Anthony Harris1, Jing Jing Li2, Karen Yong2. 1. Centre for Health Economics, 15 Innovation Walk Monash University, Clayton, VIC, 3800, Australia. anthony.harris@buseco.monash.edu.au. 2. Centre for Health Economics, 15 Innovation Walk Monash University, Clayton, VIC, 3800, Australia.
Abstract
OBJECTIVE: Deciding on public funding for pharmaceuticals on the basis of value for money is now widespread. We suggest that evidence-based assessment of value has restricted the availability of medicines in Australia in a way that reflects the relative bargaining power of government and the pharmaceutical industry. We propose a simple informal game-theoretic model of bargaining between the funding agency and industry and test its predictions using a logistic multiple regression model of past funding decisions made by the Pharmaceutical Benefits Advisory Committee in Australia. METHOD: The model estimates the probability of a drug being recommended for subsidy as a function of incremental cost per quality-adjusted life-year (QALY), as well as other drug and market characteristics. Data are major submissions or resubmissions from 1993 to 2009 where there was a claim of superiority and evidence of a difference in quality of life. Independent variables measure the incremental cost per QALY, the cost to the public budget, the strength and quality of the clinical and economic evidence, need as measured by severity of illness and the availability of alternative treatments, whether or not a resubmission, and newspaper reports as a measure of public pressure. We report the odds ratio for each variable and calculate the ratio of the marginal effect of each variable to the marginal effect of the cost per QALY as a measure of the revealed willingness to pay for each of the variables that influence the decision. RESULTS: The results are consistent with a bargaining model where a 10,000 Australian dollar ($A) fall in value (increase in cost per QALY) reduces the average probability of public funding from 37 to 33% (95% CI 34-32). If the condition is life threatening or the drug has no active comparator, then the odds of a positive recommendation are 3.18 (95% CI 1.00-10.11) and 2.14 (95% CI 0.95-4.83) greater, equivalent to a $A33,000 and a $A21,000 increase in value (fall in cost per QALY). If both conditions are met, the odds are increased by 4.41 (95% CI 1.28-15.24) times, equivalent to an increase in value of $A46,000. Funding is more likely as time elapses and price falls, but we did not find clear evidence that public or corporate pressure influences decisions. CONCLUSION: Evidence from Australia suggests that the determinants of public funding and pricing decisions for medicines reflect the relative bargaining power of government and drug companies. Value for money depends on the quality of evidence, timing, patient need, perceived benefit and opportunity cost; these factors reflect the potential gains from striking a bargain and the risk of loss from not doing so.
OBJECTIVE: Deciding on public funding for pharmaceuticals on the basis of value for money is now widespread. We suggest that evidence-based assessment of value has restricted the availability of medicines in Australia in a way that reflects the relative bargaining power of government and the pharmaceutical industry. We propose a simple informal game-theoretic model of bargaining between the funding agency and industry and test its predictions using a logistic multiple regression model of past funding decisions made by the Pharmaceutical Benefits Advisory Committee in Australia. METHOD: The model estimates the probability of a drug being recommended for subsidy as a function of incremental cost per quality-adjusted life-year (QALY), as well as other drug and market characteristics. Data are major submissions or resubmissions from 1993 to 2009 where there was a claim of superiority and evidence of a difference in quality of life. Independent variables measure the incremental cost per QALY, the cost to the public budget, the strength and quality of the clinical and economic evidence, need as measured by severity of illness and the availability of alternative treatments, whether or not a resubmission, and newspaper reports as a measure of public pressure. We report the odds ratio for each variable and calculate the ratio of the marginal effect of each variable to the marginal effect of the cost per QALY as a measure of the revealed willingness to pay for each of the variables that influence the decision. RESULTS: The results are consistent with a bargaining model where a 10,000 Australian dollar ($A) fall in value (increase in cost per QALY) reduces the average probability of public funding from 37 to 33% (95% CI 34-32). If the condition is life threatening or the drug has no active comparator, then the odds of a positive recommendation are 3.18 (95% CI 1.00-10.11) and 2.14 (95% CI 0.95-4.83) greater, equivalent to a $A33,000 and a $A21,000 increase in value (fall in cost per QALY). If both conditions are met, the odds are increased by 4.41 (95% CI 1.28-15.24) times, equivalent to an increase in value of $A46,000. Funding is more likely as time elapses and price falls, but we did not find clear evidence that public or corporate pressure influences decisions. CONCLUSION: Evidence from Australia suggests that the determinants of public funding and pricing decisions for medicines reflect the relative bargaining power of government and drug companies. Value for money depends on the quality of evidence, timing, patient need, perceived benefit and opportunity cost; these factors reflect the potential gains from striking a bargain and the risk of loss from not doing so.