Mendix on April 9, 2013
Our recent webinar, featuring Life Science IT Innovator Jim Winburn, expounded on a few of the most pressing challenges-to-innovation that Clinical Research Organizations and life science companies in general, face in their application delivery initiatives. Our audience of technology and business professionals from various life science companies learned the importance of strategic partnerships that enable innovation, and how to overcome barriers common to this industry. Before we get into the question and answer session, a little background on our featured guest: Jim is a veteran executive and innovator in the life science industry, his career spans 30 years in information technology both as strategic consultant and IT executive, most recently as Executive Director of IT at PPD, Vice President – Clinical Architecture at PRA International, and CTO and Co-Founder of Acorn Applications LLC. John Bates, Chief Technology Officer at Progress Software, puts it best “Jim is a technology visionary, an experienced technology practitioner and architect and a wonderful person.”
Q: Are you seeing shifts in the way Life Sciences organizations need to deliver applications?
A: At the highest level, the business of clinical research development has been evolving for several years and we’re under increased pressure to become more cost effective. If you look at this business ten years ago, it was significantly paper processes, and it’s evolved over the past 10 years to significantly more electronic processes in conducting monitoring visits and trial conduct. Most of the systems that were developed for study execution were built on these paper processes, so there was the intent to automate them.
We are under exponentially increasing pressure to be more cost effective, and this pressure is coming from many directions; sponsors and regulatory pressures, but also consumers.
Q: In your opinion, what’s driving these changes in the industry?
A: For example, US healthcare cost per capita is higher than the next 34 countries by a factor of three. So you’re seeing a lot of motivation in how we execute. An example of that is something that all CROs are aware of is this concept of a strategic partnership. Those partnerships are really driving us more towards a fixed cost model, and to participate in a partnership, we also are seeing RFPs that have targets for efficiency – for instance, reducing the average duration of a monitoring visit by 30%. Again, you have to do something different and make decisions much faster, and that’s a pretty significant challenge.
Q: The tension between IT and the business exists in nearly all vertical markets, and while some tension can be healthy, it seems particularly acute in life sciences organizations. I was wondering your thoughts on why that might be.
A: This is a pretty straight forward answer. If you go back ten years ago, when we were mostly paper processes, the need for IT was very minimal. IT was provisioned as very basic organizations and their main skill, which they were required to do, was maintain regulatory compliance and validation – mostly with Article 11. These were IT organizations that were developed for a purpose – there’s nothing wrong, that’s just what they were intended to be, and that’s what we currently have. In the last ten years, we’ve seen significant changes in capabilities around technology, and now we’re in this period where the demand for change is accelerating. In this accelerated environment, the business is asking IT organizations to do something very different – something they’re really not prepared to do. It really has to do with the level of tech people’s skill at the leadership level. It’s my opinion certainly, but I think that’s actually what causes this.
Q: Continuing with that line of question, why do you believe there’s so often a disconnect between what the business needs and what IT is able to deliver?
A: The business is now on the precipice of considerable change. Purpose built IT organizations that were built and provisioned for one purpose, are now expected to deliver for something very different. Fundamentally the requisite skills are now more advanced and considerably different from what they were in the past. I think you’re seeing this – lots of IT organizations in this space are trying to develop better technical skills and better analytical skills, combined with a really solid working understanding of the business, what pressures it’s under, and how IT can help enable solutions to these problems.
Q: We hear a lot about Business Process Management tools being embraced by life science organizations as an alternative to off-the-shelf solutions or custom development. Some organizations tell us that the BPM ship has sailed, however. What are your thoughts on that front?
A: After leading enterprise architecture in two CROs, I’ve certainly formed an opinion about this. And I hate to admit it, because I certainly believed at one time that BPM was a possible solution and I’m not fully divorced of that idea. It’s significantly more complex to raise an adequate SOA platform with BPM than we once thought. Very few organizations have achieved really high levels of effectiveness with these platforms, because they are complex to develop, and you still have to connect that all together in some meaningful data model. Some conical model of how you intend to integrate systems and build solutions within BPM.
Q: I’ve heard you say that the modern clinical system is overly deterministic, and that fact is rarely called into question. Can you explain that, and what are the ramifications for the modern enterprise?
A: Let me define what I mean by deterministic. If I take the computer science definition, it’s an algorithm in which given a particular input it will always produce the same output with the underlying machine always passing through the same sequence of states. The reason these systems are deterministic is that they’re built upon those original paper processes. Although this is reasonable and logical, from the perspective of moving from a paper process to an electronic process – it doesn’t really address the process of how execution actually occurs. And something that’s very entrenched in how you approve execution is the nature of variation in how we actually conduct this work.
If you could’ve gone back in time and had more of an engineering approach, you might have realized that it would have been better to rethink the processes based on a real clear understanding of these attributes. It just wasn’t done, and it’s not unusual for a business model in this level of maturity to do such a thing. The result has been several siloed systems that had been provisioned from other business purposes. EDC is one that’s very specific to the data collection of a CRF, but the systems around management – CTMS systems, startup systems – these things really have to be examined at the core. What is the fundamental problem you’re trying to solve?
And if you’re buying systems from vendors who’ve basically provisioned their existing technology to do this, you’re not likely to have something that meets your needs. Of course I can give you the example of a CTMS built on a CRM system. Although at a very high level it looks like it may be rationale to do that, it’s very different. CRM systems are built to track and identify opportunities while we’re trying to track activity in a study with a much richer, much more multi-dimensional problem. So folks have had varying degrees of success with these systems and most have been through many iterations, costing multiple millions of dollars, and they really haven’t improved their ability to conduct a study.
It really boils down to the most fundamental thing that you’re trying to do, is to address the variation in the execution of these processes.
One of the approaches is you’re seeing the application of things like Six Sigma and other ideas that have originated in manufacturing. Generally speaking, that really doesn’t address the variation in study conduct as well. The salient thing about this is that study conduct is emergent and ephemeral in nature, not deterministic. And that’s why these processes, or these methods, will not deliver the results that research companies need. The sources of variation are very different – in manufacturing it’s typically very deterministic and stationary, in studies it’s stochastic and ephemeral in nature. So if you can get to understand what that variation at the very fundamental level is, just the recognition of this elucidates a very different approach as to how to build or develop systems that address the needs of study conduct.
Receive Mendix platform tips, tricks, and other resources straight to your inbox every two weeks.