As both public and private payers seek to control pharmaceutical expenditures, formal cost-effectiveness analysis (CEA) has become a common approach for assessing the value of drugs, vaccines and other health technologies.  CEA usually is applied within a broader framework of health technology assessment (HTA), whereby the technology is assessed for use and/or reimbursement price considering incremental health-related effects and costs compared to existing treatments.  Australia was the first to adopt such a policy, in 1993, followed quickly by New Zealand and several provinces in Canada.  Since then, the UK has established NICE (1999) and several other European countries  request economic analyses for at least some new medicines, including Belgium, Finland, Ireland, Norway, The Netherlands, Portugal and Germany.  Similar policies have been adopted recently in some Eastern European countries, Asia and Latin America.  CEA also is used in the United States by some Medicaid programs, private insurers and health care providers, although its expansion under the 2010 health care law remains doubtful.

Adrian Towse has collaborated with outside experts in preparing a chapter that  addresses key  issues in the  forthcoming Oxford Handbook on the Economics of the Biopharmaceutical Industry. The authors outline the evolution of CEA through its use as part of HTA. They also explore the theoretical and practical issues that have arisen as the result of using CEA of drugs to make decisions about resource allocation, pricing and use.  The authors’ conclusions, in part, are that:

• CEA has developed as part of broader HTA.  The routine use of CEA rather than cost-benefit analysis makes health care different from other areas of public policy.

• the use of CEA for pharmaceuticals ‘if done well,’ is the most efficient form of regulation, both theoretically and in practice, but doing CEA well is the challenge.

• incorporating a preference-base outcome measure in CEA is important.  QALYs need to be made more robust and alternatives, such as willingness to pay estimates and stated-preference discrete choice experiments, need to be explored.

• the key parameters in CEA are the assessments of the clinical effectiveness of the health care treatment and program being evaluated, and the value of the resources they consume.  Clinical trials data usually need to be supplemented with other data and synthesized in a decision analytic model.  This remains a methodological and practical challenge.

• harmonization of economic evaluation methods and requirements across jurisdictions face a major issue, i.e., whether estimates are transferable from one location to another.  Often this is not the case, requiring adjustments or collection of local data. Payers and manufacturers are only beginning to understand the implications of this issue.

• in deciding the allocation of health care resources, attributes in addition to health gain may be considered – for example, the broader impact of improved health on the economy and fairness, or equity, in health care provision. A transparent approach for weighting the various attributes is essential and more structured approaches to decision making are needed, including the use of multiple-criteria decision analysis.

• in some cases, it may be helpful to determine a “cost-effectiveness threshold” of willingness to pay, reflecting the opportunity cost of the resources consumed by treatments and programs at the margin.  Determining the level of such a threshold is not a simple task and may make sense only in health care systems with a “hard” budget constraint.

• no matter how well CEAs are conducted, uncertainty always will exist about the estimates.  The use of “value of information” analysis can help the decision maker determine whether investment in further research might be worthwhile. One approach is to offer “coverage with evidence development” (CED) for new health technologies, granting the new medicine or device market access on condition that further research into its costs and effectiveness is completed.

• As with all uses of scarce resources, CEA itself must deliver value for money to decision makers. CEAs need to meet decision makers’ needs without being overly complex, if this adds to cost but not to value.  Mechanisms also must exist to implement the findings of CEA, otherwise investment in CEA cannot deliver a return.

Towse, A., Drummond, M. and Sorenson, C. Measuring value: Pharmacoeconomics in theory and practice.  In Danzon, P. and Nicholson, S.  Oxford handbook on the economics of the biopharmaceutical industry. Oxford: Oxford University Press.  (forthcoming).