Evaluating the Cost-Effectiveness of Medications for Type 2 Diabetes: An Interview With Jeffrey D. Dunn, PharmD, MBA

January 25, 2020

Jeffrey D. Dunn, PharmD, MBA, has served as vice president of clinical strategy and programs and industry relations at MagellanRx, a national pharmacy benefit manager (PBM); cofounder, senior vice president, chief clinical officer, and board member of VRx, a regional PBM; and pharmacy director of SelectHealth, part of Intermountain Healthcare. Dunn was instrumental in opening and running the VRx retail pharmacy in downtown Salt Lake City, Utah. He is a board member of Care Pharmacies Cooperative, a chain of 100 retail pharmacies.

An editor from The American Journal of Managed Care® recently conducted a question-and-answer session with Dunn regarding the recent Institute for Clinical and Economic Review report on the cost-effectiveness of oral semaglutide.

The American Journal of Managed Care (AJMC


): How does the presence of comorbidities affect the management of type 2 diabetes? How can managed care organizations encourage a more holistic approach to managing type 2 diabetes and its comorbidities?

Jeffrey D. Dunn, PharmD, MBA: Patients with diabetes have many comorbidities, including obesity, cardiovascular disease, and renal insufficiency, which should all be considered when making prescribing decisions or management decisions. The challenge is, how do you align what the providers are doing and what the payer is doing? Payers are not prescribing; we’re setting policies for coverage and reimbursement. Also, we generally have a formulary. We can’t have multiple formularies based on comorbidities or other variables—that would just be impossible to do.

Many aspects of diabetes care are managed through care management. I think there’s an obligation for payers to get involved and help improve the outcomes of their members, and that includes initiatives to provide education, help improve adherence, and increase patient engagement, for example. But it’s difficult to reconcile what happens at a population level and what happens at the level of the individual patients.

Also, for diabetes drugs that are coming to market, the FDA is now requiring that [pharmaceutical companies] have outcomes studies. Thus, it’s going to be a little bit more difficult to differentiate between products. Having said that, we do practice evidence-based medicine, and so, all things being equal, we would like to grant preferred formulary status to those drugs that have outcomes data. At the same time, a lot of this is considered class effect. Because the FDA is requiring outcomes data for new drugs, there are a lot more similarities than dissimilarities within categories.

It’s difficult to factor in comorbidities from a formulary perspective. Factoring in comorbidities needs to be done at the prescriber level, and then payers can get involved in care management and case management with patients who need additional help navigating their health status.



: How do managed care organizations (MCOs) and pharmacy benefit managers (PBMs) view and use reports from the Institute for Clinical and Economic Review (ICER)?

Dunn: We’re very familiar with ICER, and there’s a lot more emphasis on this organization based on the large number of specialty medications entering the space, the price tags associated with them, and the need for figuring out ways that we can better afford these medications moving forward.

Historically, we have always taken into account value and cost-effectiveness when developing our formulary. An example of that is: If we have 4 DPP-4s [dipeptidyl peptidase 4 inhibitors] and we don’t have head-to-head data, what we try to do is take studies with similar designs and look at the delta between drug and placebo for efficacy, divide into that net cost, and come up with a cost per outcome. And that’s a way for us to do some indirect comparisons.

So we’ve kind of always been supportive of the process, and right now, ICER is the entity that is getting the most attention. But from the payer perspective, we don’t have a formal process for using ICER [reports] in our P&T [pharmacy and therapeutics] committees and other review processes. We also don’t really have a lot of insight into their methodology.

If a drug is considered cost-effective by ICER, that doesn’t mean a lot to us; it doesn’t resonate a lot. However, if a drug gets a negative recommendation from ICER, then we pay attention to that and use that. For example, we can use that data to either negotiate better pricing or confirm a certain position on our formulary.

The concept of value or cost-effectiveness is important, and that’s what we take into account. ICER reports specifically do not necessarily help us make formulary decisions just yet. Moving forward, we need to figure out how to do a better job of combining clinical and financial data to make the most cost-effective decision. That also requires buy-in from guidelines and physicians, to continue to do that and pay attention to the pricing of medications.



: What was your initial view of oral semaglutide in terms of efficacy and overall cost-effectiveness versus other medications for type 2 diabetes?

Dunn: Anecdotally, we would probably consider oral semaglutide similar to the other GLP-1 [glucagon-like peptide-1] receptor agonists in terms of efficacy and safety. The difference is the route of administration, which is not technically 1 of the main value drivers for a payer. We tend to focus more on comparative efficacy, comparative safety, and comparative cost. Route of administration is not something that we will necessarily be supportive of paying a premium price for. There may be certain patients who prefer a drug that is administered orally. I understand that—needles have historically been an issue with insulin. However, we have a lot of options in the GLP-1 space: We have drugs with weekly dosing and drugs with daily dosing, and it’s not difficult to do these types of injections. I think oral semaglutide fills a particular unmet need, but clinically, I think the perception is that it’s not clinically differentiated [from the other GLP-1 receptor agonists], and in this case, the price is going to be very important.

As far as other classes, the way the guidelines are set up, they are set up by MOA [mechanism of action], and the way pharmaceutical companies do contracting, they contract by MOA. When we’re contracting and making formulary decisions, we’re doing it in the DPP-4 space, we’re doing it in the SGLT-2 [sodium-glucose cotransporter-2] space, and we’re doing it in the GLP-1 space, for example. We’re not yet comparing a drug of 1 MOA to another drug of another MOA, in terms of formulary or contracting. For example, after metformin, we’re not stepping a DPP-4 behind an SGLT-2. That’s not supported by the guidelines right now, so we’re not doing that.

Having said that, I think that perceptually, oral semaglutide and the GLP-1s in general are considered 1 of the more cost-effective classes or MOAs that we have versus some of the other oral options. Again, though, a lot of that is going to be left up to the provider, and currently and in the near future, we’re only managing within subcategory or MOA, until something changes.



: How has your view of oral semaglutide changed after the release of the ICER report?

Dunn: It hasn’t fundamentally changed, and the reason is that the ICER report didn’t find a difference, really, between liraglutide and semaglutide. So, it doesn’t necessarily help [payers] when 1 of the outcomes is that semaglutide is cost-effective, but maybe it’s not as cost-effective as an SGLT-2. What would be more helpful is saying that 1 GLP-1 is more cost-effective than another, and that’s where something could be addressed within the auspices of guidelines or contracting. So again, based on my understanding of the ICER report and how the comparisons were done, the results are not helpful from a formulary perspective.



: What are some challenges to ICER’s methodology in evaluating cost-effectiveness?

Dunn: Most payers are not spending a lot of time digging into ICER reports. They’re looking at what ICER says. Is a drug cost-effective or not? They are not digging into subpopulations and inclusion/exclusion criteria. I don’t think most payers are, though there may be certain payers [who do this].

As far as the methodology, payers are familiar with the quality-adjusted life-year, but it’s not something that we talk about in P&T, for example, because that would be lost in a clinical conversation in a P&T committee. Also, I’m not sure if a $50,000 per quality-adjusted life-year is an appropriate threshold anymore. Plus, it doesn’t help if you have, let’s say, 4 drugs that have the same MOA, and they’re all considered cost-effective; they all meet the threshold of $50,000 per quality-adjusted life-year. That doesn’t help us make a formulary decision or help us make comparisons and decide which drug is most cost-effective.

Overall, ICER is important, and the concept of evaluating cost-effectiveness is important. We need to do a better job of driving toward value and not chasing rebates. And prescribers need to be more aware of relative cost. I think we’re going to move toward using this type of approach to inform decisions, meaning that we’ll use more exclusions; we’ll do different things based on these types of analyses. At the same time, to be helpful from the formulary perspective, the analyses need to be more specific in how they compare drugs and what they report.



: The ICER report compared drugs of different classes—a GLP-1 receptor agonist (semaglutide) and an SGLT-2 inhibitor (empagliflozin). In your opinion, should a drug’s mechanism of action be taken into account when making comparisons?

Dunn: I would say yes, it should. Mechanism of action informs a lot of things right now. Treatment guidelines are set up by MOA; right now, the guidelines say [to] start with metformin, and [after that], match the right drug to the right patient type, based on what’s going on with that particular patient. Guidelines are usually set up by class, and contracting is set up within class, so MOA is important.

The other issue, though, is that we’re not preferring certain classes over other classes, for the same 2 reasons I mentioned: Guidelines aren’t doing that, and contracting is not doing that. Thus, we’re not preferring a GLP-1 ahead of an SGLT-2, for example.

What we really need, which is going to be really hard to do, is for ICER or another entity to show us that 1 specific drug of an MOA is more cost-effective than another drug of the same MOA. That would be more informative than saying an SGLT-2 is more cost-effective than a DPP-4. That doesn’t help us in terms of what we can do right now from a formulary perspective.

If we were to limit access to certain classes of medication because other classes are deemed more cost-effective, it would be interesting to see how the providers would react, because they want access to different MOAs based on what’s going on with their particular patients. And again, the current guidelines are set up that way. So until things change, yes, the MOA is important—that’s how we’re subdividing our formularies.