Connecting the Dots: Part D Design, the IRA, and Biosimilar Uptake

Although the biosimilar market is expected to continue growing over the next few years, adoption thus far has been somewhat slow and uneven. Recent research published in Health Affairs takes a look at how upcoming provisions of the Inflation Reduction Act (IRA) are likely to impact biosimilar coverage, concluding that both Medicare beneficiaries and the government could realize substantial savings if Part D formularies resembled those of employer-sponsored plans.

In Benefit Design and Biosimilar Coverage in Medicare Part D: Evidence and Implications From Recent Reforms, authors Luca Bertuzzi of Charles River Associates and Luca Maini, an assistant professor of Health Care Policy at Harvard Medical School, review how the dynamics of Medicare Part D benefit design influence plans’ coverage decisions over time.

Using MMIT coverage data, the authors found that Part D plans were more likely than employer-sponsored plans to prefer reference biologics to biosimilars. While Part D biosimilar coverage did increase following the reforms implemented by the Bipartisan Budget Act of 2018, Part D plan savings for biosimilars were nominal, due to high government subsidies in the catastrophic coverage phase. However, the authors expect that full implementation of the IRA will shift plans’ incentives to favor biosimilars over biologics, as government reinsurance subsidies will soon decline from 80% to 20% of drug costs.

In this Q&A, we spoke to Luca Maini about the implications of this research and his thoughts on how further benefit reforms could help us achieve lower drug costs.

How did you first get interested in exploring how the design of Part D impacts cost? 

My co-author and I are broadly interested in studying drug regulation, so we’re always looking for how price or coverage regulations affect market equilibrium. For this research, I can’t recall if we first noticed the phenomenon in the data, or if we first had a hypothesis.

But one way or the other, we realized that some of these policies that were part of the original design of Medicare Part D had some counterintuitive implications for what plans would want to cover. We used MMIT data to investigate, and sure enough, we found that essentially plans were behaving the way the program incentives encouraged them to, which is probably not how you’d want them to behave if you were designing the program from scratch.

The goal of our research is not necessarily to criticize the way Medicare Part D is set up, because obviously it’s trying to strike a balance between a lot of different competing interests. We’re just pointing out some causes and effects that might have otherwise escaped notice.

How did you end up using MMIT data for this research? 

My co-author and I started working with MMIT data back in 2019. Among the sources we looked at, MMIT had the longest historical data, going back to 2011. For this particular paper, we didn’t need to go that far back. But we do a lot of work on historical price trends and looking at the evolution of the pharmaceutical market, and MMIT’s data is very helpful for this research.

It’s also relatively easy to use and clean, so we’ve kept using it. We’ve written a number of papers now at this point using the original data we bought. There are clear markers for the channel, and then within the commercial channel we can also identify the employer segment.

Your research suggests that the IRA will have a much bigger impact on biosimilar use than the Bipartisan Budget Act did. Do you expect this will be a sharp change, or a more gradual one?

I think it will be a sharp change. The IRA is introducing four reforms to benefit design that will become operational in 2025. The first one is that it caps members’ out-of-pocket spend at $2,000 per year. It also replaces the 70% manufacturer discount in the coverage gap with a 10% manufacturer discount during initial coverage, and 20% discount after a member exceed their out-of-pocket maximum.

Most consequential is this reform: the IRA will reduce the amount paid by the government in the catastrophic phase of coverage from 80% to 20%. In our paper, we give the example that plans would save $3,385 if a patient were to use the biosimilar Zarxio instead of its reference biologic, Neupogen. That is more than three times what they save under the current benefit design.

This 60% drop in the reinsurance amount the government will provide is a huge change, and there’s no phasing it in. Plans have had time to think about what they’ll do when this happens. By and large, the incentives are fairly clear and have been known well in advance. So my best guess is that change should be relatively quick, and both 2024 and 2025 will be much more successful years for biosimilar adoption.

Once these IRA provisions go into effect in 2025, how do you expect formularies to change? 

I think formularies will increasingly forsake the originator biologics in favor of one or possibly multiple biosimilars, because as far as we can tell, the biosimilars will still be cheaper. It’s also possible that the originator biologics will compete even more aggressively on price, and will offer much larger discounts in order to secure a good position on formularies.

However, that strategy will be less powerful now than it was a few years ago. Early on, plans had an incentive to care about rebates regardless of net price. That meant that a payer might be willing to accept a slightly higher net price if it came with a large rebate. When you think about it, that is kind of silly, because the rebate shouldn’t ultimately matter; what matters is what you end up paying.

Consider the retail market. When you buy something on Amazon and Amazon crosses out the “original” price to convince you that you’re getting a great deal, that “original” price is essentially a figment of the imagination. At the end of the day, what you’re actually paying is what matters, not the perceived rebate.

Medicare Part D plans currently has these strange incentives to care about the rebate per se, but after the IRA provisions take hold, those incentives will be much smaller. Given that the size of the rebate is how a lot of originator biologics managed to keep their formulary coverage, biosimilars should have a much easier time obtaining market share.

The trend of manufacturers offering both a high WAC with rebate and low WAC options for biosimilars is so interesting. Did your research indicate if this trend impacted Medicare Part D uptake? 

We did not specifically look at that, but I think that biosimilar manufacturers started offering that pricing structure precisely because of the incentives that Part D was generating. From now on, there should be less of a need to do that.

And in a sense, the fact that manufacturers are offering the same product for essentially the same net price, but under these two different arrangements—well, it’s a red flag that there’s something weird going on. Obviously, this is a very complicated market with complicated contracts, but if I had to guess, the IRA should help to remove the need for these odd payment arrangements going forward. Hopefully, we can move towards a more straightforward, more transparent pricing structure.

The news has been flooded with IRA impact pieces, not to mention articles on the unexpectedly slow uptake of biosimilars. But I haven’t seen many pieces connecting the two. Why do you think that is? 

I definitely think payers and manufacturers are aware of what’s coming, and the implications. But the IRA is a huge reform program, and many other provisions initially stole the headlines. The industry was preoccupied with worries about the Medicare Drug Price Negotiation Program, while patient advocacy groups were happy about the out-of-pocket copay cap. Those provisions are also easier to talk about, as it’s clear what their intentions are.

For these second-order provisions, it’s more difficult to think through the consequences, which means there’s been relatively less attention paid to them. But I think they will have a significant effect on the market, and this is one positive consequence that we’re glad to highlight.

Aside from what we’ve been talking about with the IRA, what other benefit reforms should be made to incentivize cheaper drugs?

I think one necessary reform is how we handle protected classes. Medicare Part D plans have to cover “all or substantially all” drugs in six protected classes: immunosuppressants, antidepressants, antipsychotics, anticonvulsants, antiretrovirals, and antineoplastics. That last category has the biggest impact, as it covers all chemotherapy drugs used to treat cancer.

Essentially, the existence of those protected classes removes a very important bargaining chip from health plans, because they can no longer threaten the manufacturer with exclusion. Manufacturers of protected-class drugs know that plans will have to cover their products regardless of the price they set. Of course, the plans can still use utilization management tools like prior authorizations, step therapy, and high co-insurance levels to help control costs.

But a lot of these strategies are not very effective, because the demand for many of these drugs is quite inelastic. A cancer patient needs certain therapies, period, so utilization management is not as effective as exclusion. Also, these restrictions impose a huge burden on patients and providers, as all these additional bureaucratic steps delay treatment and add expense.

Some research has looked at the impact of protected class regulation on pricing, and in fact I’m currently working on another paper on this topic using MMIT data. But there’s already some existing research that suggests that drug prices are much higher because of these restrictions. If protected lass regulation results in increased prior authorization and cost sharing, then it’s not delivering the intended outcome for patients.

In an ideal world, if you were in charge, would you scrap protected class regulations altogether? 

It’s probably a good thing that I’m not in charge! I don’t believe I’d scrap them altogether, but I would curb them significantly. Even though exclusion sounds bad—as it could potentially lead to cases where patients can’t access the drug they need—it does come with a trade-off, which is that there are a lot of other drugs that are now much more easily accessible.

In some of these classes, there are a lot of drugs that have very good and close substitutes, and maybe there is no need to cover them all. The other, more subtle point is that if we allow payers to make the threat to exclude a drug, it doesn’t need to translate into an actual exclusion to be effective at lowering costs.

One of the questions I’m investigating in my current research is what is the exclusion rate of these protected drugs in commercial plans, which have no restrictions on what they can exclude?

It turns out that commercial plans are not excluding drugs at a much higher rate than Medicare plans, where exclusion is prohibited. This comes down to the art of achieving equilibrium in negotiations. If a manufacturer is threatened with exclusion, they will give payers a lower price to avoid it, and then the drug is covered at a lower price.

If we reformed how Medicare approaches protected classes, I believe the ultimate outcome would not be formularies that exclude half of these drugs, but formularies that occasionally exclude one or two of these drugs. The vast majority would still be covered, but at much lower prices.

What overall learnings did you take away on the connection between drug coverage and the impact of cost sharing on Medicare Part D beneficiaries? 

This specific study didn’t really have anything to say about utilization, but there’s a long history of research papers that study what happens to utilization when drugs are moved to different formulary tiers. Our research takes it as a given: coverage directly impacts utilization.

In a sense, the reason why our results are meaningful is that we know that coverage has a very strong effect on utilization, on which drugs patients use and how much patients spend. We must be mindful that when drug prices are higher, premiums are also higher, and cost sharing is usually rising along with them.

Clearly, how we structure benefit design can inadvertently affect plans’ coverage decisions, potentially to the detriment of patients. Rethinking benefit design can lead to considerable government savings that then trickle down to Part D beneficiaries in the form of lower cost sharing, which is a win-win scenario for everyone involved.

Learn more about MMIT’s Analytics solution to see how comprehensive payer policy and restriction data can inform your brand’s market access strategy