Hostname: page-component-7b7567c8c-wxvwp Total loading time: 0 Render date: 2026-11-15T10:19:43.960Z Has data issue: false hasContentIssue false
  • English
  • Français

Can Outcomes-Based Pharmaceutical Contracts Reduce Drug Prices in the US? A Mixed Methods Assessment

Published online by Cambridge University Press:  01 January 2021

Elizabeth Seeley,
Susan Chimonas and
Aaron S. Kesselheim
Rights & Permissions [Opens in a new window]

Abstract

To improve the value of pharmaceutical spending, some manufacturers and payers have introduced outcomes-based contracts, where rebates are tied to specified outcomes. We reviewed the literature and interviewed key experts to assess these contracts' potential to slow pharmaceutical spending. We found that while outcomes-based contracts are increasingly common in the US, they are still limited by multiple factors — including the lack of meaningful outcomes data. Moreover, there is no evidence to date that they slow pharmaceutical spending or increase access. Experts favored having CMS test and rigorously evaluate this model to achieve a better understanding of the implications.

Information

Type
Independent Articles
Information
Journal of Law, Medicine & Ethics , Volume 46 , Issue 4: Next Steps in Health Reform , Winter 2018, pp. 952 - 963
Copyright
Copyright © American Society of Law, Medicine and Ethics 2018

Rising health care costs are making essential care unaffordable for many patients, prompting payers to raise deductibles or limit covered services,Reference Claxton, Levitt, Long and Blumenkranz1 and straining the US economy. In response, new models of reimbursement have been proposed to improve the value of care. For example, the prevalent fee-for-service payment structure can incentivize overuse of expensive services not shown to enhance patient outcomes (like MRIs for low back pain), so recent Medicare and Medicaid initiatives have sought to engage providers in risk-based payment arrangements that tie health outcomes to reimbursement.2 The goal is to improve patients' health while controlling spending.

While value-based purchasing for hospital and physician care has been at the forefront of payment reform, such models have lagged in the US pharmaceutical market. Inpatient and outpatient prescription drugs currently account for about 20% of health care spending, mostly driven by costly brand-name drugs, which comprise about 12% of prescriptions but 72% of spending.Reference Kesselheim, Avorn and Sarpatwari3 Recently, however, some pharmaceutical manufacturers and private payers have tried to address questions about the value of pharmaceuticals with “outcomes-based contracts.” These commercial contracts tie rebates or other cost savings on expensive prescription drugs to the outcomes observed in the patients who receive them.

For example, in 2015, 2016 and 2017, Amgen reported outcomes-based contracts for evolocumab (Repatha), a low-density lipoprotein (LDL) cholesterol-lowering drug with a list price of over $14,000 per year. Harvard Pilgrim, Cigna and other payers agreed to initial discounts with additional rebates if evolocumab did not achieve reductions in LDL cholesterol similar to those observed in the drug's key clinical trials.Reference Staton4 The most recent contract with Harvard Pilgrim stipulated a full refund for patients experiencing the relatively rare occurrence of a heart attack or stroke.Reference Sagonowsky5 For evolocumab, the degree of health benefits relative to statins was uncertain at the time of approval.Reference Staton6 Under outcomes-based contracting, Amgen shared some of the risk with payers should a drug not achieve the expected health outcome.

Despite the promise of and high expectations for outcomes-based contracts, little research has explored their advantages and limitations. One study found that, while outcomes-based contracts have existed for over a decade, they have been slower to take hold in the US than in Europe. In the US, there were only 7 publicly disclosed outcomes-based contracts during the study period, 1993-2013.Reference Garrison7 A more recent survey has shown growing popularity of such contracts in the US, with one-quarter of health plans reporting at least one outcomes-based contract in place by spring 2017, although we were unable to determine how loosely the payers in this survey defined outcomes-based contracts.Reference Choe and Learch8

To better inform stakeholders and policymakers about outcomes-based contracting, we reviewed the publicly-reported landscape of such contracts, using the combined methods of a systematic literature review and qualitative interviews with experts in the field. Our analysis focused on contracts' potential to slow pharmaceutical spending and their direct impacts on patients, payers, and manufacturers. We also sought to evaluate how these contracts are structured.

Methods

Landscape Analysis of Outcomes-Based Pharmaceutical Contracts

We first systematically surveyed the literature to identify and summarize the evidence on outcomes-based contracts. In PubMed and Google Scholar, we used the following search terms to identify our initial sample: value-based pharmaceutical contracting, value-based drug pricing, outcomes-based pharmaceutical contracts, pharmaceutical pay for performance, risk sharing agreements for drugs. From our original sample (N=217), we then excluded any articles that did not describe U.S.-based contracts tying pharmaceutical reimbursement to health outcomes, such as those that focused on financial based risk sharing agreements.

To better inform stakeholders and policymakers about outcomes-based contracting, we reviewed the publicly-reported landscape of such contracts, using the combined methods of a systematic literature review and qualitative interviews with experts in the field. Our analysis focused on contracts' potential to slow pharmaceutical spending and their direct impacts on patients, payers, and manufacturers. We also sought to evaluate how these contracts are structured.

We used the results of this literature review to develop a database of publicly disclosed outcomes-based contracts in the US. For each contract, we extracted the drug, manufacturer, payer, disease area, outcome metric (along with additional contract components) and date. We also supplemented our data with information from IHS Global Insights' outcomes-based contract database.

Qualitative Interview Study Design

We next conducted semi-structured interviews with policy experts and representatives from pharmaceutical manufacturers and public and private payers to ascertain the benefits and limitations of outcomes-based pharmaceutical contracts in the US, focusing on their potential impact on spending. This qualitative approach provided a rigorous way of capturing diverse perspectives and insights among a cohort of subjects sharing a common experience.Reference Miles and Huberman9 Qualitative methods are also ideal for exploratory research and hypothesis generation, both relevant considerations given the paucity of published data about these contracts.Reference Crabtree and Miller10 The study was approved by the ethics review board at the Harvard T. H. Chan School of Public Health.

Sample Construction and Recruitment

We recruited interview subjects using purposive sampling of manufacturers, policy experts, and payers, including both pharmaceutical benefit managers and health plans. Potential participants were identified through the literature review (e.g. quotations from news articles and public announcements of contracts). Participants had direct experience either evaluating or implementing outcomes-based contracts. All contacted individuals and organizations participated, except one manufacturer, one payer, and one policy expert, each of whom did not respond to two emails we sent introducing the study and inviting participation. We interviewed nine manufacturers, six policy experts, and eight payers.

Data Collection

One investigator (ES) conducted all interviews via telephone between November 2016 and March 2017. All interviews but one were recorded and transcribed by an independent professional transcriptionist (one interviewee did not consent to audiotaping). The interviewer also took extensive notes.

Interviews followed semi-structured discussion guides exploring key thematic areas. We developed separate guides for manufacturers, payers, and policy experts. The guides for manufacturers and payers covered a similar range of topics including motivation, contract structure, data challenges, regulatory barriers, implications for spending and testing in Medicare and Medicaid populations (Appendix 1). Interviews elicited information on how many outcomes-based contracts each organization had along with detailed information about the general structure of the contracts, including applied populations (e.g. commercial vs Medicare) and reimbursement mechanism. We also asked manufacturers and payers to discuss the role of pharmaceutical benefit managers and health plans in negotiating and allocating rebate dollars; we especially probed health plans on whether they used PBMs to negotiate the rebates in the outcomes-based contracts. While there was not yet evidence from recently-enacted contracts on outcomes and implications for total spending, we also asked payers about the size of rebates. Finally, we asked each manufacturer and health plan whether they would be interested in participating in future pilots from the Centers for Medicare and Medicaid (CMS) to test outcomes-based contracts in the Medicare and Medicaid populations.

The guide for policy experts covered broader topics associated with outcomes-based contracts. We first asked the policy experts to define outcomes-based contracts and to discuss their advantages and disadvantages. We then prompted them to discuss regulatory barriers for outcomes-based contracts. We also asked them to assess the potential for outcomes-based contracts to reduce prescription drug spending. We also asked what potential there may be for CMS to test outcomes-based contracts in the Medicare and Medicaid population and whether there would likely be political support for this. Finally, interviewers shifted to the topic of patients and whether patients' access to these medicines and/or out of pocket costs are impacted by outcomes-based contracts.

Prior to the interviews, we sent an email to each participant explaining the study goals and our motivations for doing this research — to contribute to evidence on outcomes-based contracts in the US. Interviewees also received a copy of the interview guide in advance. One interviewee elected to provide written responses to the guide in place of being interviewed.

Data Analysis

Interviews lasted 45-60 minutes. Interview notes and transcriptions were analyzed using standard coding techniquesReference Constas11 and the constant comparative method of qualitative data analysis.Reference Bradley and Holmboe12 The investigators developed an initial set of codes based on the interview guide. During the coding process, new codes were included to capture additional concepts and themes that emerged through the interviews. The final coding structure consisted of 12 broad themes spanning 31 discrete codes.

One investigator (ES) then used the final coding scheme to code the transcripts and interview notes using the QDA Miner software package.13 The software was also used to conduct and organize the content analysis. Before submitting this manuscript for publication, we asked three participants to serve as internal reviewers — one manufacturer, one payer and one subject expert. Appendix 2 shows the consolidated criteria for reporting qualitative research (COREQ) for this study.

Results

Literature Review Findings

The systematic literature review uncovered 90 articles, including 28 public announcements of outcomes-based contracts, 57 editorials describing the challenges involved in outcomes-based contracts, and 5 qualitative studies of outcomes-based contracts (most funded by the pharmaceutical industry).Reference Choe and Learch14 One study reported that growth of outcomes-based contracts in the US appeared to plateau in 2014.15 The other studies identified barriers to the use of outcomes-based contracts.16

The literature review identified over 30 publicly-disclosed outcomes-based contracts in the US, the majority of which were for high-cost, brand-name drugs (Exhibit 1). Disease areas included heart disease, diabetes, osteoporosis, multiple sclerosis, rheumatoid arthritis, hepatitis C virus infection, and cancer. Manufacturers often set up contracts with multiple payers for a given drug. In these cases, the outcome metric was often the same across contracts with different payers, although the rebate levels and other contract provisions varied.

Qualitative Interview Findings

The majority of manufacturers and payers interviewed reported having at least one outcomes-based contract. Most had confidential outcomes-based contracts in addition to publicly disclosed ones.

Our qualitative analysis identified 6 key issues affecting the viability of outcomes-based contracts in the US: contract structures and data challenges, regulatory barriers, implications for spending, CMS testing, and implications for patients.

Contract Structures and Data Challenges

The majority of payers, even for those with Medicare Advantage plans, reported that their outcomes-based contracts were limited to commercial markets. Some payers used PBMs to help them negotiate rebates for their outcomes-based contracts, while other payers negotiated rebates directly with manufacturers for their outcomes-based contracts. For PBMs that negotiated their own outcomes-based contracts with manufacturers, they needed to determine whether to include the health plan or employer in this contract, or whether to keep the outcomes-based contract separate from the contract they had with purchasers.

Outcomes-based contracts were structured a number of ways. Respondents reported that most outcomes-based contracts were modeled on a population level, in which manufacturers provided additional rebates for all of the drugs utilized under the contract if a specified percentage of the population did not achieve the desired outcome. For example, in the case of some diabetes drugs, additional rebates were given if specified hemoglobin A1C blood levels were not met in the patient population. Amgen's contract with Harvard Pilgrim for evolocumab was the only example of a contract set on an individual level, in which Amgen provided a refund for each patient who experienced a heart attack or stroke.

All respondents reported that a critical step was defining the outcome metric and ensuring that it was readily available in existing databases. One payer reflected on this tradeoff between achieving a rigorous outcome metric and using feasible data sources:

Let's say hypothetically there was a focus on multiple sclerosis. Then as you know claims data isn't going to be a great measure of relapse so we're going to need more. We can look at durable medical equipment or UTIs, but that's pretty distal. So we need to capture some sort of patient-reported outcomes. That's sort of the process for coming to consensus around measures. Part of the consensus around measures is going to be not just “What's the right measure?” but also feasibility and understanding the challenges around collection. Those decisions aren't going to be made in a blue-sky environment, but considering the real-world constraints of doing it in the population.

In the case of drugs that treat heart disease, such as sacubitril/valsartan (Entresto) and prasugrel (Effient), heart attacks or hospitalizations are relatively straightforward outcome measures to track, using claims data. Prior authorization data may also be used to supplement claims data. Other outcome metrics, such as LDL cholesterol or hemoglobin A1c must be obtained through lab data reports. This process of obtaining timely data and analyzing it could be costly and burdensome and was cited by some health plans as a deterrent to setting up outcomes-based contracts. Health plans or neutral third parties often conduct the data analyses; manufacturers were given the right to appeal contentious results, but were careful to remain at arms-length from the data to comply with privacy regulations.

In the majority of cases, the outcomes measures were defined to be consistent with the drug's clinical trials. Respondents reported that this was largely due to regulatory requirements that marketing for drugs be consistent with the FDA-approved usage. One large difference between clinical trials and real-world data is that adherence rates are often higher in clinical trials. As a result, manufacturers sometimes included adherence thresholds that the health plan must achieve before additional rebates were triggered in the event that the defined outcome was not met. This effectively guaranteed a minimum market share for manufacturers, offsetting revenue lost from higher rebates with additional revenue from increased utilization. The purchasers usually administered the adherence programs. Some manufacturers reported that while they may share adherence information with purchasers from their clinical trials experience, anti-kickback statutes prohibited them from playing a direct role in adherence programs. Other manufacturers have been more willing to challenge the validity of this anti-kickback concern by providing support to these adherence programs.

Another advantage of outcomes-based contracts is that they can provide additional data on how medicines perform in the real world, outside of clinical trials. Clinical trials often have strict selection criteria for stage of disease and other comorbidities. In disease areas with accelerated review pathways, they can also operate in short time frames of three years. Outcomes-based contracts can offer data on longer term, more meaningful outcome metrics, such as strokes or heart attacks versus cholesterol levels in the case of evolocumab. In doing so, they reduce uncertainty for payers and reward truly innovative drugs.

We found one example of an outcomes-based contract tying reimbursement to comparative effectiveness data. Eli Lilly and Harvard Pilgrim signed an outcomes-based contract for the diabetes drug dulaglutide (Trulicity), a once per week injectable drug, in which the defined hemoglobin A1C outcome metric was defined such that Lilly would pay additional rebates to Harvard Pilgrim if patients taking dulaglu-tide did not achieve better A1C blood sugar levels than competing oral drugs.

Barriers: Real and Imagined

One often-discussed issue was the Medicaid Best Price rule. Under this rule, manufacturers must extend any additional rebates given under outcomes-based contracts to the entire Medicaid market. Some respondents identified this as a barrier to manufacturers' ability to provide larger rebates under outcomes-based contracts. However, because the contracts' rebate levels are confidential, there was no public information on whether the additional rebates under outcomes-based contracts even approached the Medicaid Best Price. As a result, we could not determine whether the Medicaid Best Price acts as a price floor or whether advocates were using outcomes-based contracts as an opportunity to seek changes to this rule.

Almost all respondents said that while their legal teams must construct outcomes-based contracts carefully in light of certain regulations (the Medicaid Best Price rules, anti-kickback statues, labeling requirements), these issues were not deterrents. Only one respondent out of the 17 manufacturers and payers cited regulatory barriers as a reason for not engaging in outcomes-based contracts. In the other cases in which the respondents did not have outcomes-based contracts, the primary reason cited was either lack of interest generally, or skepticism that contracts could deliver savings.

Unclear Impact on Pharmaceutical Spending

Some manufacturers reported using outcomes-based contracts to ensure coverage of their drug, especially in cases in which the payer questioned the value of a drug or in which a competing product was available. Manufacturers' representatives also viewed these new contracts as opportunities to focus the value discussion on the benefit side of the equation:

There's payment innovation happening with hospitals and providers that isn't necessarily happening with pharmaceuticals. This is a way to advance that. Fee for service is much easier, much faster and more predictable, allowing payers to understand the rebate back to them on day 1. These contracts create a floating payment mechanism that is pretty unique and different. We want to flush out the payers that are truly interested in spending the time to build something together versus the ones that are really in it for the bigger rebate.

Payers, on the other hand, reported a strong desire to contain pharmaceutical spending with a value-based framework such as outcomes-based contracts:

We're ok with people using these drugs more broadly than we might otherwise like as a payer and we'll let you do that, but if there is not a response, don't send us a bill.

Most respondents reported that their outcomes-based contracts were still too early in the contract cycle to have demonstrated any clear impact on pharmaceutical spending, although the analyses and implications for total spending were likely to remain confidential even once they become available. Respondents in our study expressed mixed views on the size of the rebates, suggesting that many factors other than outcomes-based contracts likely impact rebate size, such as the payer's market power and whether the drug has close therapeutic substitutes. A few of the payers that had outcomes-based contracts reported that they were able to negotiate large rebates by giving preferred formulary status to drugs with outcomes-based contracts. In this case, if the larger rebate was triggered because the drug does not perform, and there was not an offsetting increase in utilization, there could be reductions in total spending. This also assumes that manufacturers are not factoring additional rebates into their pre-rebate price. One payer was optimistic about the potential for outcomes-based contracts to achieve savings:

It is complicated, but depending on what the therapeutic area is, it may be worth the effort. In a large population, it should be scalable and easy to measure. In a small population, the difference between a $100,000 and a $250,000 bill depending on whether the patient responded may be worth the effort in picking up the phone to measure the response. While measuring some of these things is not always easy, if the dollar difference is sufficient, we can figure out how to do it.

Other payers reported that the additional rebates negotiated were not large enough to provide meaningful savings. In these cases, when the drug did not perform well, the expected rebates under outcomes-based contracting did not provide savings in excess of the guaranteed rebates under traditional contracting. In addition, payers pointed out that the rebates they would receive in the event that the outcome was not met would not be enough to offset the increase in costs associated with collecting and analyzing the data, as well as the predicted broader use of the drug:

We have attempted it and moved on. It can take 6-9 months to agree on the terms of an outcomes-based contract. One has to address data challenges such as determining how many lines down on the coding of the EMRs, whether there are enough patients over the contract period to run out the data, how to account for non-adherence, etc. At the end of the day, we didn't believe pharma would really come to the table with a plan that would save us money and we feared that we would get less money than a traditional rebate if the drug doesn't work.

Indeed, most of the payers that did not have outcomes-based contracts reported relatively small rebate differences as being the primary reason for not engaging in outcomes-based contracts.

The majority of manufacturer and payer respondents reported that a clear driver of outcomes-based contracts was improved public relations. Manufacturers have touted outcomes-based contracts to Congress and the public in response to the growing scrutiny over high pharmaceutical prices.17 Private payers have also used outcomes-based contracts to signal their broader mission of improving population health and delivering high value care to employers. Interviewees suggested that the contracts help show a willingness to collaborate and share in risk.

Outcomes-based contracts that include adherence programs may benefit patients to the extent that the programs provided genuine support to patients. One respondent recalled reports of one payer automatically filling and mailing patients' monthly prescriptions to them to increase the medical possession ratio, without any follow-up to ensure that the patients received or took their medicines.

Implications for Patients

In response to cost pressures, a growing number of private payers have tried to reduce drug prices by adopting more restrictive formularies that exclude products. Participants in our study reported that outcomes-based contracts may provide an alternative, allowing drugs to remain on formularies in exchange for outcomes guarantees. Some payers reported that a manufacturer's willingness to engage in an outcomes-based contract may help that drug receive exclusive or preferential formulary status. In these cases, the outcomes-based contract may improve patients' access to that specific drug, although it may also decrease access to other competing drugs that are excluded or given lower formulary status as part of the deal. Other payers said that they separate the outcomes-based contract from the formulary decision — that is, they first decide whether to cover the drug. Following that decision, they determine whether the drug would be eligible for an outcomes-based contract.

Respondents pointed out that outcomes-based contracts tend not to affect patients' out-of-pocket costs and are unlikely to result in lower premiums. Patient costs occur at the point of sale, while the contracts provide retrospective rebates from manufacturers to the payers that might be calculated months (or years) after the prescription is filled. In addition, the contracts would not be able to help patients who decide not to fill needed medications because they could not afford out-of-pocket costs.

Respondents also explained that outcomes-based contracts were dependent on data availability, limiting the scope of drugs to which outcomes-based contracts apply as well as the meaningfulness of the outcomes on which the rebates are based. All respondents agreed that in the near term, outcomes-based contracts would not apply to the majority of new drugs because the outcomes available for these contracts generally only include those that can be measured in claims databases. Electronic health records could theoretically offer additional opportunities, but in most cases, are too difficult to access or analyze for these purposes. Even laboratory values may not be available in many claims databases. Finally, using patient reported outcomes through disease-specific instruments such as EDSS in MS may be a way to make outcomes more meaningful as payers are increasingly able to integrate data from different sources into these models.

Respondents also pointed out another limitation of outcomes-based contracts: that outcomes measures must be available in short time frames. As a result, the negotiated outcomes are often “surrogate” measures, such as changes in laboratory values or other easy-to-obtain results that may not closely or directly correlate with actual clinical outcomes that are more central to patients' health. For example, hemoglobin A1c level is a common measure of diabetes control, but drugs that slightly reduce this number may not be associated with improvements in adverse outcomes from diabetes like kidney damage or reductions in diabetes-related cardiovascular death.Reference Seeley and Kesselheim18 Respondents noted that since hemoglobin A1c can be easily measured and may change in the course of a few months, it was a common basis for a contract. In the case of oncology, surrogate measures may include tumor shrinkage or discontinuation of treatment after a short period of time, but these were not necessarily indicative of more meaningful outcomes such as cancer survival rates.

Finally, outcomes-based contracts that include adherence programs may benefit patients to the extent that the programs provided genuine support to patients. One respondent recalled reports of one payer automatically filling and mailing patients' monthly prescriptions to them to increase the medical possession ratio, without any follow-up to ensure that the patients received or took their medicines:

There have been some allegations whereby the manufacturer basically had this pharmacy forcing refills on patients. The patients were getting mail-order refills that they weren't actually using.

It is therefore crucial to ensure that the adherence programs are run appropriately by engaging pharmacists, patients, and physicians.

High Interest in Testing Outcomes-Based Contracts

Most manufacturers expressed interest in testing outcomes-based contracts with Medicare or Medicaid, and most reported that they have already discussed this prospect with CMMI. One manufacturer's representative elucidated some of the potential challenges in introducing outcomes-based contracts to the Medicare population:

There is the option to do both the commercial and Medicare populations but we have to think about the differences. One of the things we have to think about is continuous patient enrollment. If you're trying to follow someone over a period of time and all of a sudden they drop out of your plan enrollment, then you can no longer follow their activity whether on claims, prescription claims or medical claims. For Part D, there's a high degree of turnover in the plan year after year versus in commercial where potentially for employer-sponsored plans you have a little bit more stickiness in terms of their enrollment.

Despite these challenges, the manufacturer remained strongly interested in working with CMS to test outcomes-based contracts:

We have actively talked to CMS about the value-based arrangements. Last summer there was a notification that was published about manufacturers and plans having value-based arrangements and a recognition that this type of activity was actually happening in the marketplace. That notification opened up a mailbox essentially requesting for companies and plans to feel free to meet with CMS and provide information, and we have actively done that. To the question of CMMI and whether there is an opportunity to pursue a pilot or a demo, the answer would be “Yes, we are open to that.”

According to one policy expert, there was no specific reason why CMMI had not yet engaged in pharmaceutical outcomes-based contracts pilots, save their limited resources and competing interest in other value-based purchasing areas such as ACOs, bundled payments, and patient-centered medical homes.

Discussion

This mixed-methods study explored the question of whether outcomes-based contracts offer the possibility of moving the US health care system closer to the goal of paying for prescription drug value. Payers want the value discussion to focus on the potential for cost savings. If payers have been wasting resources on drugs that do not perform as well outside of clinical trials, tying rebates to outcomes can shift pharmaceutical spending to more effective treatments. Manufacturers, by contrast, emphasize the benefit side of the value equation, hoping to be rewarded for their drugs with therapeutic superiority. As a result, manufacturers and health plans alike are showing increasing interest in this model. Yet this hope that outcomes-based contracts might lower pharmaceutical spending must be tempered by what is known about their application thus far: that they apply to a limited subset of drugs and markets, are often tied to surrogate measures that do not directly reflect patient health outcomes and do little to substantially lower drug spending.

There is limited publicly available information with which to assess the impact of outcomes-based pharmaceutical contracts, in part because these contracts have been primarily applied to the private insurance market. Lacking adequate evidence, it is difficult to ascertain how much they improve value. Any estimation of savings must also take into account the increase in labor and information technology costs that are required to execute and evaluate these costs. Structuring contracts in favor of more effective drugs is more consistent with value-based purchasing in other parts of the health care system. However, it should not be viewed as a solution to high prescription drug prices without a clearer understanding of these contracts' implications on health and spending. Another crucial point of debate is whether Medicare or Medicaid should consider broader use of outcomes-based pharmaceutical contracts. The best way to determine whether this approach can improve value is to test these arrangements through the CMS. Shortly after the conclusion of this study, CMS announced that it had signed a contract with Novartis for tisagenlecleucel (Kymriah), its new CAR-T therapy, in which Novartis had agreed to provide a full refund if the CART-T therapy was not observed to have worked after one month (an extremely short period of time in which to judge whether response has occurred). This seemed to signify CMS' willingness to engage in outcomes-based contracts for new, high cost drugs. Then in July 2018, a year after the agreement, CMS cancelled this outcomes-based contract. Although information on what drove their decision was not made public, it appears not to have been a fundamental change in approach, since around the same time CMS granted approval for Oklahoma's outcomes-based contracts for the schizophrenia drug aripiprazole lauroxil (a new formulation of an older product) and antibiotic oritavancin (see Exhibit 1). In November 2018, CMS also granted permission to Michigan's Medicaid program to engage in such contracts.

Many other dimensions would need to be considered to encourage wider use of outcomes-based contracts under CMS, beyond individual state Medicaid programs. Currently, CMS ties bonuses to the quality performance of Medicare Part D plans using a star system that measures outcomes, intermediate outcomes, patient experience, access and process.19 These incentives would need to be made consistent with outcomes-based contracts, perhaps by adding outcomes-based contracts to the list of criteria used to award stars. In addition, Medicare Part D plans are currently required to report rebate savings to CMS, which CMS then recoups by factoring these savings into lower per member per month rates that it pays to plans.20 Thus, Medicare Advantage plans that administer Part D would need to be given sufficient incentives to engage in outcomes-based pharmaceutical contracts. One option could be to allow plans to retain some of the rebate savings by easing the per member per month downward adjustments. These savings could be applied to offsetting the other costs associated with administering outcomes-based contracts. An understanding of these complexities will be crucial for policymakers as they consider whether to expand outcomes-based contracts to additional new drugs.

There is limited publicly available information with which to assess the impact of outcomes-based pharmaceutical contracts, in part because these contracts have been primarily applied to the private insurance market. Lacking adequate evidence, it is difficult to ascertain how much they improve value. Any estimation of savings must also take into account the increase in labor and information technology costs that are required to execute and evaluate these costs.

Thus, the decision to test outcomes-based contracts under Medicare Part D should be made independent of other potential changes in pharmaceutical reimbursement, such as direct pricing reform. Contracts between manufacturers and health plans could include private plans administering Medicaid and Medicare, as well government-run plans. Price and health outcomes data should be kept confidential, but rigorously and independently analyzed. Any outcomes-based contracting pilot should remain voluntary until there is public information on its implications. A voluntary program with public payers may prompt manufacturers to engage in outcomes-based contracts for drugs that are unlikely to result in them losing significant revenue. Thus, it is important to note that testing outcomes-based contracts through a voluntary program would be unlikely to achieve meaningful savings in comparison to a model in which CMS requires outcomes-based contracts for all new, high-cost drugs in exchange for coverage.

This study combined qualitative stakeholder interviews with a systematic literature review to assess outcomes-based pharmaceutical contracts in the US. Respondents' views in this study do not necessarily reflect the views of all US manufacturers and payers. All of the purchasers interviewed represented health plans or PBMs. We did not interview health care systems/providers, which can also be purchasers of pharmaceuticals. One of our respondents noted that their health plan is vertically integrated with a health care system and that one of the hospitals in this health care system is currently negotiating their own outcomes-based contract. In addition to their role as purchasers, providers are also critical in ensuring the success of outcomes-based contracts during the diagnostic and treatment determination process. For example, to achieve optimal outcomes in oncology, contracts depend on physicians being able to appropriately identify each patient's specific mutation and the best treatment to target that mutation. Further research is needed on the role of providers in outcomes-based contracts as both purchasers and prescribers. An important research question concerns whether Accountable Care Organizations are becoming increasingly interested in pursuing outcomes-based contracts.

Exhibit 1.

Publicly Disclosed Outcomes-based Pharmaceutical Contracts in the US

Source: Authors' analysis and IHS Global Insights.

Another area in which further research could yield important findings is gene therapies. Our study did not specifically probe the potential future application of outcomes-based contracts to genetic therapies because the approval of gene therapies is still in its infancy in the US. However, the literature often highlights the possibility of outcomes-based contracting being used as a reimbursement tool to address the high up-front costs for genetic therapies.Reference Marsden and Towse21

A commonly cited example is the use of annuity payments, in which payments are made over multiple years based on the outcome of the gene therapy, as in Harvard Pilgrim's contract with Luxturna.Reference Sagonowsky22 A study focusing on how outcomes-base payments could be modeled for these types of drugs could advance the discussion on outcomes-based contracts more broadly.

Finally, it is worth noting that the outcomes-based contracts to date appear highly concentrated amongst a small number of payers. We did not find a clear answer as to why certain plans have been reluctant to engage while other plans such as Harvard Pilgrim are at the forefront of this. Factors may include whether the health plan has an internal champion and the degree to which the health plan has decided to focus on population health/value based contracting approaches. Certain structural elements can also make a difference such as the payer's ability and experience with data analysis and whether the payer is able to negotiate rebates outside of PBMs. Further research is needed on the drivers of which payers and which manufacturers choose to enter into these agreements.

The high costs of brand-name drugs are squeezing the budgets of patients and public and private payers alike, often without clear health gains. Outcomes-based pharmaceutical contracts are increasingly identified as a potential solution, but many questions remain regarding whether they can deliver quality improvements or savings. Stakeholders' willingness to share the results of their pilot contracts in the commercial market would help policymakers evaluate whether the risks outweigh the potential gains. Broader testing of outcomes-based contracts in Medicare and Medicaid populations would be an important first step in determining whether these contracts can truly improve value in pharmaceutical care, or whether their primary value is as a public relations strategy.

Footnotes

This project was funded by a grant from the Commonwealth Fund. Dr. Seeley and Dr. Chimonas do not have any conflicts of interest or disclosures to make. Dr. Kesselheim's work is also supported by the Laura and John Arnold Foundation, with additional support from the Harvard-MIT Center for Regulatory Science and the Engelberg Foundation. These organizations had no role in the study design or in the collection, analysis, and interpretation of data.

References

Claxton, G., Levitt, L., Long, M., and Blumenkranz, E., “Increases in Cost-Sharing Payments Have Far Outpaced Wage Growth,” Peterson-Kaiser Health System Tracker, October 4, 2017, available at <https://www.healthsystemtracker.org/brief/increases-in-cost-sharing-payments-have-far-outpaced-wage-growth/?post_types=brief#item-start> (last visited November 9, 2018).+(last+visited+November+9,+2018).>Google Scholar
The Centers for Medicare and Medicaid, “What are the Value-Based Programs?” available at <https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Value-Based-Programs/Value-Based-Programs.html> (last visited November 9, 2018).+(last+visited+November+9,+2018).>Google Scholar
Kesselheim, A. S., Avorn, J., and Sarpatwari, A., “The High Cost of Prescription Drugs in the United States: Origins and Prospects for Reform,” JAMA 316, no 8 (2016): 858–871.Google Scholar
Staton, T., “Amgen Scores PCSK9 Exclusive in Pay-for-Performance Deal with Harvard Pilgrim,” FiercePharma, November 9, 2015, available at <https://www.fiercepharma.com/sales-and-marketing/amgen-scores-pcsk9-exclusive-pay-for-performance-deal-harvard-pilgrim> (last visited November 9, 2017); T. Staton, “Cigna Inks Results-Based Deals on Pricey Amgen, Sanofi PCSK9 meds,” FiercePharma, May 11, 2016, available at <https://www.fiercepharma.com/pharma/cigna-inks-results-based-deals-pricey-amgen-sanofi-pcsk9-meds> (last visited November 9, 2018).+(last+visited+November+9,+2017);+T.+Staton,+“Cigna+Inks+Results-Based+Deals+on+Pricey+Amgen,+Sanofi+PCSK9+meds,”+FiercePharma,+May+11,+2016,+available+at++(last+visited+November+9,+2018).>Google Scholar
Sagonowsky, E., “Amgen puts Repatha Outcomes Data to Work in Refund Deal with Harvard Pilgrim,” FiercePharma, May 3, 2017, available at <https://www.fiercepharma.com/pharma/amgen-inks-repatha-refund-arrangement-harvard-pilgrim> (last visited November 9, 2018).+(last+visited+November+9,+2018).>Google Scholar
Staton, T., “It's Not Just Payers Holding Back on PCSK9 Outcomes Data. Doctors are Holding Back Too,” FiercePharma, February 16, 2016, available at <https://www.fiercepharma.com/sales-and-marketing/it-s-not-just-payers-waiting-on-pcsk9-outcomes-data-doctors-are-holding-back> (last visited November 9, 2018). (last visited November 9, 2018).' href=https://scholar.google.com/scholar?q=Staton,+T.,+“It's+Not+Just+Payers+Holding+Back+on+PCSK9+Outcomes+Data.+Doctors+are+Holding+Back+Too,”+FiercePharma,+February+16,+2016,+available+at++(last+visited+November+9,+2018).>Google Scholar
Garrison, L. et al., “Private Sector Risk-Sharing Agreements in the United States: Trends, Barriers, and Prospects,” American Journal of Managed Care 21, no. 9 (2015): 632–640.Google Scholar
Choe, S. and Learch, C., “Health Plans Are Actively Exploring Outcomes-Based Contracts,” Avalere Health, May 30, 2017, available at <http://avalere.com/expertise/life-sciences/insights/health-plans-are-actively-exploring-outcomes-based-contracts> (last visited November 9, 2018).+(last+visited+November+9,+2018).>Google Scholar
Miles, M.B. and Huberman, A.M., Qualitative Data Analysis: An Expanded Sourcebook (Thousand Oaks, CA: Sage Publications, 1994).Google Scholar
Crabtree, B.F. and Miller, W.L., Doing Qaualitative Research (Thousand Oaks, CA: Sage Publications, 1999).Google Scholar
Constas, M.A., “Qualitative Data Analysis as a Public Event: The Documentation of Category Development Procedures,” American Educational Research Journal 29, no. 2 (1992): 253–266.Google Scholar
Bradley, E.H., Holmboe, E.S., et al., “A Qualitative Study of Increasing Beta-Blocker Use after Myocardial Infarction: Why do Some Hospitals Succeed?” JAMA 285, no. 20 (2001): 2604–2611.Google Scholar
QDA Miner Lite, Qualitative Text Analysis Software, available at <https://provalisresearch.com/products/qualitative-data-analysis-software/freeware/> (last visited November 9, 2018).+(last+visited+November+9,+2018).>Google Scholar
See above Garrison et al., supra note 7; Choe, S., Learch, C., “Health Plans Are Actively Exploring Outcomes-Based Contracts,” Avalere Health, May 30, 2017, available at <http://avalere.com/expertise/life-sciences/insights/health-plans-are-actively-exploring-outcomes-based-contracts> (last visited November 9, 2018); A. Sexton Ward et al., “Regulatory, Legal Uncertainties Are Barriers to Value-Based Agreements for Drugs,” Health Affairs Blog, November 4, 2016, available at <https://www.healthaffairs.org/do/10.1377/hblog20161104.057443/full/> (last visited November 9, 2018); P. Keckley, “Value-Based Pricing for Pharmaceuticals: Implications of the Shift From Volume to Value,” Deloitte Issue Brief, 2012, available at <http://deloitte.wsj.com/cfo/files/2012/09/ValueBasedPricingPharma.pdf> (last visited November 9, 2018); P. Neuman, “Risk-Sharing Arrangements that Link Payment for Drugs to Health Outcomes Are Proving Harvard to Implement,” Health Affairs 30, no. 12 (2011): 2329–2337.+(last+visited+November+9,+2018);+A.+Sexton+Ward+et+al.,+“Regulatory,+Legal+Uncertainties+Are+Barriers+to+Value-Based+Agreements+for+Drugs,”+Health+Affairs+Blog,+November+4,+2016,+available+at++(last+visited+November+9,+2018);+P.+Keckley,+“Value-Based+Pricing+for+Pharmaceuticals:+Implications+of+the+Shift+From+Volume+to+Value,”+Deloitte+Issue+Brief,+2012,+available+at++(last+visited+November+9,+2018);+P.+Neuman,+“Risk-Sharing+Arrangements+that+Link+Payment+for+Drugs+to+Health+Outcomes+Are+Proving+Harvard+to+Implement,”+Health+Affairs+30,+no.+12+(2011):+2329–2337.>Google Scholar
See Garrison et al., supra note 7.Google Scholar
Eli Lily and Company and Anthem, Promoting Value Based Payment Arrangements, January 29, 2016, available at <https://lillypad.lilly.com/WP/wp-content/uploads/LillyAnthemWP2.pdf> (last visited November 9, 2018); See Sexton Ward et al., supra note 14.+(last+visited+November+9,+2018);+See+Sexton+Ward+et+al.,+supra+note+14.>Google Scholar
The Cost of Prescription Drugs: How the Drug Delivery System Affects What Patients Pay, Full Committee Hearing to the US Senate Committee on Health, Education Labor and Pensions, June 13, 2017, available at <https://www.help.senate.gov/hearings/the-cost-of-prescription-drugs-how-the-drug-delivery-system-affects-what-patients-pay> (last visited November 9, 2018).+(last+visited+November+9,+2018).>Google Scholar
Seeley, E. and Kesselheim, A., “Outcomes-Based Pharmaceutical Contracts: An Answer to High Drug Spending?” Commonwealth Fund Issue Brief, September 27, 2017, available at <https://www.commonwealthfund.org/publications/issue-briefs/2017/sep/outcomes-based-pharmaceutical-contracts-answer-high-us-drug> (last visited December 11, 2018).+(last+visited+December+11,+2018).>Google Scholar
Centers for Medicare and Medicaid, available at <https://www.cms.gov/Newsroom/MediaReleaseDatabase/Fact-sheets/2016-Fact-sheets-items/2016-10-12.html> (last visited November 9, 2018).+(last+visited+November+9,+2018).>Google Scholar
Concerns with Rebates in the Medicare Part D program, Department of Health and Human Services OIG report, March 2011, available at <https://oig.hhs.gov/oei/reports/oei-02-08-00050.pdf> (last visited November 9, 2018).+(last+visited+November+9,+2018).>Google Scholar
Marsden, G. and Towse, A., “Pearson S. Gene Therapy: Understanding the Science, Assessing the Evidence, and Paying for Value. A Report from ICER and OHE,” March 2017, available at <https://icer-review.org/wp-content/uploads/2017/03/ICER-Gene-Therapy-White-Paper-030317.pdf> (last visited November 9, 2018); T Beaton, “Outcomes-Based Contracts Offer Payers New Pharmaceutical Options,” HealthPayer Intelligence, January 2018, available at <https://healthpayer-intelligence.com/news/outcomes-based-contracts-offer-payers-new-pharmaceutical-options> (last visited November 9, 2018).+(last+visited+November+9,+2018);+T+Beaton,+“Outcomes-Based+Contracts+Offer+Payers+New+Pharmaceutical+Options,”+HealthPayer+Intelligence,+January+2018,+available+at++(last+visited+November+9,+2018).>Google Scholar
Sagonowsky, E., “Spark Sets off Gene Therapy Debate with $850K Sticker on Luxturna,” FiercePharma, January 2018, available at <https://www.fiercepharma.com/pharma/spark-prices-gene-therapy-luxturna-at-850k-grabbing-top-spot-pharma-s-costliest-drugs> (last visited November 9, 2018); Harvard Pilgrim Healthcare news, “Harvard Pilgrim is the First Health Plan to Directly Negotiate Outcomes-Based Agreement for Groundbreaking Gene Therapy,” January 3, 2018, available at <https://www.harvardpilgrim.org/public/news-detail?nt=HPH_News_C&nid=1471914707173> (last visited November 9, 2018).+(last+visited+November+9,+2018);+Harvard+Pilgrim+Healthcare+news,+“Harvard+Pilgrim+is+the+First+Health+Plan+to+Directly+Negotiate+Outcomes-Based+Agreement+for+Groundbreaking+Gene+Therapy,”+January+3,+2018,+available+at++(last+visited+November+9,+2018).>Google Scholar
Figure 0

Exhibit 1.Publicly Disclosed Outcomes-based Pharmaceutical Contracts in the US

Source: Authors' analysis and IHS Global Insights.