|
Centralized Drug Review Processes In Australia, Canada, New Zealand, And The United Kingdom
Many countries have centralized the clinical and economic assessments necessary for evidence-based drug coverage policy. We analyze such processes in Australia, Canada, New Zealand, and the United Kingdom. These countries apply comparable approaches to the assessment and appraisal of evidence but apply the processes to different types of drugs and use the reviews within different decision-making contexts. Review processes applied to all medicines and clearly tied to coverage decisions appear to influence national drug use. Rigor of process and transparency of data and rationale are believed to be important for maximizing the impact and political acceptability of the processes.
POLICY MAKERS FACE TOUGH CHOICES when making drug coverage decisions. In a world of limited resources, the ever-increasing availability of medicines of varying cost and clinical benefit forces drug benefit providers to set limits on which products to cover and under what circumstances. Subsidies would ideally be allocated in light of the best available evidence concerning drugs’ comparative impacts on relevant health outcomes and costs.1 The information required for such evidence-based coverage goes beyond that collected by regulatory agencies such as the U.S. Food and Drug Administration (FDA). Consequently, to inform drug coverage decisions, many countries have established additional national processes for the critical review of comparative clinical and pharmacoeconomic data. We describe such processes in Australia, Canada, New Zealand, and the United Kingdom and discuss common features of and challenges faced by centralized drug review processes.
Centralized drug review processes involve two stages: the initial assessment of evidence; and the subsequent appraisal of that evidence, which leads to a coverage recommendation.2 The first stage, initial assessment, involves expert evaluation of the quality and meaning of scientific evidence submitted by manufacturers or gathered by other means (for example, by systematic review). This includes clinical and economic evaluation to determine the probable impact of the product on patients’ health, drug costs, and the overall health care system, each relative to appropriate treatment comparators. Assessments also indicate the degree of uncertainty concerning anticipated long-term outcomes in practical settings, which is often significant when a product is first marketed. Clinical and economic evaluations are typically conducted separately. Because the quality of economic analysis rests on the validity and reliability of clinical data, clinical experts may review certain assumptions of economic models. The evaluators may reside within the centralized agency or, as is more common, may be drawn from academe and the health professions. Assessments provide answers to what is likely to happen if a drug is applied in various patient populations. The second stage, appraisal of the evidence, involves making recommendations about which drugs ought be funded, for whom, and under what circumstances. Appraisal requires a balance of clinical expertise, informed judgment about the reliability of data and modeling, and information about the priorities and values of populations affected by funding decisions. Appraisal necessarily involves choices: In the face of limited resources, not all "effective" medicines can be funded. Typically, a committee independent of the scientific assessment conducts the appraisal stage. This separates the positive and normative functions of the drug review process, potentially reducing scientific bias while avoiding the potential pitfalls of purely technocratic decision making. In some processes, the appraisal committee only makes recommendations to drug plan managers who retain final decision authority—further separating technology assessment from priority setting. Other appraisal committees make decisions that are tied more directly to coverage or price negotiations. However, even in the latter case, recommendations may require endorsement from government officials or a board that is ultimately accountable to the population served by the decisions. Reviews of medicines are typically conducted for the purposes of establishing coverage under a "positive" formulary. Such formularies list medicines that are covered under a given drug program; they also provide information about the subsidy level and conditions under which it is offered. Drugs not listed on a positive formulary are not generally covered. In these environments, products must be reviewed for coverage to be obtained. Other drug review systems—notably that in the United Kingdom—operate within the context of a "negative" formulary. In such situations, virtually all medicines are covered unless listed on the negative formulary or otherwise restricted by local decisionmakers. Centralized reviews help determine what should or should not be placed on the negative list, assist local decisionmakers with their coverage decisions, and help establish a minimum standard of coverage that must be provided by all regions. Drugs needn’t be formally reviewed to be eligible for coverage within systems that function with negative formularies. As products are reviewed in a centralized process, information is often exchanged between the reviewing agency and the drug manufacturer. Agencies may also provide opportunities for other stakeholders to give feedback on assessments conducted at intermediate stages of the review process. The degree of feedback by consumers and other interest groups depends largely on the type of information that the review agency is permitted to disclose. Finally, when a decision is published, stakeholders might have the right to appeal. These appeals are generally limited to claims of procedural failure or the introduction of new data, rather than complaints about the nature of the final decision.
Australia. The Australian government provides virtually all residents with a subsidy for drugs listed on a national, positive formulary: the Pharmaceutical Benefits Scheme (PBS). The 1953 National Health Act that established the PBS also created the Pharmaceutical Benefits Advisory Committee (PBAC), a statutory body that reviews all applications for PBS listing. The PBAC makes listing recommendations to the minister for health and ageing, who is directly responsible for coverage decisions but cannot list a drug without a positive PBAC recommendation to do so. The PBAC membership includes consumers, health economists, practicing community pharmacists, general practitioners, clinical pharmacologists, and specialists. It has two subcommittees that conduct specific assessment components of the overall review process. The Economics Sub-Committee assesses the economic analyses required of all applications to the PBS. The Drug Utilisation Sub-Committee assesses forecasts of drug use. Manufacturers are invited to comment on these subcommittees’ assessments. Assessments are then considered by the PBAC, which may seek further outside expert opinion in its appraisal process. After appraising the evidence, the PBAC makes a listing recommendation. If the health minister accepts a positive PBAC recommendation, the drug is referred to the Pharmaceutical Benefits Pricing Authority, a separate organization that negotiates prices with manufacturers. Since October 2005, public summary documents have been published on the Internet, providing the final coverage decision, its rationale, and summary clinical and cost-effectiveness data. Where there is disagreement with final decisions, manufacturers may resubmit an application to the PBAC or seek an independent review. Independent review outcomes are reported back to the PBAC, which makes the final listing recommendation. The total submission process is on a seventeen-week cycle, excluding price negotiations. The PBAC reviews more than 100 drugs per year, including generics. Canada. Governments fund approximately half of prescription drug purchases in Canada through a patchwork of sixteen federal, provincial, and territorial drug plans. All of these drug plans operate their own positive formularies. In 2003 the Common Drug Review (CDR) was implemented to coordinate reviews of applications for drug coverage on behalf of all government drug plans except for the provincial drug plan in Quebec. The CDR reviews only new chemical entities (NCEs) and new combination products, leaving the review of generics and new dosage forms of existing medicines to the individual drug plans. The CDR’s recommendations are not binding; each participating drug plan retains independent authority over its formulary listing decisions. The Federal/Provincial/Territorial Conference of Deputy Ministers of Health appoints an eleven-person committee of experts—the Canadian Expert Drug Advisory Committee (CEDAC)—to conduct appraisals and make the CDR recommendations. When a manufacturer submits an application to the CDR, individual reviewers are contracted to conduct preliminary assessments of clinical and economic evidence. Reviewers can be internal or external to the CDR (predominantly from academe). Manufacturers are not told the identity of the reviewers but are given the opportunity to comment on the assessment documents. With assessment documents and manufacturers’ comments, CEDAC appraises the evidence and makes a single listing recommendation. The initial recommendation and rationale are sent to the manufacturer and participating drug plans in confidence. The manufacturer may appeal for reconsideration. If reconsideration is not requested, a public "notice of final recommendation" is issued. A one-page summary of the recommendation and rationale is posted on the CDR Web site, although none of the data or assessment documents are made public. The entire CDR process takes twenty to twenty-six weeks. Approximately twenty-five drugs are expected to be reviewed annually. New Zealand. Residents of New Zealand receive public subsidy for prescription drugs through their District Health Boards. The Crown entity PHARMAC (the Pharmaceutical Management Agency) was established in 1993 to manage a national, positive formulary: the Pharmaceutical Schedule. The Pharmacology and Therapeutics Advisory Committee (PTAC) reviews all new drug applications that come to PHARMAC and provides PHARMAC with expert advice on managing the Pharmaceutical Schedule. PHARMAC is not bound by PTAC recommendations. The PTAC comprises registered medical practitioners nominated by professional bodies and appointed by the director-general of health. It has numerous subcommittees with expertise in particular clinical areas and can also consult with outside experts. Assessments of evidence are conducted by these expert subcommittees but are not made public for comment. However, the minutes of PTAC meetings are available on PHARMAC’s Web site, and the minutes of subcommittee meetings are available on request under New Zealand’s Official Information Act. The PTAC bases subsidy recommendations by considering the evidence against a set of nine decision criteria concerning population health needs and the expected benefits of medicines in community use. The PTAC recommendation is not a final listing decision. PHARMAC further appraises the PTAC recommendation in light of cost-effectiveness criteria. Specifically, because PHARMAC operates within a fixed budget for expenditure on medicines, consideration is given to other drugs that must be forgone or price concessions that must be obtained to fund new medicines. PHARMAC therefore takes the PTAC recommendation and negotiates with the supplier to reach a provisional listing agreement. If an acceptable proposal is achieved, the Pharmaceutical Schedule is updated. There is no formal appeal mechanism for PHARMAC decisions; however, manufacturers may resubmit applications with additional information. There is no predetermined timeline for a PTAC review. Approximately thirty to forty reviews are conducted each year. United Kingdom. Local Primary Care Trusts (PCTs) of the National Health Service (NHS) of England and Wales subsidize pharmaceuticals for regional populations. The NHS does not have a national list of drug benefits; rather, it has a "negative list" of drugs that are excluded from NHS subsidy. The National Institute for Clinical Excellence (NICE) was established in 1999 to address variations in health care across regions that resulted from regional discretion regarding which technologies to cover. PCTs are required to fund all medical technologies (including drugs) reviewed and recommended by NICE. NICE only reviews technologies likely to have major health implications, budgetary impact, or controversy over effectiveness. The Technology Appraisal Committee (TAC) appraises the evidence and makes recommendations regarding the drugs’ clinical and cost-effectiveness. NICE is not bound by these recommendations in constructing its final "guidance" for the NHS. There are more than sixty TAC members, including statisticians, physicians, pharmacists, economists, NHS management, patient advocates, and industry representatives. NICE consults with patient organizations, professional bodies, manufacturers, and other experts to develop the scope of each technology review that it conducts. An independent academic center is commissioned to prepare the technology assessment report. Stakeholders, including manufacturers and patient groups, are invited to comment on the assessment documents. TAC then appraises the evidence and publishes a recommendation to be submitted to NICE. Stakeholder organizations may appeal the recommendation. If there is no appeal, and if NICE accepts TAC’s recommendation, guidance is sent to the NHS and it is to be adopted by the PCTs within three months. TAC’s initial assessments and Final Appraisal Determinations are posted on NICE’s Web site; these documents contain cost, clinical, and economic data. NICE’s appraisal process takes fifty-two to sixty-two weeks. NICE completes, on average, eleven appraisals per year for products or classes of products.
To compare the listing decisions and their impact on cost and use, we chose seventeen of the top-selling drugs in 2003 for each of the four countries and the United States (Exhibit 1  ). These products represent a set of extraordinarily high-volume brands, within which there are wide ranges of therapeutic novelty and cost. Several of these drugs, for example, are relatively new entrants into long-established drug classes, while others are intensely marketed "breakthroughs" for which there had (as of 2003) been limited data concerning safety, effectiveness, and costs in the real world. Thus, we expected that some products in this list would not be covered by public drug plans in some of the countries studied. This would stem from differences in the assessment of clinical evidence (for example, some jurisdictions might have taken a more cautious approach to interpreting data on COX-2 inhibitors) or differences in the appraisal of products against national priorities (for example, some jurisdictions might have covered the many antidepressants other than those on this list). The coverage information in Exhibit 1 describes national coverage policy, if any. Data on use and spending are expressed in terms of the relative per capita use and spending (respectively) in the given country compared with the average across all five countries.3
Because central drug review is necessary for determining national coverage decisions in Australia and New Zealand, the respective centralized assessment agencies had reviewed all of the drugs listed in Exhibit 1  . The Australian PBS listed fifteen of these drugs with some form of restriction and did not list two of them. PHARMAC in New Zealand provided unrestricted coverage for two products, listed seven with restrictions, and did not list eight others. By comparison, the experiences in the United Kingdom and Canada reflect the lack of national funding processes. NICE (U.K.) focuses only on individual drugs when there is potential for major controversy and regional variation. NICE therefore reviewed none of the top-selling products in isolation. However, NICE had provided guidance on treating conditions for which five of the products are indicated. The CDR (Canada) was only established in 2003; thus, each of Canada’s public drug plans had separately reviewed the drugs in Exhibit 1. Only four of these were subsequently covered by all of Canada’s provincial drug plans. Comparison of listing decisions. In general, national listing decisions were not perfectly concordant across the four countries. This reflects differences in what is appraised, interpretations of evidence, coverage of therapeutic alternatives, and negotiations with suppliers. The two processes tied to positive, national formularies (Australia and New Zealand) were comprehensive in their review of all products. These processes tended to result in coverage policies involving restricted listings. The other two countries had a lesser amount of active review of the leading products in 2003.
Impact on use.
As shown in Exhibit 1 Impact on spending. Spending levels followed closely usage levels in all countries. It is notable that in cases where Australia and New Zealand had higher-than-average usage levels, they often had lower-than-average spending levels. This likely reflects the fact that these countries tie national coverage decisions to negotiations over the prices of products. This practice is not unlike policies of large U.S. insurers. It is notable that because of relatively high costs per unit, the United States tended to be a greater outlier in per capita spending than in per capita usage. This likely stems from the fact that data presented here do not reflect all discounts provided directly from manufacturers to insurers; U.S. insurers that covered some of these drugs in high volume would have likely secured lower prices than implied by these data.
After reviewing the drug processes of the four countries, we conducted semi-structured telephone interviews to elicit expert opinion regarding the impact and perceived acceptability of the processes. In fall 2004 we identified and interviewed four key informants from each country: one expert from the review agency, an informant from a major drug benefit plan, an industry association representative, and an informed member of academe. Interviews were structured around questions to determine how the central drug review process worked in practice within the given country, what was perceived to be "working" or not, and what key lessons could be drawn from experiences in the country.4 Although official government policy statements tend to steer clear of terminology such as priority setting and (even more so) rationing, there was little doubt in the minds of the experts we interviewed that the intent of these processes is to inform coverage policy. Informants were virtually unanimous in the view that centralization of scientific review processes can foster improved priority setting. Moreover, most viewed the establishment of arms-length processes as generally positive. As one informant put it, in this politically contentious area, policymakers can allow science to "take some of the heat" by having an independent review process to apply "what appear to be established standards of evidence." Related to this last point, the political sustainability of a drug review process is widely believed to lie in the perception (at the very least) that it applies appropriate standards of evidence and that whoever makes decisions in light of that evidence has legitimate authority to do so. Three lessons from the interviews stand out in relation to this: The process must be rigorous, the roles of experts must be clear, and the rationale behind decisions should be transparent. Rigor of process. Setting high standards of evidence and consistently maintaining those standards for every drug review improves the impact and acceptability of a drug review process. The standards and processes should be such that disinterested and well-informed people will find them reasonable.5 Experts in the four countries shared belief in a "hierarchy of evidence," as is commonly used to determine the scientific merit of a study design (for example, ideally blinded, randomized controlled trials). However, they also cautioned against undue focus on trial designs: "Simply saying ‘we’ve got a good randomized trial here’ is usually not the full picture about the strength or the overall influence of the evidence base on the decision making." Informants suggested that standards of evidence for review processes should include a "hierarchy of outcomes" that would, for example, place greatest emphasis on all-cause mortality and least emphasis on subclinical, surrogate outcomes. Finally, experts also suggested that review processes assess the "relevance" of scientific data to the coverage decision at hand: For example, "Is [the finding] in the patients that we wanted to apply the medicine to? Is it for the outcomes that we’re interested in?" Clarity of roles. Drug review processes can improve assessment of evidence by investing in expertise and maintaining, to the extent possible, the independence and objectivity of this expertise. It was widely believed that a drug review process is improved by clearly defining roles and responsibilities for such experts. Just as conflicts of interest with manufacturers discredit the process, so people charged with assessing scientific evidence should not also be responsible for appraising evidence and making final coverage recommendations. Independence of scientific process implies that both manufacturers and funding agencies must "live with" the results of the scientific assessment, even if the science suggests findings that are financially or politically costly. Division of assessment and appraisal roles can help to maintain the necessarily high standards of scientific evidence, leaving appraisal and recommendations to representative committees who weigh competing objectives against appropriately critiqued and summarized scientific data. Such division does not, however, imply that appraisal is a dispensable part of the drug review process. Assessment of evidence, in and of itself, provides only part of what evidence-based coverage policy requires. Appraising evidence, by judiciously considering its meaning and relative importance to policy questions, completes the decision-supporting function of centralized drug review processes. Transparency of rationale. Transparency of information and decision-making rationales was believed to support centralized drug review. Accountability and political defensibility of coverage decisions is increased when stakeholders understand the reasons for decisions. A central review agency can foster understanding by publishing assessments of scientific evidence and the rationales behind comparative appraisals and coverage recommendations. One of the major obstacles to transparency remains commercial confidentiality imposed by manufacturers. Centralized review processes might help to ensure greater accountability of both decisionmakers and manufacturers by requiring full disclosure of all reasonable evidence relevant to a drug’s potential coverage. Should this be enforced uniformly, it would not place any single firm at a competitive disadvantage and would dramatically increase the level of debate concerning acceptability of coverage decisions made on the basis of critically appraised scientific evidence.
The four centralized processes for reviewing drugs to inform coverage decisions that were studied here evolved out of different policy circumstances and have slightly different objectives. Australia and New Zealand. From its inception in 1953, the Australian PBS required a process for determining what should be provided as a national benefit. When New Zealand’s PHARMAC effectively "nationalized" drug benefits in 1993, it too required a process for formulating coverage policy. The relationships between the central drug review process and drug coverage decisions are therefore very close in Australia and New Zealand. This increases the impact of recommendations made and creates the opportunity to tie price considerations to listing on the national formulary. U.K. and Canada. The relatively new systems of drug review in the United Kingdom and Canada arose out of quite different circumstances and have less direct impact on coverage policy. NICE emerged in 1999 out of concern that NHS funding mechanisms created undesirable variations in access to certain technologies. National standards can play an important role in reducing undesired regional variations; however, reviewing and making recommendations in only these cases does not assist with the majority of coverage decisions made by regional authorities. It remains the case that unless otherwise declared at the national level, all products are ostensibly eligible for public coverage in the United Kingdom. Regional authorities are therefore likely to engage in a greater degree of implicit rationing at a regional or clinical level, or both, than would be the case in comparator countries. The Canadian CDR arose in 2003 out of a desire to coordinate drug review processes for many public drug plans. However, the recommendations of the CDR are effectively just that: recommendations. Each participating drug plan retains the right to make final coverage decisions based on local priorities and resource constraints. This harmonizes and probably increases the standard of evidence used in coverage processes, while leaving final decisions to regional discretion. Combined with the fact that government pays for a minority share of expenditures, Canada’s pluralist approach to drug coverage reduces the impact of recommendations from its centralized review process. United States. The United States is evolving a number of health technology assessment initiatives for the purpose of informing coverage decisions. One such example is the Drug Effectiveness Review Project (DERP), a "centralized" process for conducting regular systematic reviews of evidence concerning the comparative performance of drugs within leading therapeutic classes of medicine.6 DERP is funded by a collaboration of private and public organizations that choose which drug classes are to be studied. It conducts the "assessment" stage of the drug review process by reporting what is likely to be the effectiveness of alternative products based on systematic review; it does not conduct an appraisal of the evidence in the sense that DERP does not recommend what drugs ought to be covered in light of the systematically reviewed evidence. Potential for improvement. The experiences of Australia, Canada, New Zealand, and the United Kingdom indicate that the effectiveness of processes like DERP depends on the rigor and transparency of the process. DERP upholds a high level of rigor and independence and publishes the results of its systematic reviews online. However, to have maximum impact on coverage and subsequently utilization, the drug review process must be consistently applied to all medicines and be clearly tied to the decision-making process that involves appraisal of the evidence and coverage determinations. Although DERP assists a wide range of decisionmakers evaluate competing drugs within the major therapeutic classes, an even greater impact might be had through expanding the process to review all new products (in addition to reviewing entire classes of products), and by incorporating the results of such reviews directly into an appraisal process and the coverage policy of major drug plans. THE TRANSPARENCY, RIGOR, AND ROBUSTNESS of centralized review processes are critically important for them to be effective decision-supporting policy tools. If drug review processes are designed with these attributes in mind, they can help policymakers make "tough" but evidence-based choices in this health care sector. In the words of an expert informant, what a review agency ultimately wants the public to say is, "I understand how they reached that decision. I can see the process. I followed the methods. I may not necessarily have made exactly the same judgment, but they’ve made a judgment, which in the circumstances can be justified."
Steve Morgan (morgan{at}chspr.ubc.ca) is a health economist at the University of British Columbia (UBC); an assistant professor at the Centre for Health Services and Policy Research; and research lead for the Program in Pharmaceutical Policy. Meghan McMahon was an intern at the center during this project and is completing her master of science degree in health policy at the University of Toronto. Craig Mitton is an assistant professor on the Faculty of Health and Social Development at the UBC Okanagan campus. Elizabeth Roughead is an associate professor at the Sansom Institute in Adelaide, Australia. Ray Kirk is deputy director and senior lecturer at the Health Sciences Centre, University of Canterbury, in Christchurch, New Zealand. Panos Kanavos is a lecturer in international health policy and a research fellow in pharmaceutical economics at the London School of Economics. Devidas Menon is a professor of public health sciences at the University of Alberta in Canada. This research was supported by the Commonwealth Fund, a New York City–based private independent foundation. Steve Morgan is supported in part by the Canadian Institutes of Health Research and the Michael Smith Foundation for Health Research. Craig Mitton is supported in part by the Canadian Priority Setting Research Network. The authors are indebted to expert informants in Australia, Canada, New Zealand, and the United Kingdom who participated in interviews and provided valuable insights. The views presented here are solely those of the authors and not necessarily those of the Commonwealth Fund or the agencies studied herein.
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Structure Of The Processes
Overview Of The Processes
