I had a rare and refreshing experience yesterday afternoon. In meeting with a startup team from a big pharma company, the subject of source data verification came up. We like to use a digital pen CRF as source data, obviating the need for source data verification, one of the most inefficient steps in the research process. Why waste a third of a study budget comparing transcribed values to source data when you can eliminate transcription?
The reason for such large-scale waste on SDV is that change is difficult for everybody—CROs, sponsors and sites. In the case of the study under discussion, our understanding was that the site’s standard practice was to record data first in their EMR system. We checked and the EMR system couldn’t grant selective access to study-specific data, therefore we couldn’t use CRFs as source data. We also couldn’t access data to do remote monitoring. We were stuck with verifying transcriptions from the EMR system and doing tedious, costly onsite data verification. We couldn’t be efficient because it was a given that we couldn’t ask the site to change its workflow and the site wouldn’t change if asked.
When we meet to plan clinical studies, the givens always have a seat at the table. In this case, it was a given that we couldn’t interfere with how a site works. I recall a similar issue coming up more than two decades ago when we started using EDC (RDE at that time for those of you old enough): most site staff barely knew how to use computers so it was a stretch to ask them to enter data at the computer keyboard. We chose sites then based on criteria such as therapeutic expertise and patient access and we still do. The difference is that back then, we couldn’t do EDC because it was a given that site staff couldn’t deal with PCs and enter data. Now everybody can do EDC, but it’s often a given that we can’t use CRFs as source data because of site workflow.
What made yesterday’s meeting so refreshing was that things were different. This time, the sponsor said, let’s have the sites change their workflow a bit. We hadn’t even considered that possibility. It was the sponsor who was willing to break with routine and ask a site to make changes. I can’t tell you how different this is from dealing with big pharma companies that specify performing 100% SDV onsite for all of their studies and aren’t interested in hearing about alternative data-capture and monitoring methods.
Result: the digital pen will be capturing data as source (our database receives an electronic copy immediately and the paper CRF is left behind in the patient’s file). A report will be on the sponsor’s desk within minutes. The sponsor will spend a lot less on monitoring because of reduced SDV. No transcription errors will get past a bleary-eyed monitor and into the study database. Perhaps best of all, rapid access to actionable data will enable timely decision making that keeps the study on track. A bit of change at the site will provide an enormous benefit to the sponsor who was bold enough to ask.
We all acknowledge a crisis in pharma productivity that demands bold thinking and big changes. We need fresh thinking about our whole approach to clinical development with three phases and white space and p-values and delays that have been givens for decades. We need to think big. At the same time, we shouldn’t forget to think small. This sponsor wasn’t willing to accept familiar routines as an excuse for wasting resources and asked for a small change at the site. The lesson for us all is to look for potential improvements in the areas where we don’t even think about change because well, it’s a given that that’s how things are done.