For three decades, US pharmaceutical companies have agreed to sell their products to Medicaid programs at the lowest price available in the market. In return, Medicaid programs cover and pay for every Food and Drug Administration (FDA)-approved product the companies sell, even when the lowest price is still excessive.
Pharmaceutical companies want out of the first part of this longstanding deal, and the administration has capitulated in proposing a gaping loophole in the Medicaid best price rules. Prices resulting from contracts between pharmaceutical companies and purchasers that link a drug’s price to its performance—what the drug industry calls “value-based purchasing” agreements—would not be available broadly to Medicaid programs. According to the proposal, a Medicaid program could only have access to the drug for that lower price if it agreed to the same value-based purchasing agreement that produced the low price in the first place. But there is no indication that drug companies would be required to even notify Medicaid programs of the existence of those agreements, never mind avail them of the terms on which they are based.
The Medicaid Best Price Rule
It’s worth remembering how we got to the balance the administration proposes to upend. The Medicaid Drug Rebate Program has been in place since 1991. By agreeing to cover every drug a company sells no matter its price, Medicaid programs give up most of their negotiating leverage. But by being guaranteed access to prices other purchasers pay—such as large health systems and drug wholesalers—Medicaid programs are assured that at least they will pay no more than the best free market price that pharmaceutical companies have shown they are willing to accept in the US. Philosophically, the agreement seeks to harmonize two conflicting goals: ensuring access to any beneficial therapy for our nation’s neediest residents while achieving affordability in a program that (inclusive of the Children’s Health Insurance Program) serves more than 70 million of our nation’s neediest individuals, half of whom are children.
The program has other rules, too. At a minimum, all drugs get discounted between 13.0 percent (for generics) up to 23.1 percent (for brand drugs) if the best price in the country is not less than those discounted amounts. Prices paid by government purchasers such as the Veteran’s Administration and Medicare Part D plans are not included in the calculations. Medicaid programs can try to obtain discounts (called supplemental rebates) by preferring some drugs over others. If companies raise their prices faster than inflation, which they do routinely in the US, although not in other high-income countries, Medicaid gets some of the money earned through that scheme returned to it.
What Is A Value-Based Purchasing For Drugs Anyway?
It is also worth considering what these “value-based purchasing” agreements the administration seeks to enable actually entail, given that the administration’s justification for creating the best price loophole is to foster them in the market. In the proposal, the contracts that would qualify as “value-based purchasing” cover a broad range of agreements, from ones that in any way link prices to existing evidence of a treatment’s effectiveness, to ones where prices are adjusted based on a treatment’s performance observed prospectively. In the proposal itself, these two types of arrangements are referred to as “evidence-based” and “outcomes-based,” respectively.
Taking these in order, it makes little sense that a price that is in some sense based on an analysis of existing clinical evidence of effectiveness should be unavailable to the Medicaid program. Rather, basing prices on evidence of effectiveness would align Medicaid program spending better with the program’s overarching goal of improving health affordably. Achieving such rationalization in drug pricing is of course the raison d’etre for the Institute for Clinical and Economic Review.
It is concerning that Medicaid would not automatically have access to such rationally determined prices. More concerning is just how easy it will be for drug companies and purchasers to declare that a contract meets this standard, thus avoiding the need to avail Medicaid of the prices associated with that contract, even if it is lower than the best price the company reports. This loophole is one reason I anticipate that large purchasers and integrated delivery networks will support the administration’s proposal, anticipating that they can achieve lower prices for themselves by leaving Medicaid programs out in the cold. They should be urged to remember that added cost, like rising water, always finds its way through the system. If Medicaid spends more on drugs, it will be able to cover fewer people, be forced to trim benefits, and will thus subsidize other shared resources less.
As for contracts depending on outcomes-based measures, they are routinely touted by the industry as the salvation for high drug prices, even though the approach has at least four foundational defects.
- Outcomes-based agreements cannot scale. The market for drugs involves hundreds of billions of dollars and trillions of transactions annually. Any change in payment that requires following every treated patient over time to see if a pre-specified outcome was achieved can only be done for a handful of treatments, at enormous cost. A recent review of Italy’s experience with these arrangements found that pursuing the approach at a national scale still produced only a “trifling refund.” The US experience has been no better, with Steve Miller, the chief clinical officer of Cigna suggesting that these agreements tend to “collapse under their own weight.” Michael Sherman, the chief medical officer of Harvard Pilgrim Health Care listed a number of challenges that impede scaling the agreements, including that they are “difficult to negotiate and carry out, many drugs don’t have readily measurable outcomes through claims data, and patients may switch plans before longer-term results can be assessed.”
- Outcomes-based agreements shift the risk and cost of drug performance assessment. A central tenet of drug development is that the company first pays to document for the FDA that its new drug provides a therapeutic benefit, and then the FDA grants it a monopoly in the market. How much it can charge is related (although somewhat imperfectly) to how large and certain the treatment’s benefits are. Outcomes agreement allow the company to command a price for the new product based on projected rather than documented benefit. In fact, going to the trouble of executing one of these agreements only makes sense if there is a concern that the gap between projected and documented benefit is large. While traditionally, a manufacturer would invest to reduce this concern by conducting more robust clinical studies prior to setting a launch price, it is apparent why the drug industry favors the outcomes contracting approach. If a high price is available for treatments based on the expected rather than documented benefit, why not take it? The drug company can spend less money evaluating its product prior to launching it and get to market and revenue sooner. If the product performs as expected, the company ends up capturing the same price as it would have anyway. As one analogy for how these contracts work in favor of the companies, it is like a car dealership charging you for the gas when you test drive the car it is trying to sell you.
- Outcomes-based agreements do not necessarily result in prices aligned with value. Contracts in which refunds are given when treatments underperform could, in theory, result in alignment of price with therapeutic value after the fact. However, there is little reason to believe this outcome will be arrived at reliably. Most of the existing agreements are confidential, but the ones that have been analyzed include fundamental defects that keep the ultimate net price well above one commensurate with the observed treatment value. An analysis of Amgen’s pricing agreement for its anti-cholesterol treatment Repatha concluded that across a range of hypothetical implementations, the price per quality-adjusted life year (QALY), which was $324,000 at the start of the contract, would at most be lowered to $314,000 per QALY if the drug underperformed. A standard benchmark for a drug price being aligned with value is that the price per QALY is less than $50,000. Novartis, in launching its gene therapy Zolgensma, has proposed reconciling payments if the treatment does not continue to help the patient through four years after administration, even though the price Novartis is charging is predicated on it working for more than 15 years. Some outcomes agreements seem to not be about value whatsoever. Merck has a contract with Cigna regarding its diabetes treatments Januvia and Janumet. One provision has the price of the treatment being reduced if measures of diabetic control improve, even though such an event would be a sign of treatment benefit.
- Outcomes-based agreements are fundamentally human subjects research without safeguards or benefits. Although outcomes contracts are often labeled “innovative,” there is nothing particularly innovative about characterizing how well treatments work by administering them to patients and then observing the outcomes. The process is referred to as therapeutic experimentation and generally is considered the riskiest type of human subjects research. Multiple safeguards exist for human beings who enroll in such research, including that participation is voluntary, design is overseen by an independent body with no financial interest in the outcome, and the analysis will likely produce useful generalizable knowledge. If the data from the study are to be used for regulatory purposes or scientific publication, the protocol and results are posted to gov. Calling a prospective collection of data by another name does not change the reality that these agreements involve subjects being evaluated to gauge a treatment’s effectiveness. But the human subjects are denied the benefits of the protections they deserve.
How The Administration Can Improve Its Proposal
The administration is to be credited for several recent well-considered efforts to tackle high drug prices. While some have succeeded and others foundered, policy advances routinely involve both fits and starts. But enabling the pharmaceutical industry to side-step its obligation to offer Medicaid programs the lowest price they otherwise accept in the marketplace is a poor decision. It’s a surprising misstep, in that it is well understood in Washington, D.C., that when the drug industry proposes an “innovation,” you should grab your wallet with both hands. The administration will likely proceed anyway; the drug industry has a powerful lobby. If so, there are some modifications that would be helpful:
- Eliminate any provision that would allow contracts based on existing evidence of effectiveness to be exempt from Medicaid best price.
- Clarify that the lowest price arising from each individual outcomes agreement is the Medicaid best price, available to all Medicaid programs without participation on the same contractual terms. That price would be the average of all prices paid under that agreement (that is, a bundled price, as the administration proposes) and would include all reconciliation payments and refunds, averaged over the units of all drugs sold under that agreement. If refunds are delayed due to long time horizons of evaluation, interest should accrue based on the statutory late payment penalty rate. Reporting should be by contract and by time period.
- Articulate that outcomes-based contracts in which patient outcomes are tracked for the purpose of payment reconciliation are human subjects research. The protocols should be publicly posted before the contract commences, the design of the intervention and outcome assessment should be overseen by an independent review body, the patients treated under them should be given informed consent and an option to not participate, and the resulting data should be posted to gov in a timely fashion.
- Indicate that if any Medicaid program wishes to engage in any available market contract, it should be able to, under the same terms.
Even though there is limited merit to enabling value-based purchasing approaches, and it is an open question whether there truly is an issue with Medicaid best price reporting under them, these changes would, at least, achieve two objectives. They would ensure that these agreements will produce useful insights regarding the treatments covered by them and that resulting savings will flow to Medicaid.
Author’s Note
Dr. Bach is the director of the Drug Pricing Lab at Memorial Sloan Kettering Cancer Center in New York City. He discloses receiving speaking fees from entities across the health care sector, consulting fees and stock from GRAIL and EQRx, and serving as a director of oncology analytics.