Robert Groves, MD, and Bhavesh Shah, PharmD, comment on the role real-world evidence plays in creating cost/value analyses for psoriasis and metabolic syndrome.
Ryan Haumschild, PharmD, MS, MBA: One last question for you, Dr Groves, because you’ve been a great sport in answering these questions, how would you validate some of the findings that we’re looking at, the study that we’re trying to produce here in real-world evidence? And do you believe that this would provide a real opportunity for saving in total cost of care within the patient population?
Robert Groves, MD: Yes. I always get nervous when somebody calls me a great sport. I’m not sure exactly how to take that. But what I would tell you is that there’s a reason we do randomized, blinded, controlled trials, but they are imperfect in their own way. There’s also a way to use practice-based evidence that can be effective in helping us both learn more and come to conclusions, and also inform the kinds of placebo-controlled randomized trials that are done. Thus, without having all of the depths of nuance with the way that the information is collected, whether there are opportunities for bias to be introduced, how it is compared, what the comparison group is, is it a historical control? Is it a matched control? All of those things are critically important when you do anything other than a randomized controlled trial. And it’s because we, as human beings, are biased. And even when you talk about algorithmics, some cases of AI [artificial intelligence], I’ll pull out of this mix, but we can build biases into our algorithms that end up leading us in the wrong direction. All of us need to be very skeptical of everything. You can take that too far as well, as we’ve learned in our recent political shenanigans. But my point is this, absolutely, I think that what you’re doing will bring value. The question of how much value and how effective it will be in convincing others really depends on all of those nuances about how the data are collected, how they are used, and how reliable, ultimately, we feel they are based on those study strategies.
Ryan Haumschild, PharmD, MS, MBA: Dr Shah, I’ve got a question for you building on some of the points that were just shared. How do you utilize some of these real-world analyses of outcomes for certain patient populations, and doing that cost analysis? Do you utilize these within your organization? And some of those challenges that were brought up, how do you overcome those to produce meaningful data that are specific to your patients?
Bhavesh Shah, PharmD: Yes. I feel like even though we have access to all these data from the plan, from the medical benefit side, pharmacy benefit side, we still have challenges in cutting the data to the right levels that we need to actually make those actionable decisions. But I think that more and more, we’re getting more sophisticated in using the data and targeting patients specifically who are driving that total cost of care. One of the things we had mentioned, our targets are diabetes, hypertension, and hyperlipidemia, which are all tied to this disease, but we’re not targeting that disease, specifically. Thus, it’s kind of an oxymoron in regard to how we’re targeting these specific total cost of care population health models, that we’re really working toward caring for these patients.
I think we would definitely benefit having a better perspective on this, incorporating the disease and all of those downstream adverse effects of the disease, such as metabolic syndrome, how do these treatments really play out? And how can we be more targeted in terms of how we treat these patients? One of the biggest things that we do is academic detailing, and having pharmacists intervening in a lot of the patients who are responsible for this total cost of care and identifying essential interventions that we can make toward their diabetes, or their hypertension, or changing it to the most optimal therapy, that may be more effective in that specific patient population. Using all the tools at the IDN [integrated delivery network] level, and the population health level to drive some of the changes, using the data and academic detailing, I think can bring more value to the entire model.
This transcript has been edited for clarity.