Buddy Holly, and then Linda Ronstadt, sang enthusiastically that “It’s So Easy” (in their case, to fall in love). Lately, leading EDC vendors have been singing the same song about implementing their technologies. Unfortunately, both of these songs are very misleading.
The pitfalls of falling in love I presumably do not need to tell you about. The potential pitfalls in implementing EDC, on the other hand, apparently need some re-iteration. Listening to EDC vendors today, one would conclude that EDC is as easy to adopt as a mildly complex desktop tool like MS Project¢â. This is the worst kind of déjà vu, reminding me of the mid-1990’s, when EDC technology and vendor support was very immature, and yet EDC vendors would routinely tell their prospective customers that EDC was “simply an electronic CRF”, that their software interface was “intuitive”, that study startup would accelerate, and that study closeout would be “overnight”. The failure of EDC in the 1990’s to deliver these features, at least as perceived by pharma customers, was a huge contributor to the derailment of EDC progress in that decade. It would be a great disservice to pharma clinical development if the EDC train derailed again.
Yes a lot has happened in the past decade. Technology and vendors are stronger. Sponsors know what the abbreviation EDC means, and are modestly more aware of what EDC implies. But the fundamental change which EDC triggers, which indeed can and should extend far beyond making the CRF computerized, is only recently being realized, conceptualized, and the implications understood.
I am not sure why “easy EDC” is becoming such a popular marketing pitch. Are the vendors frustrated at the pace of EDC adoption? Are their investors impatient? This would be ironic, since EDC adoption is finally taking off. Do they think this will overcome a major selling obstacle? This too would be ironic, because misleading the customer is the best way to create a selling obstacle, as the ¡®90s proved.
Clearly some of the very small EDC vendors are using this approach to try and break into the market, differentiate themselves from more capable vendors, and try to appeal to biopharmas with low budget or low tolerance for complex reality. This is understandable as a time-honored marketing tactic. What is more distressing is to see market leaders — capable vendors working on a large scale — who now want to “move to the next level” by assuring customers EDC is not as hard as they may think.
When vendors say that implementing EDC can or should be easy, they are thinking about it from their own perspective, naturally. So they would say that these things should not be complicated:
-learning the tool (by which they mean training the sites and monitors in the software interface)
-following a checklist of technical steps to set up a study, or derive an report
-setting some software switches (yes, query writing has gotten easier)
-writing a SAS export routine
-basic aspects of how staff roles change, or that the time for study startup has to be planned more carefully.
These are all good things to learn and maybe one can learn them quickly.
But implementing EDC is much more about things that no seminar can “transfer knowledge” about:
-governance of the EDC change effort (who’s in charge, who has the money, how conflicting interests among CDM, clinical operations, IT and others get accommodated, etc.)
-the business model and budgeting (there are so many ways to “buy” or “rent” EDC technology and support services, including these days “EDC monitors”, plus there is the question of who should be paying: is this coming from the study budget, is it a capital expense, etc.)
-support issues across the board, for the site, monitor, patient (in the case of ePRO), IT, study manager, and more
-examining the cleaning processes in your company and long-entrenched roles (this is so diverse than no standard seminar can begin to address it specifically)
-raising “site consciousness” (responsiveness to site needs, workflow, usability) which is nearly non-existent in most companies
-”change management” in the classic sense of attitude changes and acceptance of new roles
-and coping with the tough personnel problems that come from change which is not (never will be) universally embraced.
Let’s look at just one example in more detail, an example that points the diversity and detail that has to be coped with in implementing EDC – data cleaning choices. Even though we are all in the same business, and trials are nominally the same from a data processing standpoint, sponsors vary enormously in their policies and approaches to data cleaning. For instance, sponsors vary by how they define the role of the monitor in data cleaning: it runs the gamut from on-site data scrubber to completely hands off. Sponsors using EDC, therefore, vary accordingly in what they allow their monitors to do with the EDC software – some take advantage of the ability of monitors to review data online before a site visit, while others prohibit it! Sponsors vary notoriously in their source data verification policies (100% of datapoints, or a subset). From trial to trial, even within the same sponsor, teams have different styles of querying the data, from complex queries to simple. And teams differ in who queries the data (data managers, monitors, statisticians, physicians) and when (early and often, or late in the game). Sponsors also vary widely in how standards (for CRF design, data structures, common questions) are applied – by whom, how rigorously, and why.
With all this variation, it is not enough for any third party to say either “this is the way we do these things with our tool” or “decide all that later, let’s get started”. For a study team who is conscientious, or rigorous, or conservative (pick your adjective), this drive for simplicity is misleading and sure to ultimately generate disappointment.
It’s not that there is anything wrong with vendor-driven quickie seminars about EDC-induced process change. Just getting vendors to admit there is process change around EDC adoption, and that there is more for sponsors to consider than purchasing software, is definitely progress. The value added in such seminars, assuming quality content, is unquestionable. It is the assertion of sufficiency, that all the issues will be covered, that EDC has been “made too complicated”, which is so dangerous, and simply inaccurate.
Why does this matter? Why can’t EDC be simple? The answer has at least two levels:
– The Clinical Development process overall is not simple now. Perhaps it could be, and EDC will make it simpler, but you have to get from here to there, and understand the path. Ample experience at sponsors over the past decade shows us that teams and staff will stumble, duplicate work, maintain unnecessary work, and even unwittingly resista change without careful re-design of roles and processes, participation of those needing to change, agreement on how/when to change, and some trial and error.
– Secondly, over-simplification of EDC matters because without careful preparation, the resulting unnecessary inefficiencies, delays (or lack of time compression), and other expected benefits will result in widespread disappointment at the failure to deliver the promised impact (speed, reduced resources, higher quality with less effort, etc.). And this is deadly. Because biopharmas cannot afford to sit out another decade without widespread clinical development innovation, of which EDC is only one piece among many.
It’s easy to fall in love, and it’s easy to believe that putting a CRF page up on a computer screen should make clinical trials equally simple. Some day it will be, but it won’t be tomorrow. Saying it doesn’t make it so, only hard work does.