Accounting for the realities of the marketplace when judging healthcare innovations

TURN THE PAGE

Last week, Drs. Hirsch, Phares, and Neumann put a spotlight on something that has largely been omitted (at least quantitatively) from the value assessment of healthcare innovations. In a recent Health Affairs Forefront piece, they discuss how health system barriers influence the uptake of new interventions, and how accounting for these within value assessment would promote more realistic evaluations. The list of potential systemic barriers is long and includes things like reimbursement challenges, payment complexities, policy constraints, geography, administration at specialty clinics, complex diagnostic processes, shortage of qualified providers, and complex manufacturing requirements.

Last week, we talked about the enormous societal value of direct-acting antivirals for hepatitis C. In the Forefront piece, the authors discuss that at the time of launch for these direct-acting antivirals, numerous systemic barriers hindered access and uptake, including “barriers associated with screening and diagnosis; restrictive insurance policies such as authorization criteria (for example, Medicaid and other insurers limited coverage to patients with advanced disease); and other challenges such as lack of awareness and stigma, patient’s unease with drug regimens, transportation impediments, and a shortage of treating providers in some areas.”

What are the consequences of value assessment not accounting for these systemic barriers? The authors say: “Cost-effectiveness analyses and budget-impact assessments that ignore real-world access barriers can themselves exacerbate barriers to access because they can distort coverage and pricing policies and diminish public trust.”
Budget impact models are often done around the time a new innovation enters the market. The purpose is, as the name suggests, to model the impact of adopting a new healthcare innovation on a payer’s budget. Rather than informing value or efficiency, findings from a budget impact model inform affordability considerations. As compared to cost-effectiveness analyses, a budget impact model typically has a shorter time-horizon (e.g., 1 to 5 years) and includes a different set of model inputs (e.g., size of the population, uptake of innovation, treatments displaced).

Findings from a budget impact model can inform utilization management. For example, say a new healthcare innovation is supposed to have an enormous impact on the payer’s budget (based on the population size and uptake modeled). The payer might implement utilization management strategies (e.g., prior authorization, step therapy, quantity limits, placing the innovation on a higher formulary tier) in an attempt to control the budget impact.

What Drs. Hirsch, Phares, and Neumann are pointing out is that the population size, uptake, and treatments displaced are dependent on systemic real-world barriers. If the inputs used in the budget impact model don’t account for these barriers, the analysis could overestimate the budget impact, which could lead to access restrictions that might not have been necessary if the systemic real-world barriers had been considered. Not accounting for these systemic barriers could lead to restrictive policies that are unnecessary.

Drug developers are not only carrying the risk that the drug they are studying will work and eventually get approved, but they are also carrying the risk that the drug (if it gets approved) will make it to the end user, and as this paper exposes, there are market complexities that can make that challenging. When a drug gets approved, it’s certainly a milestone for the company, but society doesn’t cut them a huge check upon approval. The company still has to get the drug to patients, and that can be complex and full of barriers—barriers associated with manufacturing; patient diagnosis, education, and transportation; provider specialty, prescribing, and administration; payer coverage, utilization management, reimbursement; and government policies (to name only a few).

The authors go on to discuss “the potential mismatch between the potential value of medications and the realities of the marketplace”. If the reward that the manufacturer receives for bringing an innovation to the market is being judged (by health technology assessment groups, payers, policy makers, government, etc.), then accounting for these systemic barriers is also essential. I have said this a lot recently, but cost-effectiveness analyses (and subsequent threshold-based prices from them) produce per-person estimates (such as a per-person annual price to meet a cost-effectiveness threshold). However, the “reward” to the manufacturer is not a per-person price, but rather a function of price, volume, and margins. These systemic barriers impact volume and thus impact the reward to the manufacturer.

I argue that modeling the total societal benefit of a drug by accounting for expected volume (and considering these systemic barriers in the volume projections) over a product’s lifecycle can incorporate these barriers and more realistically inform questions around efficiency.

SO WHAT

The prices of branded drugs are highly scrutinized, and we are now living in an era with government price negotiation during the patent period. Regulating the prices of branded drugs may seem like a straightforward solution to reduce drug spending, but the impacts of these regulations on subsequent innovation cannot be ignored. The prices we pay today incentivize tomorrow’s innovation.

In a recent USC Schaeffer Center white paper, Drs. Filson, Van Nuys, Lakdawalla, and Goldman review literature around the elasticity of pharmaceutical innovation. Elasticity is just economic talk for how one thing (in this case pharmaceutical innovation) changes in response to another (in this case US pharmaceutical revenue).

They found that “for every 10% reduction in expected U.S. revenues, pharmaceutical innovation—such as clinical trial starts or new drug approvals—is expected to ultimately fall by 2.5% to 15%.” This translates to a long-run innovation elasticity of US revenues between 0.25 and 1.5.

They explain that the elasticity of pharmaceutical innovation in the US isn’t fixed and likely varies with pricing, population size, development costs, market dynamics, etc. Although different estimates for the magnitude exist, the elasticity of pharmaceutical innovation in the US is not zero and it’s not negative. Any push to reduce revenues must also consider the tradeoff of reduced innovation.

Reduced innovation can be measured by things like clinical trial starts or new drug approvals. It is true that not all clinical trials will result in an approved drug, and thus there could be fewer clinical trials without resulting in fewer approved drugs. However, when companies are developing drugs and starting clinical trials, they do not know which ones will end up getting approved and ultimately reach patients—hence the reason for the trial.

SAD BUT TRUE

Patient out-of-pocket costs for some generic drugs are hundreds of dollars more than the actual acquisition cost of the drug. Dr. Inma Hernandez and colleagues analyzed spending on the top 50 generic drugs using claims data from Medicare Part D beneficiaries from the year 2021. They compared the total patient liability and the total reimbursement amount to the drug acquisition costs as defined by the National Average Drug Acquisition Cost (NADAC).

The mean patient liability was HIGHER than the mean acquisition cost for 5 out of 6 (83%) Plan D sponsors. And this is just the patient liability! The total reimbursement was even higher, but the fact that the patient’s out-of-pocket liability alone is higher than the acquisition cost of a generic drug should make us all furious.

I have talked a lot about imatinib’s price in previous newsletters. The authors looked specifically at imatinib and found that one Medicare Part D sponsor “reimbursed pharmacies a mean of $4880 per 30-day supply prescription, which is $4749 above the drug acquisition cost from 2021.” Looking only at the patient liability, Medicare beneficiaries paid $1,102 OUT-OF-POCKET for this 30-day prescription of generic imatinib. The patient liability alone is $972 more than the drug acquisition cost in 2021. In the case of generic imatinib that had an acquisition cost of $126 per 30-day supply at the time of the analysis, “PBMs reimbursed at an average of $3,780, or 30 times acquisition costs”.

The authors explain that “the large markups paid by Part D sponsors to pharmacies for selected generic drugs resulted in Medicare beneficiaries being liable for hundreds of dollars above product acquisition cost when using their Part D insurance to fill generic prescriptions”.

I did a quick check today and the 30-day acquisition cost for imatinib is around $78 based on NADAC or around $35 from the Mark Cuban Cost Plus Drug Company. Generic competition was successful in making the acquisition cost for imatinib fall off a cliff after the patent period, but those reductions in acquisition cost must translate to reimbursed amounts.

In yesterday’s Cost Curve, Brian Reid summarized some comments made by Dr. Mehmet Oz during his testimony last week. Related to drug pricing, Dr. Oz said that President Trump wants him to reduce drug prices, not only for the government-funded health insurance programs, but also their beneficiaries. Dr. Oz suggested international reference pricing as a way to do that (he must not have read my February 25^th newsletter). I hope he looks to research like this study from Dr. Hernandez and colleagues and realizes there are opportunities to reduce drug spending for payers and for patients without reducing incentives for innovation.