Outcomes-Based Contracting for Disease-Modifying Therapies in Multiple Sclerosis: Necessary Conditions for Paradigm Adoption

In 2018, specialty medications accounted for approximately 1% of prescriptions filled in the United States, but for 45% of the total drug spending.¹ Consequently, payers have begun to adopt agreements with drug manufacturers called “outcomes-based contracts,” in which “the payment terms for medication(s) or other health care technologies are tied to agreed-upon clinical circumstances, patient outcomes, or measures.”² In therapeutic areas with uncertainty regarding clinical outcomes that are dependent on high-cost specialty medications, outcomes-based contracts are an attractive vehicle to reduce financial risk for payers. Through such contracts, payers receive assurances that they will “get their money’s worth,” and drug manufacturers are provided earlier access to their drugs and have significant sales for successful drugs. A typical mechanism of risk sharing in an outcomes-based contract is a rebate or supplemental rebate to the payer if the drug does not meet its agreed-on outcomes measures (ie, end points).

Several outcomes-based contracts have been implemented in healthcare over the past decade, and the pace of their implementation is quickening.³ A total of 62 outcomes-based contracts have been publicly announced in the United States between 2009 and 2019.⁴ A 2018 survey conducted by Avalere Health showed that more than 25% of all health plans have an outcomes-based contract in place.⁵ In all, 85% of those health plans expressed interest in pursuing additional contracts.⁵ A well-known example of an outcomes-based contract is between Novartis and multiple private insurers for its heart failure combination drug, sacubitril plus valsartan; under these contracts, hospitalization rates may determine whether additional rebate payments are paid to insurers.⁶

Multiple sclerosis is a compelling therapeutic area for outcomes-based contracts, because of the number of disease-modifying therapies that are on the market for this disease, spanning 3 routes of administration, high drug costs,⁷ and variability in patient responses to therapies.⁸ In clinical trials, US Food and Drug Administration (FDA)-approved disease-modifying therapies for multiple sclerosis have demonstrated a reduction in the number of disease relapses, a delay in disability progression, and the ability to limit new disease activity (as seen on magnetic resonance imaging [MRI]); these drugs are the mainstay of treatment for patients with multiple sclerosis.⁸

According to Express Scripts’ 2018 Drug Trend Report, in 2018, medications for the treatment of multiple sclerosis were the fourth costliest drug therapy category, averaging $55.81 per member annually.⁹ Drug manufacturers and payers have begun experimenting with the use of outcomes-based contracts for disease-modifying therapies for multiple sclerosis.^7,10 The 2018 survey by Avalere Health revealed that 10% of outcomes-based contracts target immune or inflammatory disease states,⁵ yet little is known about the details of those contracts.

The widespread adoption of outcomes-based contracts in the United States involves significant challenges.^2,11 Organizations such as AMCP (Academy of Managed Care Pharmacy) and ISPOR (International Society for Pharmacoeconomics and Outcomes Research) have identified some challenges and provided strategies to develop, evaluate, implement, and monitor these contracts.^2,12 To further understand these challenges and offer some insights on how to overcome them, we conducted a series of market research interviews with industry representatives. The interviews’ objectives were to:

1. Identify payer and manufacturer perspectives on outcomes-based contracts
2. Learn how end points are selected for use in outcomes-based contracts
3. Explore the potential role of blood-based clinical biomarkers as end points in outcomes-based contracts
4. Explore the applicability of outcomes-based contracts to disease-modifying therapies for multiple sclerosis
5. Identify the necessary conditions for the widespread adoption of outcomes-based contracts.

Methods

A total of 58 representatives from payers (national, regional, and local health insurance plans, pharmacy benefit managers), drug manufacturers, and industry consultants were invited to participate in our study. The invitees were targeted based on their experience with outcomes-based contracts and, for the manufacturers, their disease-modifying therapy portfolios related to multiple sclerosis. Specific individuals were identified through the AMCP; the Pharmacy Quality Alliance; the primary investigator, Octave Bioscience; and LinkedIn. The invitation included the study purpose, the summary interview questions, and an offer of a $75 gift card. The interviews took place in fall 2018.

We developed a 17-question interview guide for this study (Table 1). Semi-structured phone interviews lasted 30 to 45 minutes and were deidentified; we reported the data in aggregate. Each interview was transcribed and organized using QSR International’s N-ViVo 11 software.¹³ Throughout the coding process, an iterative approach was used to reexamine the original data as themes defined as codes were merged, expanded, deleted, or divided.

Results

A total of 17 individuals, representing payers, drug manufacturers, and industry consultants, responded to the study invitation, all of whom had some involvement in outcomes-based contract development or evaluation.

In all, 7 of the participants were from payer organizations, representing commercial and government health plans and pharmacy benefit managers. Of these participants, 6 had previous involvement with outcomes-based contract development or implementation, and 4 of them had experience specific to disease-modifying therapies for multiple sclerosis.

A total of 5 participants were drug manufacturer representatives. Of these participants, 4 represented drug companies with disease-modifying therapies for multiple sclerosis that had been approved by the FDA or were still in development; 3 of the representatives had experience with outcomes-based contracts.

Finally, 5 participants were healthcare consultants who had experience with payer, drug manufacturer, and outcomes-based contracts.

The study participants identified challenges that have hindered the widespread adoption of outcomes-based contracts (Table 2). Data issues represented the most frequently reported category of challenges, including selecting end points. Other cited challenges were related to confounding factors, regulatory issues, and level of risk mitigation.

In an outcomes-based contract, participants reported that both parties must agree on the end points, what data to collect, how the data will be collected and interpreted, and how much time can pass before the data will be available to payers. The first data challenges reported are identifying, defining, and agreeing on meaningful end points. The end points need to be simple, easily measurable in available data sets, objective, and clinically relevant. The payers reported a preference for end points that can be measured with pharmacy and medical claims data. In addition, the participants agreed that the parties must concur on the population to be measured and the time frame of the contract. In our study, the reported target contract durations to measure outcomes successfully ranged from 3 months to 3 years, with a median of 1 year.

The preferred number of end points varied across the respondents. Some respondents preferred multiple end points, particularly for complex diseases. Other respondents preferred keeping only 1 or 2 end points, making the interpretation of success simpler.

The frequently measured outcomes-based contract end points included medication adherence, event avoidance (eg, relapse, strokes, myocardial infarction), hospitalizations or admissions, laboratory values (eg, hemoglobin [Hb]A_1c, low-density lipoprotein [LDL] levels), and disease progression.

In the study, payers and manufacturers reported challenges with data collection and analysis. The participants noted that there needs to be an adequate data collection infrastructure in place for measuring and monitoring outcomes. They also noted that it is important that the data collected are accurate, are collected in a timely manner, and do not require a significant cost to access. Payers expressed concern about needing enough sample sizes, whereas drug manufacturers reported challenges with reconciliation and the transfer of data between parties.

The study respondents reported that end points should be related to the use of the drug and should not be influenced by confounding factors. For example, they noted that manufacturers require good medication adherence for patients’ data to be included in an outcomes-based contract. The common measures of medication adherence include medication possession ratio and proportion of days covered with a required threshold of 80%.

The respondents reported that comorbid conditions may influence the appropriate interpretation of an outcomes-based contract end point. For example, if a 6-minute walk test measures the disability level of a patient with multiple sclerosis, the test’s outcome may be affected if that patient also has rheumatoid arthritis. However, although manufacturers reported a desire to control for comorbid conditions, payers in the sample did not. Consequently, the parties may be more likely to agree on using end points that are less sensitive to comorbid conditions.

Regulatory barriers, including the Anti-Kickback Statute and statutory rates for Medicaid best price law, have limited the adoption of outcomes-based contracts according to the participants. Until appropriate “safe harbors” are developed, the widespread adoption of outcomes-based contracts will likely continue to be limited.

The respondents noted that an outcomes-based contract’s level of economic risk mitigation must be significant enough to justify the effort in its development and implementation. Payers reported that some outcomes-based contracts had a rebate level that was too small to mitigate the payer’s risk appropriately. Effort should equal reward.

Disease relapse rates are the most often used end points in outcomes-based contracts for multiple sclerosis. However, payers reported that disease relapse rates are subjective and are difficult to capture and monitor in the claims data. To identify a relapse, payers have used a hospitalization or a course of steroids as proxies. Other end points that are currently used in outcomes-based contracts for multiple sclerosis reported by participants include changes in MRI, emergency department visits, and the number of symptom-free days. End points that would be based on patient-reported outcomes, such as the number of symptom-free days, will require a method to capture this information, according to the participants.

There was skepticism among the study participants related to the value of MRI data in outcomes-based contracts in multiple sclerosis. The respondents questioned the direct correlation between MRI data and disease progression. In addition, payers reported that using MRI data as outcomes-based contract end points would increase the costs of testing, data capture, and analysis.

The respondents were receptive to the use of a blood-based clinical biomarker to measure multiple sclerosis disease activity as an outcomes-based contract end point, and they believed that the development of this type of biomarker could increase the adoption of outcomes-based contracts for multiple sclerosis. A biomarker provides an objective outcome measure that may provide timely results and be less sensitive to comorbid conditions in measuring disease activity than physician observation or patient-reported symptoms.

One payer reported that a biomarker for multiple sclerosis would “help you see in a reasonable time frame something that might otherwise be hard to measure in a short period of time.” The adoption of blood-based biomarker data for outcomes-based contracts for multiple sclerosis would require the biomarker to be established and validated as predictive of long-term clinical outcomes (eg, disability, disease progression), and its data must be made available to the participants.

Our findings showed some encouraging consensus regarding the future use of outcomes-based contracts for multiple sclerosis among payers, manufacturers, and industry consultants (Table 3). The respondents agreed that high-cost specialty medications, including disease-modifying therapies for multiple sclerosis, are particularly attractive subjects for outcomes-based contracts. The respondents also agreed that the cost of data collection and analysis needs to be sufficiently low to make the economics of the outcomes-based contract acceptable.

Discussion

Although the respondents agreed that end points must be measurable, clearly defined, objective, and realizable in a relatively short period of time, differences surfaced when the respondents reported which end points to use to measure the success or failure of a drug. Payers wanted to use claims data that were easily accessible, and they were concerned about not having access to clinical data beyond the standard medical and pharmacy claims data sets. Tracking the success of an outcomes-based contract may be more challenging when the end point, such as a laboratory value, only resides in the patient’s electronic health record. If a payer does not have access to these data, the payer must look for surrogate end points to develop an outcomes-based contract.

This approach can be problematic, because surrogate measures do not necessarily correlate with the clinical outcomes of focus. Manufacturers were concerned that other data, such as laboratory or electronic medical record data, were needed to ensure that the claims data accurately portrayed a patient’s response or nonresponse to medication therapy.

The end points in an outcomes-based contract may include patient-centered outcomes, such as disease relapse rates or cardiovascular events, or intermediate end points, such as HbA_1c values or LDL levels. Overall, the simplicity of outcomes measurement of an outcomes-based contract will be key.² Easily accessible drug outcomes measurement data that are agreed to by both parties will be a critical element of success of outcomes-based contracts and will minimize the resource utilization and cost associated with contract execution.¹²

Payers reported challenges in selecting end points for outcomes-based contracts in a progressive neurologic disorder, such as multiple sclerosis. For example, given their subjective definitions, disease relapses are viewed as an unpopular choice for an end point. Other end point candidates, such as changes in MRI over time, do not always directly correlate with multiple sclerosis disease progression,¹⁴ especially in advanced stages of the disease.¹⁵ Notwithstanding, because of the high cost of disease-modifying drugs for multiple sclerosis on the market, the number of these agents, and the variability of responses across patients, multiple sclerosis is a compelling therapeutic area for the adoption of outcomes-based contracts. Successful early control of disease progression in patients with multiple sclerosis can provide value to payers and to manufacturers and can prolong the ability of patients to remain active.⁸

An end point that can quickly measure whether a disease-modifying therapy for multiple sclerosis is meeting therapeutic goals would be a significant asset to prolonging a patient’s quality of life and slowing disease progression. The study respondents agreed that clinical biomarkers that meet specific criteria that correlate to clinical outcomes (eg, disability, disease progression) may serve as an objective basis for payers and for manufacturers to measure the success of disease-modifying therapies under an outcomes-based contract and may increase the adoption of outcomes-based contracts related to disease-modifying therapies for multiple sclerosis. Consequently, if a blood-based clinical biomarker can be identified that could accurately measure disease progression, an outcomes-based contract could be developed around this measure.

Contract duration and timing may vary depending on the target population and the specific outcome. Today, the efficacy of a disease-modifying therapy for multiple sclerosis is typically evaluated in an 18- to 24-month time frame. However, most patients are enrolled in health plans on an annual basis; an outcomes-based contract in which end points cannot be realized within a 6- to 12-month time frame may affect the value for payers when member turnover occurs. Ultimately, the payer must pay in full for the drug at the time of dispensing and may not realize any value-based outcome payment until years later. This could create a challenge from a budgeting perspective in trying to estimate the potential return on investment of outcomes-based contracts.

There was consensus by the respondents that once the sources of the data and the end points are agreed on, having the data easily accessible to all stakeholders is important. Because most laboratory data (and presumably biomarker data) are more easily obtained from the electronic health record as opposed to standard claims files, health plans would need to identify an alternate way to capture and monitor this agreed on end point, assuming they are unable to obtain the outcomes measurement directly from the laboratory tests. Accessibility of the data is critical, and all stakeholders need to have confidence that the data will be collected, shared, and interpreted in an unbiased manner.²

Some health plans have used the prior authorization and reauthorization process to capture data from the physician at agreed-on checkpoints for prescription renewals. A check on disease progression could be a required question regarding prior authorization reauthorizations. Alternatively, or in tandem, a purpose-built software platform could be engaged to allow stakeholders access to the end point measures. Payers may need to work with physician providers and laboratory vendors (eg, Quest Diagnostics or LabCorp) to gain access to relevant laboratory data. Independent third-party outcomes measurements are also an option used by some organizations to reduce bias and effort associated with the measurement of outcomes.¹⁶

The respondents in our study also agreed that it is important that the contract terms are clear and prescriptive with respect to the collection, integration, and analysis of the data. Not having a large enough sample size to achieve statistical significance may be a barrier to an outcomes-based contract that does not measure each patient’s success independently.

The drug in question must also be a primary driver of the outcome measure. The consensus among study participants was that outcomes-based contracts could include elements in the contract design to monitor and/or improve adherence. One way to incorporate medication adherence support in an outcomes-based contract would be to only include in the contract patients who met a prespecified threshold.² There was agreement among the respondents about the need to have a minimum threshold for patient adherence to therapy. There was disagreement as to whether other confounding factors, such as comorbid conditions, should eliminate a specific patient’s data from consideration.

Payers and manufacturers want to mitigate risk; outcomes-based contracts with meaningful economic consequences can mitigate the payer’s risk of paying for a drug that doesn’t work, while mitigating the drug manufacturer’s risk of not selling the desired volumes of a successful drug. Overall, an outcomes-based contract needs to be designed so that it provides benefit in some form to all parties involved, including the payer, the manufacturer, and the patient.²

Finally, the respondents agreed that regulations such as Medicaid best price and the federal Anti-Kickback Statute have presented hurdles for drug makers to engage in outcomes-based contracts with health plans. Changes to the legal infrastructure that include additional safe harbors will encourage a broader adoption of outcomes-based contracts.²

Limitations

The primary limitation of this study is its small sample size. Another major limitation is the broad nature of the questions.

Because of the low response rate (only 29% of participants who were contacted agreed to participate in our study), there is a likelihood of some element of response bias.

Finally, 2 participants reported minimal experience with outcomes-based contracts, and this paradigm is still nascent.

Conclusions

Payers and drug manufacturers in this study agreed that outcomes-based contracts are an effective vehicle to mitigate risk and deliver value for disease-modifying therapies for multiple sclerosis, but they noted that widespread adoption of outcomes-based contracts is tempered by 5 broad categories of challenges, including data-related issues, outcomes measurement and confounding factors, patient adherence to the therapy, regulatory barriers, and levels of risk mitigation. The 2 parties to the contract must agree on the data elements that define success; how these data shall be measured, stored, and accessed; and how to do all this in a cost-effective manner.

A software platform that stores and provides data reports to stakeholders and is sufficient to track an outcomes-based contract may help to reduce the cost of access to data. Manufacturers prefer that the end points used should be relatively insulated from comorbid conditions, which will likely fuel their desire to incorporate laboratory and electronic health record data into the outcome’s analyses; they are less confident that claims data alone will provide an accurate assessment of a medication’s effectiveness.

Key conditions are needed to facilitate widespread adoption of outcomes-based contracts. First, payers prefer to rely on claims data because of their accessibility and unambiguity. Second, blood-based biomarkers that meet specific criteria, accurately correlate to clinical outcomes, and are observable in a relatively short time frame may bridge this gap and may serve as effective end points for payers and manufacturers to measure the success or failure of a disease-modifying therapy for multiple sclerosis. Third, patients need to be engaged in an adherence support program to ensure that the medication regimens are being followed. Fourth, regulators, such as the US Department of Health & Human Services, will need to allow payments from outcomes-based contracts to be included in safe harbors from antikickback laws if beneficiaries are to benefit from outcomes-based contract arrangements. Finally, the level of risk mitigation in an outcomes-based contract needs to be significant for both parties to make the endeavor economically worthwhile.

These key conditions provide payers, drug manufacturers, and regulators a road map to start creating viable outcomes-based contract structures for disease-modifying therapies for multiple sclerosis and for other high-cost medications.

Funding Source

Octave Bioscience provided funding for this study.

Author Disclosure Statement

Dr Gray is an employee of Sanofi and a former employee of Octave Bioscience, which sponsored this study. Mr Kenney has no conflicts of interest to report.

References

1. CVS Health. 2018 Drug Trend Report. 2019. https://payorsolutions.cvshealth.com/sites/default/files/2018-trend-report.pdf?aliId=eyJpIjoiMzk4V2lHU09EVlwvWGZBNkkiLCJ0IjoiMEJ5YUZjUzhrQUNBbk5zMTV3eVFlUT09In0%253D. Accessed October 16, 2019.
2. Academy of Managed Care Pharmacy. AMCP Partnership Forum: Advancing Value-Based Contracting. J Manag Care Spec Pharm. 2017;23:1096-1102.
3. Yu JS, Chin L, Oh J, Farias J. Performance-based risk-sharing arrangements for pharmaceutical products in the United States: a systematic review. J Manag Care Spec Pharm. 2017;23:1028-1040.
4. PhRMA. Value-based contracts: 2009–Q2 2019. June 7, 2019. www.phrma.org/-/media/Project/PhRMA/PhRMA-Org/PhRMA-Org/PDF/PhRMA_ValueBasedContracts_Q2_2019.pdf. Accessed November 19, 2019.
5. Brow M, Macher D. Health plan interest in outcomes-based contracts increasing. July 12, 2018. https://avalere.com/press-releases/health-plan-interest-in-outcomes-based-contracts-increasing. Accessed May 28, 2019.
6. Novartis. Novartis 2017 US transparency and patient access report. 2018. www.novartis.com/sites/www.novartis.com/files/us-transparency-patient-access-report-2017.pdf. Accessed May 28, 2019.
7. Pagliarulo N. Biogen inks outcomes-based contract for MS meds. BioPharma Dive. December 6, 2017. www.biopharmadive.com/news/biogen-outcomes-based-contract-prime-multiple-sclerosis/512422/. Accessed May 28, 2019.
8. Costello K, Halper J, Kalb R, et al. The use of disease-modifying therapies in multiple sclerosis: principles and current evidence. A consensus paper by the Multiple Sclerosis Coalition. March 2014. Updated September 2018. www.nationalmssociety.org/getmedia/5ca284d3-fc7c-4ba5-b005-ab537d495c3c/DMT_Consensus_MS_Coalition_color. Accessed May 28, 2019.
9. Express Scripts. 2018 Drug Trend Report. 2019. https://corporate-site-labs-prod.s3.us-east-2.amazonaws.com/2019-09/Express%20Scripts%202018%20Drug%20Trend%20Report%282%29_0.pdf. Accessed May 28, 2019.
10. Prime Therapeutics. Prime Therapeutics signs CareCentered Contract with leading manufacturer of multiple sclerosis drugs: contract measures adherence, helps link price of therapy to clinical value. December 1, 2017. www.primetherapeutics.com/en/news/pressreleases/2017/ms-care-centered-contract-release.html. Accessed May 28, 2019.
11. Neumann PJ, Chambers JD, Simon F, Meckley LM. Risk-sharing arrangements that link payment for drugs to health outcomes are proving hard to implement. Health Aff (Millwood). 2011;30:2329-2337.
12. Garrison LP Jr, Towse A, Briggs A, et al. Performance-based risk-sharing arrangements—good practices for design, implementation, and evaluation: report of the ISPOR Good Practices for Performance-Based Risk-Sharing Arrangements Task Force. Value Health. 2013;16:703-719.
13. NVivo. What is NVivo? www.qsrinternational.com/nvivo/what-is-nvivo. Accessed June 13, 2019.
14. Sahraian MA, Eshaghi A. Role of MRI in diagnosis and treatment of multiple sclerosis. Clin Neurol Neurosurg. 2010;112:609-615.
15. Gajofatto A, Calabrese M, Benedetti MD, Monaco S. Clinical, MRI, and CSF markers of disability progression in multiple sclerosis. Dis Markers. 2013;35:687-699.
16. Mendelson D, Pearson CF. Inovalon announces agreement with Bristol-Myers Squibb to focus on real world outcomes & value-based contracting initiatives. May 19, 2016. https://avalere.com/press-releases/inovalon-announces-agreement-with-bristol-myers-squibb-to-focus-on-real-world-outcomes-value-based-contracting-initiatives. Accessed October 18, 2019.