of client base are you pursuing? Is this system more suitable to a particular segment of the market?
ET: We are targeting any pharma that's running clinical trials and manage their supplies; it can be anyone from a global pharma to a little biotech in Cambridge.
We're looking at multi-site, Phase II/III trials. If you have all of your patients at one site -- as in early Phase II or I -- it's not as big a consideration because all the supplies are in one spot. But as soon as it starts to get complex, that's our target market. We've got a variety of ways to approach it, and it's been architectured to accommodate both large and small companies.
JR: We're more concerned with what kinds of problems these companies are facing than we are with their size. A virtual pharma can run a Phase II or III trial, and have limited supply, and that sort of company would benefit greatly from this solution. But it's really not targeted to one kind of company or another; it's a very flexible system.
ET: A large company may receive one kind of benefit: take scale, for example. If a big company is just producing oodles of a drug and wasting millions of dollars, they can save that through this modeling. And for a small company with a very expensive drug, they can't afford to waste any of the substance, so they can benefit in that way.
CP: What's the history of the collaboration between your companies?
ET: My company's been building clinical trial support systems for quite some time. In the course of building an IVR system for a large pharmaceutical company, we learned a lot about supply chain and the problems it has. We thought, "Wouldn't it be nice if you could model this?"
JR: Ed and I got introduced a couple of years ago. He was starting to show his product around. It took some time for us to see where it was going and to make sure there a good fit. In the last 18 months we've seen tremendous development. It can integrated with our existing offerings very well.
ET: We started this up in 2004, then really ramped up in 2006. We made the alliance with Parexel and really haven't looked back. It's been pretty fruitful for both of us.
CP: What do you think was more important for developing this software and service: improvements in computing power or a greater understanding of simulation and modeling as a concept?
ET: Wow! That's a good question! This is a big, powerful application, but it runs on a desktop or laptop computer, and that certainly may not have been possible a few years ago. On the other hand, I think it's also important that we learned how to put these concepts together. Things that are known individually weren't being seen together in the big picture. That's the conceptual leap we made.
CP: So it's less about the brute force of computing and more the elegance of the architecture?
ET: I'd agree with that. We're not running this on a supercomputer or anything like that. Of course, if we did, we'd great even faster results.
JR: I think it's more about the concepts and the ability to link them together, pulling different domains together. If you look at the tool and how we're using it in the partnership, you've got the forecasting, and the simulation, and the execution aspect of it. You also need to find experts who are well versed in the area of clinical supply chain management and having a deep understanding of things like IVRS. That's what makes it special. The computing power doesn't hurt; you can get results pretty quickly, so it helps the user who wants to try out different simulations. But that's more of a benefit than a driving force.
We've been tapping our own experts within the company. As the product spreads out, we'll receive more user comments and feedback that can be incorporated into the next release.
CP: Is this release considered the "1.0" version?
ET: No, it's definitely undergone prior releases. We debuted with 1.7. We put out 2.0 last year, and are on the verge of putting out 2.2, which has some really neat features, including Monte Carlo simulations and variable dosing. The product continues to evolve and advance, and there's a lot ahead.
CP: I have no idea what a Monte Carlo simulation is.
JT: [laughs]
ET: [laughs]
CP: No, seriously. Do I have to look it up?
ET: I think in the last few years there's been a push toward simulation as a strategy for supply chain. We've taken it a step further and integrated the plan to the actual data, allowing people to reforecast every day.
CP: Does this ability to reload incoming data ever create a risk of overanalyzing?
ET: Well, the new data doesn't necessarily have to change the trial. It gives you knowledge of what's going on. At the beginning of a trial, your awareness of the situation is based on your plan and your predictions, but as real data comes in, they'll probably be better than your predictions, unless you're really good, so it's going to give you an ability to take actions if you need to.
JR: Rather than stop enrollment because an issue arises that you didn't see coming, now you'll have the opportunity to slow down enrollment in one region of a trial until more supplies are available. You can be proactive, rather than reactive, and fine tune the supply chain. It gives a forward look into issues that can arise, and then implement changes in the execution of the plan.
ET: You're going to find out about the crisis eventually!
JR: And you'd rather find out two months early rather than two days after it happens.