Join ISPOR  | Sign up for mailing List  | Follow Us: LinkedIn Twitter Facebook YouTube
The Official News & Technical Journal Of The International Society For Pharmacoeconomics And Outcomes Research
President's Message

Reimbursement of New High Cost Drugs – Deal or No Deal!

Michael Barry MD, PhD, FRCPI, 2009-2010 ISPOR President and Clinical Director National Centre for Pharmacoeconomics, St. James’s Hospital, Dublin, Ireland

The efficient use of scarce health care resources Tis higher up the agenda of decision-makers than ever before given the current fiscal climate. It appears to me that we are assessing a greater number of very high cost medicines where the evidence base is not as robust as we would like e.g. we recently assessed the orphan drug eculizumab a terminal complement inhibitor for the treatment of paroxysmal nocturnal haemoglobinuria (PNH). We recommended the product should not be reimbursed as the company failed to demonstrate cost-effectiveness. We now turn our attention to the anti-human interleukin-1 beta (IL1beta) antibody certolizumab (Cimza) for the treatment of Cryopyrin-Associated Periodic Syndromes (CAPS). These products cost the taxpayer hundreds of thousands of euro per patient per year.

An understandable tension may exist between payers responsible for ensuring prudent use of the health care budget and pharmaceutical companies and patients who could derive benefit from such high cost products. To date, much of the risk in product reimbursement has been borne by the payer; however, the appearance of new, very high cost drugs frequently supported by a weaker evidence base has increased the risk of the reimbursement decisions dramatically, with obvious consequences for decision-makers and ultimately taxpayers. How can the payer possibly reconcile this dilemma of the provision of high cost technologies frequently described as ‘breakthrough’ in the popular press yet with insufficient evidence to prove cost-effective or value for money? The answer may, in part, come from the adoption of schemes that facilitate the availability of high cost drugs to (some) patients whilst limiting the budget impact, focusing the utilisation of these products to situations more likely to prove cost-effective i.e. conditional reimbursement. These various schemes have been described in terms such as ‘performance based schemes’, ‘risk-sharing’, coverage with evidence development’ etc where reimbursement is conditional on a variety of factors depending on the scheme in question.

These schemes may be non-outcomes based or outcomes based or indeed a combination of both. Non outcomes based schemes would include price volume agreements e.g. where the payer will specify the level of expenditure (attractive as budget impact is still priority for many payers). Any utilisation above this level will be funded by the manufacturer via a mechanism such as a rebate. The payer may also indicate that reimbursement is confined to specific patients, providers or clinical centres, which aims to control technology diffusion and reduce the risk of inappropriate use of the particular drug. Allocation of market share or manufacturer funded treatment initiation are other non-outcome based approaches that may be used. For those schemes that are outcomes based they may be subdivided into conditional reimbursement (which may be coverage with evidence development or conditional treatment continuation) or reimbursement with outcomes guarantee. A number of excellent publications on coverage with evidence development (many produced by ISPOR members) have appeared in the literature recently. They highlight the distinction of coverage with evidence development where the stated aim is to generate evidence to validate current reimbursement decisions and inform future decisions [1]. The comprehensive review by Stafinski et al. (2010) indicates that access with evidence development schemes with reimbursement as part of a clinical study are mainly applicable to non drug technologies such as diagnostics, surgical and non surgical procedures [2]. In the majority of these technologies uncertainty arose because of safety issues with approximately one-third relating to issues of cost-effectiveness. The schemes related to a variety of clinical scenario’s (graft versus host disease, coronary heart disease and cancer). The average duration of the studies was over four and a half years with one-third being observational studies (e.g. registries) and two-thirds in the form of clinical trials (62% RCT’s and 38% non-comparative single arm clinical trials).

For schemes which are outcomes-based, the approach most relevant to pharmaceutical products were those where reimbursement was linked to outcomes guarantee. In the majority of cases (over 80%) this was in the form of health outcomes guarantee i.e. survival, QALY or surrogate outcome (change in levels of a known biomarker). In all cases the schemes included uncertainties around clinical effectiveness. This will come as no surprise to those of us involved in health technology assessment. The review highlighted four different ‘terms of arrangement’. The two most frequent arrangements (over 85%) included those where a) payers provided the drug for a defined period of time, with manufacturers refunding the cost of the drug in patients who did not achieve the targeted health outcome e.g. bortezomib for multiple myeloma - the Velcade Response Scheme in the UK, and b) payers purchasing the drug at half the regular price for the first treatment cycle and then purchasing at full price in those who achieved the targeted outcome and continued on therapy e.g. sorafenib as second line therapy for metastatic renal cell carcinoma in Italy. Some of the schemes incorporated a financial outcomes component e.g. ranabizumab (Lucentis) for the treatment of wet age related macular degeneration and lenalidomide (Revlimid) for multiple myeloma in the UK.

It is reasonable to query whether these schemes work or not? The answer to some extent depends on your perspective. From the decision-makers standpoint with particular reference to pharmaceuticals the attractiveness of non outcomes based and outcomes guarantee schemes or a combination of same is evident and an approach which is increasingly used. But what of coverage with evidence development? In a recent article Penny Mohr and Sean Tunis provide examples of what may be considered successful coverage with evidence development schemes namely the high dose chemotherapy with autologous bone marrow transplant (HDC-ABMT) study and the National Emphysema Treatment Trial (NETT) [3]. In contrast, the UK Multiple Sclerosis Risk Sharing Scheme (MSRSS) is considered by some as an example of a scheme which has failed to deliver. This scheme was collaboration between the UK Department of Health and four companies that manufacture disease modifying drugs for multiple sclerosis (MS). Patient level disease progression data was to be collected on a cohort of approximately 10,000 patients for the purpose of assessing whether the benefit observed in phase III trials occurred in practice. Eight years later little published data is available and last month (February 2010) the MS Society urged the UK government to scrap the risk-sharing scheme which has cost some £ 350 million [4].

Clearly there are challenges in relation to the successful implementation of coverage with evidence development schemes including the necessity for the clear delineation of the decision problem with a clearly formed study objective ensuring that the scheme is designed in a manner considered fit for purpose.

Decision-makers will have a numbers of concerns not least the cost and feasibility of post launch evidence collection. Will the information be sufficient to reduce uncertainty around the reimbursement decision? Who is going to pay for the additional information? Payers will recognise that such schemes may result in significant unanticipated commitments and a further draw on scarce resources which may even increase costs in the long term. In relation to the generation of such data will it be transferable from one setting to another? What is the time frame? Private and public payers may consider such schemes differently. There is also the issue of changing the threshold for reimbursement.

Payers will be concerned that such an approach will lower the threshold, thereby facilitating reimbursement applications backed up by insufficient evidence. Frequently we are advised of the difficulty in collecting information particularly in the area of orphan and ultra orphan therapies. Payers will recall studies such as CONSENSUS I trial back in 1987 where enalapril reduced heart failure mortality by 40% in just 6 months. Yet this study only included 253 subjects! Of course, it depends on one’s perspective and colleagues from industry will express the concern that the evidence threshold may be raised with associated implications for innovation and indeed patient access. Whatever the pros and cons it appears that one of the most important aspects of coverage with evidence development schemes is governance. Credible schemes will likely require management independent of stakeholders to ensure implementation in line with the stated objective.

It appears that stakeholders agree on certain requirements such as the necessity to clearly define the area(s) of uncertainty and the resultant evidence gap together with the need for well designed studies. Areas for further consideration will focus on issues such as funding of such schemes and preferred study design. Adrian Towse and Lou Garrison (in another excellent manuscript) considered that it was too early to tell if the recent surge in performance based risk sharing agreements was likely to be a trend or only a fad and I agree [5]. However, in relation to pharmaceuticals I feel decision-makers will consider reimbursement with evidence development an exception rather than the rule. In these challenging fiscal times I believe that payers will increasingly turn to the rather less demanding combination of non outcomes based plus outcomes guarantee schemes reflecting their increased degree of risk aversion.

If you are interested in more information about risk sharing/performance-based agreements, the ISPOR Short Course, Risk-Sharing/Performance-Based Arrangements for Drugs and Other Medical Products, to be held on Sunday, May 16, 1-5 pm priorto the ISPOR 15th Annual International Meeting in Atlanta Georgia, USA or Sunday, November 7, 2010 8:00-12:00 prior to the ISPOR 13th Annual European Congress in Prague, Czech Republic, is suggested. See the ISPOR website at for registration information.

See you all at the ISPOR 15th Annual International Meeting in Atlanta, GA, USA!


  1. Trueman P, Grainger DL, Downs KE. Coverage with evidence development: applications and issues. Int J Technol Assess Health Care 2010;26:79-85.
  2. Stafinski T, McCabe CJ, Menon D. Funding the unfundable. Pharmacoeconomics 2010;28:113-42.
  3. Mohr P & Tunis S. Access with evidence development, The US experience. Pharmacoeconomics 2010;28:153-62.
  4. McCabe CJ, Stafinski T, Edlin R, Menon D. Access with evidence development, A framework for description and evaluation. Pharmacoeconomics 2010;28:143-52.
  5. 5. Towse A, Garrison LP. Can’t get no satisfaction? Will pay for performance help? Pharmacoeconomics 2010;28:93-102.

  Issues Index | 2010 Issues Index  

Contact ISPOR @  |  View Legal Disclaimer
©2016 International Society for Pharmacoeconomics and Outcomes Research. All rights reserved under International and Pan-American Copyright Conventions. 
Website design by Eagle Systems USA, Inc.