Inclusion of cost effectiveness in licensing requirements of new drugs: the fourth hurdle
http://www.100md.com
《英国医生杂志》
1 Department of Public Health and Epidemiology, University of Birmingham, Birmingham B15 2TT, 2 Centre for Health Economics, University of York, York, 3 School of Public Health, University of Sydney, Sydney, Australia,, 4 Pharmaceutical Outcomes Research and Policy Program, University of Washington, Seattle, USA
Correspondence to: R S Taylor r.s.taylor@bham.ac.uk
Increasing numbers of countries are considering cost effectiveness in decisions about which drugs to make available for prescription. How do the different approaches work and is it time for standardisation?
Introduction
Faced with greatly increasing drug budgets (see box 1) many countries, particularly in Europe, have begun to use economic evidence in national reimbursement decisions.
Europe
Probably the most important fourth hurdle development in Europe has been the technology appraisal programme of the UK's National Institute for Clinical Excellence (NICE). In its first few years, NICE has attracted much attention and criticism and is seen by some as a potential model of a pan-European fourth hurdle agency.3-5
A few countries have introduced a formal requirement for the consideration of economic evidence as part of the pricing or reimbursement decision. These include Belgium, Finland, Norway, Portugal, and Sweden. The Netherlands has indicated the intention to introduce a formal requirement but has postponed this until 2005. Just recently, Hungary has become one of the first Eastern European countries to signal the introduction of a formal requirement for economic evidence. Germany has recently established an institute, which may have an evaluation function, and there is some health technology assessment at a regional level in Spain. In several other countries, including Denmark, France, and Italy, the submission of economic evidence (by companies) is voluntary, but may be considered when it is submitted.
Box 1: Why are drug budgets rising? (adapted from Stevens et alw15)
The unit cost of new drugs has been higher than that of the drug they replace—for example, selective serotonin reuptake inhibitors cost six times as much as the older tricyclic antidepressants,w1 taxanes are several thousand pounds per patient more expensive than previous anticancer drugs,w2 w3 and two drugs for severe rheumatoid arthritis—etanercept and infliximab—can cost nearly ￡10 000 per patient for every year that they are treated.w4
Drugs are being developed for many conditions that have previously had no treatment—for example, lifestyle drugs (such as sildenafil for erectile dysfunctionw5 and buproprion for stopping smokingw6) and designer drugs often produced by the biotechnology industry to treat uncommon diseasesw7
The introduction of a drug can place a demand on the use of other expensive technologies—for example, the effective use of the anti-flu drug, zanamivir, is aided by a near patient diagnostic test to confirm flu-like illness.w8 Moreover, delivery of drugs may depend on other medical technology—for example, insulin pumps for diabetics,w9 intrathecal morphine pumps for chronic back pain,w10 and, more recently, the development of drug eluting coronary artery stents coated with anticoagulant glycoproteins.w11
United States
Like Europe, the United States has only recently started to use economic evaluation in drug listing decisions. In 1998, Regence BlueShield, a health management organisation, began requesting clinical and economic evidence from pharmaceutical and biopharmaceutical manufacturers as a condition for formulary review.6 In 2000, the Academy of Managed Care Pharmacy (AMCP), a national professional society of managed care purchasers, developed its own version of the guidelines—the AMCP format for formulary submission.7 Although an exact figure is not known, over 50 private and public sector healthcare purchaser organisations, covering well over 120 million lives, have adopted the AMCP format or a similar process.
Rest of the world
Little has been written about use of economic evaluation outside Australia, Europe, and North America. For example, Japan, despite being the second largest health care economy in the world, seems to have done little economic evaluation and currently has no system of limiting market entry of drugs based on their cost effectiveness.8
Many countries are now implementing formal or informal fourth hurdle systems that reflect the local health economy.9 10 Box 2 outlines the requirements of these systems (see bmj.com for illustration of the differing approaches to these requirements in the Australian and UK systems).
Effect of fourth hurdle
The ultimate goal of economic evaluation may be to maximise health for a given healthcare budget, but this is difficult to assess in practice. Here, we examine the effect of the fourth hurdle according to three outcomes: quality of evidence, effect on reimbursement policies, and effect on price
Quality of pharmacoeconomic evidence
The problem of quality and related biases in pharmacoeconomic studies has long been recognised. Friedberg and colleagues found that studies funded by the pharmaceutical industry were one eighth as likely to reach unfavourable qualitative conclusions, and 1.4 times more likely to reach favourable quantitative conclusions as non-profit funded studies.11 The industry related bias in economic evaluation publications has been confirmed in several subsequent publications.12
Is the quality of pharmaceutical industry economic submissions to fourth hurdle agencies any better than published studies? Hill and colleagues examined all 326 submissions made to the Australian Pharmaceutical Benefit Scheme between 1994 and 1997.13 Of these submissions, 216 (67%) were considered to present "serious problems of interpretation." The drug benefit plan committees in British Columbia and Ontario have confirmed this Australian experience. Anis et al reported that of the 32 pharmacoeconomic studies submitted to the two committees in 1996, only 21 could be used to make recommendations, the remainder being rejected because they contained incomplete or pending information.14
We were unable to find more recent publications examining the quality of pharmacoeconomics submitted to fourth hurdle agencies. However, our collective experience is that the quality of industry reimbursement submissions has probably improved. Nevertheless, complex decision analysis models are being increasingly used to support the acceptable (or not) cost effectiveness of drugs; this move is supported by the recent technical guidance released by NICE.15 The models vary in quality and many are not very transparent, making continued independent assessment of models essential within the fourth hurdle process.
Does cost effectiveness information influence reimbursement decisions?
A central premise of the fourth hurdle approach is that those drugs deemed to achieve acceptable cost effectiveness are recommended while those that fail to meet acceptable levels are more likely to be rejected. What might constitute acceptable cost effectiveness is beyond the scope of this article and has been discussed elsewhere.16
George and colleagues examined 355 submissions to the Australian Pharmaceutical Benefit System between 1991 and 1996 and found 26 that used cost per life year gained and nine that used cost per quality adjusted life year (QALY).17 Raftery conducted a similar exercise based on the decisions of the NICE technology appraisal programme up to March 2001.18 From these two studies it is possible to compare the cost per life year gained or QALY of drugs that have been recommended for funding versus those drugs that have been rejected. The threshold above which drugs were regarded as not providing good value for money seems to be $A76 000 (￡30 000, $53 500, 44 000) per life year gained and $A42 000 per QALY in the Australian study and ￡30 000 per life year gained or QALY in the NICE study.
Box 2: Stages in leaping the fourth hurdle
Statement of information needs of decision maker
Submission of evidence
Critical assessment of evidence
Appraisal or decision making
Issuance of guidance policy
Although broadly supportive of the objective of efficiency, both case series show that decision makers do not operate against a fixed willingness to pay threshold as an absolute decision rule. Other factors that can influence policy decisions include the overall budget impact of the drug (that is, the overall cost of introducing a drug to the health system); the rule of rescue (that is, funding a drug for a serious clinical condition on the grounds that alternative therapies are lacking or inadequate); and decisions around so called lifestyle drugs (such as sildenafil).
Effect of cost effectiveness analysis on price
An inevitable consequence of the global growth of the fourth hurdle has been the development of a wide range of differing systems—both in terms of processes and methods. This is exemplified by the proliferation of guidelines for economic submissions; a recent review identified over 25 different guidelines across Europe and North America.9
Summary points
Licensing of drugs has traditionally been based on quality, safety, and efficacy
Faced with increasing healthcare costs, many countries are now also requiring evidence of cost effectiveness—the fourth hurdle
The limited evidence available suggests fourth hurdle policies have contributed to more cost effective use of drugs
Increasing international harmonisation and greater openness could improve the operation of fourth hurdle systems
Several new drugs have been deemed to have acceptable cost effectiveness in one jurisdiction but been rejected in another (for example, zanamivir for the treatment of flu, riluzole for motor neuron disease, and interferon beta for multiple sclerosis). Variations in methods and processes across these countries may be contributing to these differences. A lesser but nevertheless potentially important effect of the differing fourth hurdle requirements across countries has been the inefficiency resulting from pharmaceutical manufacturers having to invest in often substantially differing submissions for a new drug in order to meet local requirements.
Although differences in decision making procedures and societal willingness to pay pharmaceutical policies between countries are likely to continue, the time seems right for a greater degree of harmonisation of methods. Drummond has recently outlined the issues to be considered in such a convergence of economic evaluation guidelines.21
Confidentiality and openness
Commonwealth Department of Health, Housing and Community Services. Guidelines for the pharmaceutical industry on preparation of submissions to the Pharmaceutical Benefits Advisory Committee. Canberra: Australian Government Printing Service, 1992. www.health.gov.au/pbs/pharm/pubs/guidelines/content.htm (accessed 5 May 2004).
Ontario Drug Benefits Program. Ontario guidelines for drug submission and evaluation. www.health.gov.on.ca/english/providers/pub/drugs/dsguide/dsguide_mn.html (accessed 1 May 2004).
Hutton J, Maynard A. A NICE challenge for health economics. Health Economics 2000;9: 89-93.
Benjamin I, Poston G, Sherlock D. Is the NICE process flawed? Lancet 2002;359: 2120.
McKee M, Maclehose L, Mossialos E. After NICE: where now for health policy in the European Union? J R Soc Med 2003;96: 1-2.
Mather DB, Sullivan SD, Augenstein D, Fullerton DS, Atherly D. Incorporating clinical outcomes and economic consequences into drug formulary decisions: a practical approach. Am J Manag Care 1999;5: 277-85.
Sullivan SD, Lyles A, Luce B, Gricar J. AMCP guidance for submission of clinical and economic evaluation data to support formulary listing in US health plans and pharmacy benefits management organizations. J Manag Care Pharm 2001;7: 272-82.
Oliver A. Health economic evaluation in Japan: a case study of one aspect of health technology assessment. Health Policy 2003;63: 197-204.
Hjelmgren J, Berggren F, Andersson F. Health economic guidelines—similarities, differences and some implications. Value in Health 2001;4: 225-50.
Guillen AM, Cabiedes L. Reforming pharmaceutical policies in the European Union: A "penguin effect"? Int J Health Serv 2003;33: 1-28.
Friedberg M, Saffran B, Stinson TJ, Nelson W, Bennett CL. Evaluation of the conflict of interest in economic analyses of new drugs used in oncology. JAMA 1999;282: 1453-7.
Lexchin J, Bero LA, Djulbegovic B, Clark O. Pharmaceutical industry sponsorship and research outcome and quality: systematic review. BMJ 2003;326: 1167-70.
Hill SR, Mitchell AS, Henry DA. Problems with the interpretation of pharmacoeconomic analyses. A review of submissions to the Australian Pharmaceutical Benefits Scheme. JAMA 2000;283; 2116-21.
Anis AH, Rahman T, Schechter MT. Using pharmacoeconomic analysis to make drug insurance coverage decisions. Pharmacoeconomics 1998;13: 119-26.
National Institute for Clinical Excellence. Guide to methods of technology appraisal. London: NICE, 2004. www.nice.org.uk/page.aspx?o=201974 (accessed 5 May 2004).
Olsen JA, Smith RD. Theory versus practice: a review of "willingness to pay" in health and health care. Health Economics 2001;10: 201-8.
George B, Harris A, Mitchell A. Cost-effectiveness and the consistency of decision-making. Evidence from pharmaceutical reimbursement in Australia (1991 to 1996). Pharmacoeconomics 2001;19: 1103-9.
Raftery J. NICE faster access to modern treatments? Analysis of guidance on health technologies. BMJ 2001;323: 1300-3.
Bureau of Industry Economics. The pharmaceutical industry: impediments and opportunities. (Evaluation report 11). Canberra: BIE, 1991.
Salkeld G, Mitchell A, Hill S. Pharmaceuticals. In: Mooney G, Scotton R, eds. Economics and Australian health policy. Sydney: Allen and Unwin, 1999: 115-36.
Drummond MF. Will there ever be a European drug pricing and reimbursement agency? Eur J Health Econ 2003;4: 67-9.
Drummond M. Should commercial-in-confidence data be used by decision makers when making assessments of cost effectiveness? App Health Econ Health Policy 2002;1: 53-4.(R S Taylor, reader in pub)
Correspondence to: R S Taylor r.s.taylor@bham.ac.uk
Increasing numbers of countries are considering cost effectiveness in decisions about which drugs to make available for prescription. How do the different approaches work and is it time for standardisation?
Introduction
Faced with greatly increasing drug budgets (see box 1) many countries, particularly in Europe, have begun to use economic evidence in national reimbursement decisions.
Europe
Probably the most important fourth hurdle development in Europe has been the technology appraisal programme of the UK's National Institute for Clinical Excellence (NICE). In its first few years, NICE has attracted much attention and criticism and is seen by some as a potential model of a pan-European fourth hurdle agency.3-5
A few countries have introduced a formal requirement for the consideration of economic evidence as part of the pricing or reimbursement decision. These include Belgium, Finland, Norway, Portugal, and Sweden. The Netherlands has indicated the intention to introduce a formal requirement but has postponed this until 2005. Just recently, Hungary has become one of the first Eastern European countries to signal the introduction of a formal requirement for economic evidence. Germany has recently established an institute, which may have an evaluation function, and there is some health technology assessment at a regional level in Spain. In several other countries, including Denmark, France, and Italy, the submission of economic evidence (by companies) is voluntary, but may be considered when it is submitted.
Box 1: Why are drug budgets rising? (adapted from Stevens et alw15)
The unit cost of new drugs has been higher than that of the drug they replace—for example, selective serotonin reuptake inhibitors cost six times as much as the older tricyclic antidepressants,w1 taxanes are several thousand pounds per patient more expensive than previous anticancer drugs,w2 w3 and two drugs for severe rheumatoid arthritis—etanercept and infliximab—can cost nearly ￡10 000 per patient for every year that they are treated.w4
Drugs are being developed for many conditions that have previously had no treatment—for example, lifestyle drugs (such as sildenafil for erectile dysfunctionw5 and buproprion for stopping smokingw6) and designer drugs often produced by the biotechnology industry to treat uncommon diseasesw7
The introduction of a drug can place a demand on the use of other expensive technologies—for example, the effective use of the anti-flu drug, zanamivir, is aided by a near patient diagnostic test to confirm flu-like illness.w8 Moreover, delivery of drugs may depend on other medical technology—for example, insulin pumps for diabetics,w9 intrathecal morphine pumps for chronic back pain,w10 and, more recently, the development of drug eluting coronary artery stents coated with anticoagulant glycoproteins.w11
United States
Like Europe, the United States has only recently started to use economic evaluation in drug listing decisions. In 1998, Regence BlueShield, a health management organisation, began requesting clinical and economic evidence from pharmaceutical and biopharmaceutical manufacturers as a condition for formulary review.6 In 2000, the Academy of Managed Care Pharmacy (AMCP), a national professional society of managed care purchasers, developed its own version of the guidelines—the AMCP format for formulary submission.7 Although an exact figure is not known, over 50 private and public sector healthcare purchaser organisations, covering well over 120 million lives, have adopted the AMCP format or a similar process.
Rest of the world
Little has been written about use of economic evaluation outside Australia, Europe, and North America. For example, Japan, despite being the second largest health care economy in the world, seems to have done little economic evaluation and currently has no system of limiting market entry of drugs based on their cost effectiveness.8
Many countries are now implementing formal or informal fourth hurdle systems that reflect the local health economy.9 10 Box 2 outlines the requirements of these systems (see bmj.com for illustration of the differing approaches to these requirements in the Australian and UK systems).
Effect of fourth hurdle
The ultimate goal of economic evaluation may be to maximise health for a given healthcare budget, but this is difficult to assess in practice. Here, we examine the effect of the fourth hurdle according to three outcomes: quality of evidence, effect on reimbursement policies, and effect on price
Quality of pharmacoeconomic evidence
The problem of quality and related biases in pharmacoeconomic studies has long been recognised. Friedberg and colleagues found that studies funded by the pharmaceutical industry were one eighth as likely to reach unfavourable qualitative conclusions, and 1.4 times more likely to reach favourable quantitative conclusions as non-profit funded studies.11 The industry related bias in economic evaluation publications has been confirmed in several subsequent publications.12
Is the quality of pharmaceutical industry economic submissions to fourth hurdle agencies any better than published studies? Hill and colleagues examined all 326 submissions made to the Australian Pharmaceutical Benefit Scheme between 1994 and 1997.13 Of these submissions, 216 (67%) were considered to present "serious problems of interpretation." The drug benefit plan committees in British Columbia and Ontario have confirmed this Australian experience. Anis et al reported that of the 32 pharmacoeconomic studies submitted to the two committees in 1996, only 21 could be used to make recommendations, the remainder being rejected because they contained incomplete or pending information.14
We were unable to find more recent publications examining the quality of pharmacoeconomics submitted to fourth hurdle agencies. However, our collective experience is that the quality of industry reimbursement submissions has probably improved. Nevertheless, complex decision analysis models are being increasingly used to support the acceptable (or not) cost effectiveness of drugs; this move is supported by the recent technical guidance released by NICE.15 The models vary in quality and many are not very transparent, making continued independent assessment of models essential within the fourth hurdle process.
Does cost effectiveness information influence reimbursement decisions?
A central premise of the fourth hurdle approach is that those drugs deemed to achieve acceptable cost effectiveness are recommended while those that fail to meet acceptable levels are more likely to be rejected. What might constitute acceptable cost effectiveness is beyond the scope of this article and has been discussed elsewhere.16
George and colleagues examined 355 submissions to the Australian Pharmaceutical Benefit System between 1991 and 1996 and found 26 that used cost per life year gained and nine that used cost per quality adjusted life year (QALY).17 Raftery conducted a similar exercise based on the decisions of the NICE technology appraisal programme up to March 2001.18 From these two studies it is possible to compare the cost per life year gained or QALY of drugs that have been recommended for funding versus those drugs that have been rejected. The threshold above which drugs were regarded as not providing good value for money seems to be $A76 000 (￡30 000, $53 500, 44 000) per life year gained and $A42 000 per QALY in the Australian study and ￡30 000 per life year gained or QALY in the NICE study.
Box 2: Stages in leaping the fourth hurdle
Statement of information needs of decision maker
Submission of evidence
Critical assessment of evidence
Appraisal or decision making
Issuance of guidance policy
Although broadly supportive of the objective of efficiency, both case series show that decision makers do not operate against a fixed willingness to pay threshold as an absolute decision rule. Other factors that can influence policy decisions include the overall budget impact of the drug (that is, the overall cost of introducing a drug to the health system); the rule of rescue (that is, funding a drug for a serious clinical condition on the grounds that alternative therapies are lacking or inadequate); and decisions around so called lifestyle drugs (such as sildenafil).
Effect of cost effectiveness analysis on price
An inevitable consequence of the global growth of the fourth hurdle has been the development of a wide range of differing systems—both in terms of processes and methods. This is exemplified by the proliferation of guidelines for economic submissions; a recent review identified over 25 different guidelines across Europe and North America.9
Summary points
Licensing of drugs has traditionally been based on quality, safety, and efficacy
Faced with increasing healthcare costs, many countries are now also requiring evidence of cost effectiveness—the fourth hurdle
The limited evidence available suggests fourth hurdle policies have contributed to more cost effective use of drugs
Increasing international harmonisation and greater openness could improve the operation of fourth hurdle systems
Several new drugs have been deemed to have acceptable cost effectiveness in one jurisdiction but been rejected in another (for example, zanamivir for the treatment of flu, riluzole for motor neuron disease, and interferon beta for multiple sclerosis). Variations in methods and processes across these countries may be contributing to these differences. A lesser but nevertheless potentially important effect of the differing fourth hurdle requirements across countries has been the inefficiency resulting from pharmaceutical manufacturers having to invest in often substantially differing submissions for a new drug in order to meet local requirements.
Although differences in decision making procedures and societal willingness to pay pharmaceutical policies between countries are likely to continue, the time seems right for a greater degree of harmonisation of methods. Drummond has recently outlined the issues to be considered in such a convergence of economic evaluation guidelines.21
Confidentiality and openness
Commonwealth Department of Health, Housing and Community Services. Guidelines for the pharmaceutical industry on preparation of submissions to the Pharmaceutical Benefits Advisory Committee. Canberra: Australian Government Printing Service, 1992. www.health.gov.au/pbs/pharm/pubs/guidelines/content.htm (accessed 5 May 2004).
Ontario Drug Benefits Program. Ontario guidelines for drug submission and evaluation. www.health.gov.on.ca/english/providers/pub/drugs/dsguide/dsguide_mn.html (accessed 1 May 2004).
Hutton J, Maynard A. A NICE challenge for health economics. Health Economics 2000;9: 89-93.
Benjamin I, Poston G, Sherlock D. Is the NICE process flawed? Lancet 2002;359: 2120.
McKee M, Maclehose L, Mossialos E. After NICE: where now for health policy in the European Union? J R Soc Med 2003;96: 1-2.
Mather DB, Sullivan SD, Augenstein D, Fullerton DS, Atherly D. Incorporating clinical outcomes and economic consequences into drug formulary decisions: a practical approach. Am J Manag Care 1999;5: 277-85.
Sullivan SD, Lyles A, Luce B, Gricar J. AMCP guidance for submission of clinical and economic evaluation data to support formulary listing in US health plans and pharmacy benefits management organizations. J Manag Care Pharm 2001;7: 272-82.
Oliver A. Health economic evaluation in Japan: a case study of one aspect of health technology assessment. Health Policy 2003;63: 197-204.
Hjelmgren J, Berggren F, Andersson F. Health economic guidelines—similarities, differences and some implications. Value in Health 2001;4: 225-50.
Guillen AM, Cabiedes L. Reforming pharmaceutical policies in the European Union: A "penguin effect"? Int J Health Serv 2003;33: 1-28.
Friedberg M, Saffran B, Stinson TJ, Nelson W, Bennett CL. Evaluation of the conflict of interest in economic analyses of new drugs used in oncology. JAMA 1999;282: 1453-7.
Lexchin J, Bero LA, Djulbegovic B, Clark O. Pharmaceutical industry sponsorship and research outcome and quality: systematic review. BMJ 2003;326: 1167-70.
Hill SR, Mitchell AS, Henry DA. Problems with the interpretation of pharmacoeconomic analyses. A review of submissions to the Australian Pharmaceutical Benefits Scheme. JAMA 2000;283; 2116-21.
Anis AH, Rahman T, Schechter MT. Using pharmacoeconomic analysis to make drug insurance coverage decisions. Pharmacoeconomics 1998;13: 119-26.
National Institute for Clinical Excellence. Guide to methods of technology appraisal. London: NICE, 2004. www.nice.org.uk/page.aspx?o=201974 (accessed 5 May 2004).
Olsen JA, Smith RD. Theory versus practice: a review of "willingness to pay" in health and health care. Health Economics 2001;10: 201-8.
George B, Harris A, Mitchell A. Cost-effectiveness and the consistency of decision-making. Evidence from pharmaceutical reimbursement in Australia (1991 to 1996). Pharmacoeconomics 2001;19: 1103-9.
Raftery J. NICE faster access to modern treatments? Analysis of guidance on health technologies. BMJ 2001;323: 1300-3.
Bureau of Industry Economics. The pharmaceutical industry: impediments and opportunities. (Evaluation report 11). Canberra: BIE, 1991.
Salkeld G, Mitchell A, Hill S. Pharmaceuticals. In: Mooney G, Scotton R, eds. Economics and Australian health policy. Sydney: Allen and Unwin, 1999: 115-36.
Drummond MF. Will there ever be a European drug pricing and reimbursement agency? Eur J Health Econ 2003;4: 67-9.
Drummond M. Should commercial-in-confidence data be used by decision makers when making assessments of cost effectiveness? App Health Econ Health Policy 2002;1: 53-4.(R S Taylor, reader in pub)