An official website of the United States government
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Kohlhammer VW, author; Schildmann J, Buch C, Zerth J, editors. Defining the Value of Medical Interventions: Normative and Empirical Challenges [Internet]. Stuttgart (DE): W. Kohlhammer GmbH; 2021.
Limited resources for healthcare need to be allocated effectively and efficiently and in accordance with the respective value of medical interventions. Recent developments in the pharmaceutical market and a rise in the cost of drugs pose a challenge for healthcare systems. This is particularly the case for so-called “personalized medicine” and drugs for small numbers of patients (e.g. orphan drugs). Research and development costs are high, while there are risks in terms of financial return, given that these drugs only address a very small patient population. Challenges to assess the value of these drugs are considerable, especially considering the lack of robust data due to small clinical studies. Accordingly, there is a high uncertainty regarding the real-world effectiveness, leading to significant discrepancies between the pharmaceutical company's list price and the cost payer's willingness to pay.
One approach to control pharmaceutical expenditures in these cases is value-based pricing. On a microperspective level, this is employed as risk-sharing agreements between payers and pharmaceutical companies, whereby the uncertainties, i.e. risks, regarding the clinical and economic performance are shared and the remuneration for the drug is dependent on its real-world value. Depending on the inclusion of an outcome element into the scheme, we can distinguish financial-based risk-sharing schemes and performance- or outcome-based risk-sharing agreements.
This chapter will analyse the strengths and weaknesses of performance- or outcome-based risk-sharing agreements (PBRSA). In addition to a theoretical analysis, we will substantiate it by reference to a case study of a PBRSA involving the evaluation of drugs for a multiple sclerosis risk-sharing scheme in the United Kingdom.
Some medical drugs come at an extraordinary price. One of the more recent examples is Onasemnogen Abeparvovec (Zolgensma®), a gene therapy for the neuromuscular disorder spinal muscular atrophy. It was approved by the United States Food and Drugs Administration in May 2019 and induces high costs considering direct drug treatment and non-treatment costs (Pearson et al., 2019, p. 1302). Given that the drug was the most expensive medical therapy in the world at the time of approval, it attracted a lot of attention in discussions by experts and the public. However, the drug example named, which is supposed to be administered only once to the patient, represents only one side of higher expenditures for pharmaceuticals. There are other pharmaceuticals that induce high expenditures because of an enduring level of the prescription itself or changes in prescribed medications (e.g. substitutions from low-dose approaches to higher-dose ones) (Lohmüller et al., 2019).
In Germany, the expenditures of the German Statutory Health Insurance for pharmaceuticals are close to those for outpatient care. This means that pharmaceuticals already belong to the most expensive service areas (GKV-Spitzenverband, 2019, p. 4; Pfannstiel et al., 2019, p. 313). These expenditures have been increasing constantly over the last few years (GKV-Spitzenverband, 2019, p. 7). We have to differentiate two end-points of a continuum to specify potential reasons for higher expenditures, as has been already mentioned above. The increase of expenditures may be generated by (1) a rise in the number of units of drugs or (2) the proportion of drugs that can be connected with a higher price compared to permissible comparator therapy (Evaluate, 2019; Statistisches Bundesamt, 2019, p. 44). In this respect, new and possibly innovative pharmaceuticals are cost drivers because of two different reasons that may sometimes interact. First of all, higher expenditures may result from targeting a large number of patients that can be addressed by the medical innovation (population approach). Secondly, as it is the case with Onasemnogen Abeparvovec, a pharmaceutical solution has been developed for a very small population whereby pharmaceutical companies aim at recovering their higher development costs by employing price mark-ups.1 Further examples are so-called orphan drugs, which are drugs for very rare diseases, and targeted drugs, which are used, for example, for cancer patients with a certain genetic mutation (Pfannstiel et al., 2019, pp. 313–314).
Rising costs for medical drugs point to fundamental issues: The resources that are available within a healthcare system based on solidarity have to be restricted and limited. Consequently, there is an ongoing societal debate and consensus needed to find an appropriate match that citizens could claim having access to a certain level of healthcare and healthcare innovations, and, concomitantly, the additional financial burden for those citizens may only increase moderately.2 At the same time and while considering the difficulties in stemming the rise of pharmaceutical expenditures, it is important to keep or build incentives for the development of new effective medical drugs. While there are limited data available on the costs of drug development (Morgan et al., 2011, pp. 10–11), it seems fair to assume that the research and development of new drugs in many cases needs large financial investments combined with a long and uncertain research and development process. This is especially the case for those fields of applications where only a small number of beneficiaries could be targeted, such as in the case of rare diseases and those with an unmet medical need, i.e. when there is a higher investment risk. Therefore, while payers are searching for cost-containment measures, it is also important to consider options to keep pharmaceutical companies encouraged to be innovative (Danzon, 2018; Pfannstiel et al., 2019, p. 308). In the following, we provide a brief account of some of the price-regulations schemes used in different healthcare systems to reconcile the aforementioned interests of payers, the pharmaceutical industry, patients and other stakeholders.
Referring to the challenges mentioned above, healthcare systems in industrialized countries have to find an appropriate trade-off between controlling expenditures within the regulated benefit basket and fostering innovation and ensuring access to new diagnosis and treatment options (Levaggi, 2014, p. 69).3 Employing drug price regulation is one of the standard approaches employed to meet that optimisation problem and is directly connected with elaborating some forms of a value for money measurement. One specific strategy is based on so-called value-based pricing.
Value-based pricing approaches are rooted within a legitimation of a price by its impact on a perceived or estimated value to the patient.4 Different strategies of value-based pricing on the macroeconomic level have been implemented. An incremental cost effectiveness ratio is used, for example, to give an appraisal for developing a price cap for pharmaceutical reimbursement in the United Kingdom (UK), as a Beveridge-type country (OECD, 2010, p. 166). Bismarckian-type countries, such as Germany, believe in the idea that decentralized regulated competition is an additional means for elaborating “value for money”, fostering variations in preference and willingness to pay. However, regardless of the macroeconomic approaches chosen by different countries, fostering the development of new drugs while bearing in mind limited resources is the definite optimisation problem which each healthcare system faces.
Against the background of rising costs for medical drugs and limits of established models for financing new drugs, we will focus in the following on risk-sharing agreements utilising value-based pricing which have been developed as one standardized approach to control pharmaceutical expenditures.
Risk-sharing agreements can be defined as
agreements concluded by payers and pharmaceutical companies to diminish the impact on the payer's budget of new and existing medicines brought about by either the uncertainty of the value of the medicine and/or the need to work within finite budgets. (Adamski et al., 2010)
According to the logic of risk-sharing agreements, pharmaceutical companies grant some kind of warranty for the value of a medical drug. The company and the payer both have different obligations depending on the occurrence of an agreed condition (Adamski et al., 2010; Renze-Westendorf, 2010, p. 206).
Participants of risk-sharing agreements acknowledge that there are uncertainties regarding developing and marketing a medical drug, i.e. risks regarding the clinical and economic performance in the real world. It is not clear whether the effectiveness of a medical drug in clinical reality, particularly at the time of its admission, reflects the efficacy shown in clinical trials (Antonanzas et al., 2011, p. 399, 2011, p. 399; Garrison et al., 2013, p. 704). These uncertainties are especially considerable for new, innovative and expensive drugs that only address small target populations. The reason for this is that clinical studies are small in such cases and, therefore, the accompanying evidence is sparse. Accordingly, there is an outcome uncertainty in terms of the patients’ response to treatment and how that will translate into health outcomes and resource utilisation, as well as a subgroup uncertainty, i.e. which patients of the heterogeneous population should and do get treated (Carlson et al., 2010, p. 188).
From the payer's point of view, these uncertainties are even more pronounced by the inherent characteristic of drugs as a post-launch “experience” good. Consequently, there is a specific risk of an ongoing asymmetry of information between the pharmaceutical company and different cost-payers considering the results of the implementation of the new drug, especially the number of prescriptions by doctors and the valuation of the beneficiaries. Controlling the expected profit margin in these cases is very complex for the pharmaceutical company from an ex ante perspective. If payers are additionally risk-averse, i.e. they have a greater fear of incorrectly paying than not paying for a cost-ineffective technology, they will insist on paying only for the (most) effective healthcare interventions within their limited budget (Towse and Garrison, 2010, p. 94).
Considering that assumption, pharmaceutical companies run the risk that the real-world value of a manufactured drug is underestimated. In the worst case, the payer refuses to adopt the drug at all (Garrison et al., 2013, p. 704). In addition, there may be high development costs and manufacturing overheads for a potentially curative or, at least, disease-modifying therapy that is only applicable to a small market and must be compensated by high list prices (Editorial, 2019, p. 697).
Pharmaceutical companies and cost-payers have different priorities regarding the payment because they have different information and views about the value of a new drug and the allocation of the accompanying risks. However, one potential shared interest is an objective assessment of benefits and costs. Obviously, such a value-oriented assessment is especially important for very expensive pharmaceuticals because there is a lot at stake for both sides: The pharmaceutical company wants its expenses for research and development to be compensated, while the payer needs to weigh the new drug against other therapies they may no longer be able to fund. In particular cases in which evidence is limited at the time of admission of the drug, for example, due to small studies for an orphan disease, it may be possible that the payer and manufacturer cannot align their expectations of remuneration and the company decides to drop out rather than to accept the price of the payer.
In order to avoid this situation, risk-sharing schemes enable the risk mentioned to be shared between payer and manufacturer by making the value of the drug, i.e. the price or remuneration, dependent on the value of the product, i.e. the future proven effectiveness in the real world (OECD, 2010, p. 170). Hereby, the payments to be made will be fixed ex ante and are contingent on information that will be collected ex post (Antonanzas et al., 2011, p. 400). Such value-based schemes may not only help to control pharmaceutical expenditures without negatively impacting the patient populations but also enable discounts without changing the high list price. This is relevant from the perspective of the pharmaceutical company, because the list price serves as an international comparison and changes may affect manufacturers’ global revenues (Carlson et al., 2010, p. 188; Towse and Garrison, 2010, pp. 95–96).
Risk-sharing agreements can be distinguished according to either financial/financial-based schemes or outcome/performance-based models (Adamski et al., 2010).
Finance-based schemes do not usually take the patient outcome into account but concentrate more on keeping the expenditures within agreed limits (Adamski et al., 2010). Some examples of this risk-sharing approach are price volume agreements or budget impact schemes, according to which the unit price of a product is linked to the volumes sold. An increase of the volume of units results in a declining cost per unit. This is especially important when there is a possibility that the new medicine will be prescribed in a wider population than anticipated (Adamski et al., 2010; OECD, 2010, p. 170).
Another form of financial-based risk-sharing is patient access schemes, which seek to enhance the value of new drugs and improve the possibility of their funding. They involve either the use of a drug for free or with financial discounts for an agreed period of time, or they focus on controlling the financial impact from an individual patient's perspective in the form of price capping schemes. The latter are connected to a specific outcome element, for example, when the drugs are provided for free once patients have exceeded a specified number of units but need more to support a certain state of health. Such an arrangement prevents the payer from spending more than a fixed amount per patient (Adamski et al., 2010; OECD, 2010, p. 171).
Performance-based risk-sharing schemes – also called performance-based-risk sharing agreements (PBRSA) – address the optimisation of definite healthcare expenditures as well. However, different from financed-based models, these schemes are dependent on the generation of evidence regarding the real-world impact of a given drug on patients’ health. The remuneration depends on the (real-world) effectiveness of the drug according to data collection subsequent to the admission of the drug. Remuneration is determined either directly by a pre-arranged rule or indirectly through an agreement regarding the option to renegotiate prices if a certain condition has been met (Adamski et al., 2010; Garrison et al., 2013, p. 705). The outcomes, which serve as a basis for the remuneration, may be defined either in terms of outcome-related benefits (e.g. clinical response) or cost-effectiveness (e.g. cost per quality-adjusted life year5), each at the individual patient level or the aggregated level of the whole population treated (OECD, 2010, p. 172).
There are different types of PBRSA: Using outcome guarantees, the treatment costs for patients not reaching a predetermined response are (fully or partially) paid back by the manufacturer. One prominent example which has been discussed in this regard is the coverage for CAR T-cells (Jørgensen et al., 2020). Another type of PBRSA involves coverage with evidence development. This approach means that there is access to a new drug while evidence is generated within a given period of time. However, reimbursement may change during the course of this time depending on the findings regarding predetermined health outcomes. Finally, PBRSA involving conditional treatment continuation means that reimbursement only takes place if patients achieve a previously defined level of response (Gonçalves et al., 2018).
Implementing a successful PBRSA needs some serious preliminary considerations. In particular, the costs of generating additional evidence or information on treatment response must be weighed against the benefits that better resource allocation decisions bring, which is basically an investment decision. All parties have to decide beforehand whether a PBRSA is acceptable to them in the given situation and with the given risk. In addition, there is the need for consideration regarding how the PBRSA should be implemented and evaluated to be successful. Among the many questions which must be answered are the choice of the appropriate study design and outcome parameters, consideration about who will measure these parameters and in which time frame, what the reimbursement modalities will look like and a lot more (Garrison et al., 2013, p. 709). These questions are raised by different interests that have to be embedded within the PBRSA.
The PBRSA approach seems appropriate if there is a significant uncertainty regarding the effectiveness of a given drug, while, at the same time, there is the desire to provide patients with access to its potential benefit. Therefore, it is especially interesting for funding new and innovative medicines such as orphan drugs and gene therapies. In contrast to this, finance-based schemes are more useful for generics, where one assumes that the outcome is already known and does not need to be considered (Carlson et al., 2010, p. 180; Gonçalves et al., 2018).
A second interesting feature of PBRSAs is that there is the possibility of generating additional evidence. The latter is a desirable public good since it is not only valuable regarding the certain drug to be assessed, but it also offers insights into the respective disease itself and, therefore, enables a general improvement of patient care.
Given the preceding analysis, PBRSAs seem potentially desirable from an ethical and societal perspective. However, it is necessary to analyse the advantages and barriers of a PBRASA to determine its value and possible limitations in practice (Garrison et al., 2013, pp. 717–718). The multiple sclerosis (MS) risk-sharing scheme, which has been established in the UK, is an illustrative example to learn about a PBRSA in practice and its associated strengths and weaknesses. This case example will be described and, subsequently, the benefits and problems with implementation of PBRSAs in practice will be discussed in the following.
Treatment for the chronic disease relapsing-remitting MS was restricted to anti-inflammatory drugs, such as cortisone, and symptomatic treatment up until the mid-1990s. Treatment options changed when the first disease-modifying therapies, interferon beta and glatiramer acetate, were developed. While these drugs do not cure the disease, they at least reduce the number of relapses. Four different pharmaceutical companies had licensed respective products (three interferon beta products and one glatiramer) within a rather short time frame. The problem with those drugs from a payer perspective was that it was not possible to predict their long-term effect from their short-term clinical achievements shown in clinical studies. Therefore, the National Institute for Health and Wellbeing came to the conclusion in its initial assessment in 2002 that the disease-modifying therapies were not cost-effective and should not be funded by the National Health Service (NICE, 2002).
Several stakeholders were obviously disappointed by this decision and campaigned against it. One argument was that cost-effectiveness may be derived if the short-term successes could also be proven in the long-term. Therefore, the first risk-sharing scheme in the UK was established to examine the long-term development of and to enable patients’ access to the therapies.
The subject of the scheme were the four MS drugs. The stakeholders involved were the respective pharmaceutical companies, the UK Department of Health and representatives of patients and health professionals. Around 5000 patients in 70 MS specialist centres across the UK were recruited in the scheme within three years. They were monitored over ten years by capturing their Expended Disability Status Scale status every year. The natural development of the disease was projected as a comparator, based on data from a historic control group, to assess the impact of the therapies. Regarding reimbursement, the stakeholders agreed on a cost-effectiveness threshold of 36,000 £ per quality-adjusted life year, which had to be kept. Therefore, the pharmaceutical companies lowered the price of their products right at the beginning and agreed that the impact of the therapies would be analysed every two years. The health outcome (effect of the drug on the progression of disease) was determined for any drug evaluated. In the case of missing the predetermined target outcome, the price of the respective therapy would be further lowered (Department of Health, 2002, pp. 7–14).
According to this agreement, the long-term clinical and outcome impact and cost-effectiveness of the disease-modifying therapies were examined. Six years after the start of the scheme, the findings about the longer-term effectiveness of the drugs were still mixed (Palace et al., 2015, pp. 502–504), but the additional data gathered during the scheme showed that there was a clinically significant treatment effect maintained at 10 years, reducing the progression of the disease, decreasing over time. If this effect was maintained over 20 years, the cost-effectiveness target would be reached (Palace et al., 2019, pp. 257–259).
The strengths and limitations to implement this approach to share the risks of new and costly drugs will be discussed in the following based on the example of the MS scheme and the preceding theoretical analyses.
One of the positive experiences of the MS risk-sharing scheme was that it had a strong impact on the care of patients with MS in the UK in addition to the generation of evidence on the effectiveness and cost-effectiveness of the drugs involved. Most importantly, the MS specialist centres, which had been partially newly established at the time of the scheme, have now build a strong network improving the support of and care for MS patients in terms of quantity and quality. In addition, many new MS therapists, nurses and doctors have been trained and educated since this has also been part of the companies’ obligations. It should be noted that many patients were getting access to the therapies due to the scheme which would otherwise not have been possible. In addition to the change of structural aspects relevant to the high quality of care for patients with MS, the data generated within the scheme over a long time period offers valuable insights into the disease itself for all stakeholders, such as its long-term development, and enables improved care (Boggild et al., 2009).
Apart from these advantages shown in the real world, theory hints at even more benefits. First of all and relevant from the perspective of payers, they allow patients’ access to innovative medicines and, therefore, to a broader range of treatment options and potential health benefits. Furthermore, pharmaceutical companies are encouraged to develop new drugs, which contribute to the health of those patient populations where health gain, hence value, is greatest. For the payers, the additional evidence generated in the course of a PRBSA decreases uncertainties about effectiveness and informs decisions about allocation within a limited budget. On the other hand, the pharmaceutical companies gain faster market access for their innovative medicines, because they can prove their real-world value over time. Since the terms of agreement between payer and manufacturer are usually confidential, this offers the possibility of hidden discounts and, therefore, remains neutral regard list prices, which are relevant for the global market. Another advantage is that PBRSAs allow the definition of patient groups which are likely to benefit from the treatment. Thereby, both manufacturers and payers reduce the risk of using the drug in patient groups not likely to profit from the new treatment (Adamski et al., 2010; Gonçalves et al., 2018).
Given the possible strengths of PBRSAs, they seem an “understandable and logical response to increasing pressure for greater evidence of real-world effectiveness and long-term cost-effectiveness for new medicines” (Garrison et al., 2013, p. 717). However, they do not come without disadvantages and barriers for implementation. Firstly and relevant for the pharmaceutical companies is that they have to show their hypotheses of the suggested treatment effect in a real-world environment, which means there is new risk of calculating their contribution margin. This perspective may already be sufficient to hinder the implementation of a PBRSA. Even if the parties agree in principle regarding a PBRSA, implementing and monitoring PBRSAs cause financial and administrative burdens. It is important to evaluate these investment costs against the potential benefits of PBRSAs and, of course, to find a fair solution regarding carrying the financial and administrative burden.
One challenge is the difficulty of defining appropriate performance indicators that are easily measurable and, at the same time, adequate to demonstrate effectiveness. Similar to clinical studies, there may be situations in which surrogate parameters may be easy to collect, whereas direct indicators to demonstrate effectiveness, such as overall survival in cancer care, are difficult to obtain – particularly in diseases with a longer course. An additional barrier for the payer may be an insufficient infrastructure for collecting, analysing and monitoring data. There may also be ethico-legal challenges concerning data protection and the need to find agreements for further proceedings after the agreement about a scheme ends (Adamski et al., 2010; Garrison et al., 2013, p. 718; Gonçalves et al., 2018; Lorente et al., 2019, p. 30; Neumann, 2013, p. 701).
Further problems and concerns can be identified regarding the case example of the MS risk-sharing scheme in the UK. One important point of criticism in this concrete example was that the historic control group did not represent the actual state-of-the-art regarding the treatment of MS prior to the start of the scheme. In addition, there were questions concerning the neglect of quality standards. Furthermore, measuring the impact of the treatments using only the Expended Disability Status Scale scores was shown to be difficult given the heterogeneity of the presentation of MS and the differing courses of progression. Additionally, many reasons led to the delay of evidence generation. The long observation period caused difficulties as well, for example, many administrative challenges. Last but not least, new medicines against MS were developed during the course of the scheme, therefore, when the final results of the scheme came out they had mostly already been outdated (Adamski et al., 2010; Boggild et al., 2009; Palace et al., 2019, pp. 258–259).
Nevertheless, it was possible to find remedies to some of these problems during the scheme, for example, by adapting the control group and the research methodology (Palace et al., 2015, pp. 502–504). Moreover, even with the development of new treatment approaches during the scheme, the results generated have important consequences for assessing the cost-effectiveness of present and future MS drugs (Palace et al., 2019, pp. 257–259).
The application of PBRSAs keeps promises as well as the risk of failing, as has been shown in theory and in the case study included (Antonanzas et al., 2011, p. 393). Correct planning seems one key to the successful implementation of PRBSAs. This includes detailed considerations regarding unambiguous and easily measured effectiveness criteria, transparency and ethical considerations, as well as staffing and funding considerations. The PBRSAs should only be considered when there are explicit and transparent objectives and scopes, when the new drug is a novel treatment in a high priority disease area with only a few or no effective alternative treatments, and when a likely health gain can be determined within a limited amount of time. Alternatively, PBRSAs should be rejected when effective and low-cost treatment standards already exist or when health authorities could end up funding a substantial part of the new drug's development costs. The high administrative burden must be weighed against the likely health and/or financial benefits. The patients’ compliance must always be considered and addressed in the scheme proposed (Adamski et al., 2010)6.
At present, it is unclear whether and where PBRSAs can really be embedded within the beneficiaries’ focus (Levaggi, 2014, p. 72; Neumann, 2013, p. 702). However, they may be become more prevalent in the future, given the fact that there is lack of robust evidence according to which it is possible to predict real-world effectiveness at the time of admission for many drugs in the times of “personalized” or “precision” medicine. The distinction here between Bismarckian- and Beveridge-type healthcare systems may become very interesting: Considering the baseline philosophy of Bismarckian-types, PBRSAs may be interpreted as an further attempt to use controlled selective contracting in order to collect appropriate information about value for money. Considering current discussions on converging national high technology assessment approaches and methods, PBRSAs could play an important role within a European method box.
Differentiation of high-risk approaches from the population approach could be beneficial referring to traditional public health considerations Edwards and Atenstaedt (2019, p. 11).
See the chapter by Stutzin Donoso in this publication for a general discussion of the concept of health and right to healthcare.
The chapter by Alex in this publication also deals with the question of sustainable healthcare.
Further discussions on different judgements of value from healthcare interventions can be read in Vermeulen and Krabbe (2018). See the chapter by Steigenberger et al. in this publication for information about how to integrate the patients’ perspective into the value assessment of medical interventions.
The cost-effectiveness approach is further discussed in the chapter by Himmler in this publication. See the chapters by Mitchell and Ubels in this publication for more information about quality-adjusted life years and their alternatives.
How to increase the patient compliance by implementing shared decision-making is discussed in detail in the chapter by Napiwodzka in this publication.
This is an open access article licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed.
Monographs, or book chapters, which are outputs of Wellcome Trust funding have been made freely available as part of the Wellcome Trust's open access policy
Your browsing activity is empty.
Activity recording is turned off.
See more...
