GLP-1 Usage Driving Health Care Spending, Real-World Implications: Ben Urick, PharmD, PhD

Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are significant contributors to health care spending, contrary to the belief that their presence would decrease spending per patient after a couple of years, Ben Urick, PharmD, PhD, the senior director of health outcomes at Prime Therapeutics, said in an interview with The American Journal of Managed Care® (AJMC^®).

In this Q&A, Urick discusses the real-world applications of GLP-1 therapies driving health care costs.

This transcript has been lightly edited for clarity.

AJMC: GLP-1 therapies for obesity are driving major spending concerns—how would you summarize the current state of real-world evidence on their clinical and economic value?

Urick: When we think about the clinical and economic value of GLP-1s, I would say that from the clinical trials lens, it’s very positive. If you're looking at this through an obesity and diabetes lens, again, both show that these products are highly effective. But when we think about the real-world evidence, it is a little bit mixed when it comes to the spending side, but still very positive when it comes to the clinical side. However, within the spending side, you really have to think about who within that population has obesity. Are you looking at obesity only or a mixed population of members with obesity and/or diabetes?

In that mixed population, you tend to see lower increases in medical spending and some savings, potentially in that population. We know these are very good drugs for diabetes. They're also very good drugs for supporting weight loss. When you combine the 2 together, obviously, there are some synergistic effects there, right? If you look at the obesity-only population, that population that does not have diabetes, we tend to see a bit more increase in medical spending over time, and that's really due to the different clinical picture within that group.

But if you combine the pictures together, because you have different populations and different study designs, it just looks a bit more mixed on the economic side.

AJMC: We're seeing seemingly contradictory findings across real-world studies—what are the biggest drivers of that variation, and how should decision-makers interpret those differences?

Urick: One of the main drivers of contradictory findings, and we've shown this with our data, is whether or not you include members with diabetes in that obesity population. It is entirely reasonable to do so; if I think about this from a pharmacoepidemiology perspective, looking at this through the lens of users of a drug is an entirely appropriate lens to view this. If [I am] also thinking about this from a benefit decision-making perspective: the obesity-only population is really the most important population to look at, because if you have diabetes, you can get access to these products through the pharmacy benefit for diabetes, regardless. Looking at the obesity-only population is really where you get to that marginal population that is of most interest and importance there.

Those contradictory findings are really driven in large part by your patient selection, obesity only vs obesity plus diabetes. And then, within that, some of the real-world evidence has further selected the population to look at only those members. For example, who remained adherent, or only those members who are on a high-potency product or reach the max dose of a high-potency product and remain adherent. When you start selecting within that population, you also tend to see different effects in the population as a whole. Overall, in a full population with obesity only, you tend to see different things than within the population that also has a mix of diabetes and then requires adherence, et cetera. A lot of that does come down to those aspects of study design.

AJMC: From a payer perspective, how can managed care organizations confidently use imperfect or evolving real-world evidence to guide GLP-1 coverage decisions?

Urick: If you're looking at a piece of evidence and you're a managed organization, you really want to understand your population; with who's within your groups, you want to compare that to the population that's included in your study. Then, within that, you want to compare it to the decisions that you're trying to make. Ideally, you would have a piece of evidence that is well-fitted to your population and is relevant to your benefit decision-making. Some of the evidence is better fitted than others, and some of the evidence that is a bit more positive in terms of medical spending reductions is actually less well fitted to the population and the benefit decision-making that you want to consider.

Really understanding the population included in the study, making a comparison between the 2, and then from there, considering for yourself, is this or [is this] not relevant?

AJMC: Given the high upfront cost of GLP-1s, what does the evidence suggest about their potential to offset long-term medical spending, and where are the biggest uncertainties?

Urick: There are 2 pieces of this. You can look at economic analyses like the Institute for Clinical and Economic Review, for example, which has published some economic analyses that say there should be major returns on the GLP-1 investment from medical spending. You can look at adjusting for age and gender, or populations that spend more on obesity than populations that do not. In theory, if you can take the population that has obesity, give them highly effective therapies like the GLP-1 products, and help them lose weight and become non-obese. At some point in time, you should see that reduction in medical spending. You will eventually bend that cost curve if you can take somebody who is obese and bring them down to a normal weight.

But what we actually see from the real-world evidence is a lot more complicated. Within our data, over the first 3 years, we do not see a reduction in medical spending among members with obesity who initiated GLP-1 products. As a matter of fact, we see that at 3 years, medical spending remains about $900 more among the population that initiated these products.

We do expect at some point in time for that to come back down. But there's some additional complexity around members receiving needed services that they could not receive before they started this product. For example, they start the product, and they get a hip and knee replacement, which is great. That's a very good thing. If obesity prevents you from getting a hip or a knee replacement, you can lose that weight. You are now eligible for surgery. That's a fantastic thing for you. But clinically, those surgeries are also expensive.

It’s that and other things like it that tend to offset some of the reductions in medical spending that we believe are also there through improvements in cardiovascular health, for example. It's a more complicated picture than economic models might show.

AJMC: When evaluating GLP-1 therapies, which outcomes should payers prioritize—weight loss, comorbidity reduction, or total cost of care, and why?

Urick: It really comes down to individual payers’ decision-making. For our data, we are just providing evidence when it comes to payers. We're talking about health plans, pharmacy benefit managers, and employer groups. And for employer groups, it really depends on their preference. Some employer groups may be looking at this from a medical spending reduction perspective. Some employer groups may be looking at this as, “How can I create a benefit that provides the best care for my employees and their dependents?” And depending on how you weigh those 2 factors, you may come up with different decisions. It’s not our role for the data to say that; we're just providing evidence, and then ultimately, the payers can make the decision for themselves.