Clinical research professionals considering upcoming information technology investments are at a crossroads in 2010: complexity versus simplicity – which should we choose?
The decision is not at all obvious. The implications of the decision are profound: the resulting strategy will direct the spending of millions of dollars and thousands of hours, and decide the fate of dozens of vendors.
We all understand that information technology is an essential underpinning of every clinical development function. As a result, software applications have been installed for almost every function, with little ultimate regard for each other. Each individual decision has generally been justified, and paid for, by the function for which the software was written, and not primarily for how similar it is to other applications, or how well it may talk with others.
Many of us have argued for years that this “silo” approach to building and buying individual software packages, which are increasingly sophisticated servants of their intended users, is reinforcing bad organizational behavior. It enables individual micro-professions to make their data so specialized as to be unusable by others. It means multiple instances of re-entering identical data. It makes ensuring the currency of data very difficult, to say nothing of the challenge of finding the necessary linkages between silos. The answer would be, as much as possible, a single system: an interweaving of applications so transparent, based on common means of entry, storage and visualization, so that the entire enterprise breathes data as a single organism.
So on the one hand, we have a path for functionally focused, deeply featured independent software, integration and cross-talk be damned. And on the other hand, we have a path based on the belief that integration is everything, realism be damned. You can see that we cannot have both. In order to integrate effectively – to make the mammoth effort and frequent compromises worthwhile – we cannot maintain the rich depth of vertical profession-specific applications. It is too much to tie together. And vice versa. We cannot keep digging deeper for richer and richer vertical functionality and expect to tie these unique entities together usefully.
We have a magnificent example of complex simplicity in every facet of daily life today – the Internet, or more precisely, the Web. (Indeed, what better metaphor for complex simplicity than a spider’s work.) Everyone with access to some kind of Web-enabled device can easily learn how to use this vast network that enables the simplest and most sophisticated exploration, transactions, learning, discovery, entertainment and influence. It seems so simple at its face, but of course it is enormously complex, as only something which tens of thousands of people, and literally billions of dollars, can create. And besides people and money, the Web has needed repeated flashes of brilliance to invent it and sustain it and expand it. Think of the search algorithms, fiber optics, high capacity storage, wireless access, global ubiquity, and on and on. This is highly complex simplicity.
And that’s the rub. Clinical research is not a universe of a billion users justifying billions of dollars for IT innovation. How can biopharma R&D mimic what the Web did with information? We in R&D want a confluence of basic science, applied biochemistry and physiology, genomics, real patient experience, epidemiology, macro- and micro-marketing, economic forecasting, regulatory and provider/payor policymaking, and more. All of this information, together, is what we actually need to be integrated. Integration in clinical research is not about having only one contact address per investigator, or even about being able to track a physiologic parameter from mice to Phase III. We want to know whether if we invest millions of dollars in therapy X, it will help enough people at the right price to more than cover the cost of developing it. Again, this is highly complex simplicity.
One has to wonder if our industry would be ready for simple access to such a complex picture. Do we understand how to run an enterprise that knew so much? What would our flashes of brilliance be to make this flower? Outsourcing and deal-making are not the stuff of genius. What would be the R&D equivalent of search algorithms or wireless access?
And On the Other, Other Hand
Let us return to the R&D IT choice in front of us. Should we invest in the complex vision of a single integrated universe of R&D software, be it powered by SAS or Oracle or Microsoft? How far can they take us today? How long will it take to get to the point where their integrated system, having been irrevocably chosen back in 2010, will justify the commitment?
On the other hand, continuing on the path of collecting sophisticated (and expensive) vertical applications is in fact a dead end. But there is one adjustment that might make the choice of paths more distinct and dramatic. If we want an alternative to the integration nirvana that may never be reached, a better path would be to design and acquire simple vertical applications, optimized for the barest minimum of business requirements, with the leanest of function sets and the easiest of user interfaces. We would carefully but doggedly dismantle the thick infrastructure of the vertical applications we have today, by a merciless zero-sum examination of what we really need, knowing that the simpler the tool, the more likely it will be used and the more relevant the output will be. Such a collection of applications would be cheaper to acquire and easier to learn. We would worry much less about the commitment to these applications because we would intentionally plan on their obsolescence, and ruthlessly strip their purposes to the bone. If only we could find vendors willing to build and sell such things.
Remember when watching TV was really simple? There was almost nothing to watch and very little choice. Perhaps your little sister had to ground the antenna with her hands, but hey, that’s what little sisters were for. The picture resolution (a word we didn’t even know) was terrible. And the world of media was divided like the continents: TV, radio, movies, records, books, newspapers, libraries and stores were entirely separate and knew their place. Who wants to go back there again, when today all of those media and more fit in the palm of your hand? So the question is, can anyone really (and I mean really) get the world of R&D decision-making into the palm of our hands, or should we count on just turning on a TV-simple CTMS to hear the latest trial news?
The choice is a conundrum: simple complexity or complex simplicity? Which can we afford? Which do we need? Which can we make happen?