In part 4 of this Exclusive Interview, Dr. Swapnil N. Rajpathak talks with Diabetes in Control Publisher Steve Freed during the ADA 2017 Scientific Sessions in San Diego, CA about his interest in the comparative analysis of various drug classes and their real-world implications.
Swapnil N. Rajpathak, MD, MPH, DrPH is Executive Director for the Center for Observational and Real World Evidence at Merck.
Transcript of this video segment:
Steve: So Merck has a huge pipeline. I know this is not the only thing that you work on. What are you looking into the future… First of all, let me ask you the question. While you are here, have you seen heard or expect to hear some earth shattering news? What are you looking for?
Swapnil: So steve, just to again mention my role: I am more focused on the real-world evidence. So generating evidence based on observational data that once drugs are available in the market, how are they contributing to health of patients with diabetes? So I think there are other experts in my organization that can specifically talk to the pipeline and what new things will be more exciting. But from my perspective let me say that I have seen big change in or rather a bigger interest in having things like real-world evidence of observational data, maybe comparative effectiveness research, comparing different drugs. And I saw several abstracts here. Now that we have several drug classes in diabetes and they have been utilized for a relatively good enough period of time, we have the opportunity to compare them with each other to understand what outcomes are better depending on the drug classes or within the drug class itself. Now, obviously real-world evidence and observational data have its own limitations. They are not randomized studies. They have issues of confounding, so interpretations of these observational studies should definitely be done by the methodological rigour of individual studies. But just for the field, there is availability of tons of data and there is also newer methods that are coming that actually help us address some of the real-world comparative questions in respect to diabetes treatments.
Steve: How many patients were involved in the real-world study?
Swapnil: So every study we have done is different. The one that I was referring to before was done in a market scan database. It’s a large database. We had approximately, I believe, 300,000 patients in each category of the drug class and what we did was we looked at the combination treatments as well, but it was more focused on the whether they had DPP4s or whether they had SUs. Now when you translate that into the enhanced data, obviously we are projecting it to national estimate, so the numbers will be very large on patients that are on DPP4s currently or on SUs currently. But the model itself, because the numbers were large enough, to run a predicated model, the model has its own limitations as well. I mean no sample size is infinite. So there is always a restriction of the sample size itself and then of course how our assumptions going into the model are also limitation of the study. But it’s a large enough sample size that we can drive some meaningful results. Of course, with considering the limitations of the study.