Are cancer drugs less likely to be recommended for listing by the Pharmaceutical Benefits Advisory Committee in Australia?

BACKGROUND

The hurdle of cost effectiveness for the selection and reimbursement of drugs in Australia limits access to new medicines based on an assessment of cost relative to clinical benefit. Those drugs that are expensive and provide modest benefits will be less likely to receive a government price subsidy. There is concern that the cost-effectiveness hurdle will limit access to new cancer treatments because of their high costs and modest benefits.

OBJECTIVE

To test the hypothesis that Ceteris paribus, cancer drugs are less likely to receive a recommendation for reimbursement on the Pharmaceutical Benefits Scheme (PBS) than non-cancer drugs.

METHODS

We reviewed public summary documents (PSDs) on all major submissions considered by the Pharmaceutical Benefits Advisory Committee (PBAC) from July 2005 to March 2008. Each PSD includes summary information on the clinical, economic and utilization considerations of the PBAC in arriving at a recommendation. A total of 227 PSDs were reviewed, from which 243 PBAC recommendations were identified. Logistic regression was used to determine the effects of drug type (cancer vs non-cancer) and other potentially confounding variables on the outcome of PBS approval versus non-approval.

RESULTS

There were 243 PBAC recommendations in 227 published PSDs: 108 for rejection (44%), 10 deferrals (4%) and 125 (51%) recommendations for listing. Recommendations for listing were made somewhat more often for non-cancer drugs than for cancer drugs: 104/191 (54%) versus 21/52 (40%), respectively; p = 0.07. Based on the results for univariable analyses, there is evidence that four variables have some association (p < 0.25) with PBAC approval, but only type of application, economic modelling and estimated cost to the PBS remained statistically significant at p < 0.05 in the multivariable model. No interaction terms were statistically significant. Cancer drug submissions tend to have a modelled economic evaluation and have a higher cost per QALY than non-cancer drugs (29% vs 15% of cancer and non-cancer drugs, respectively had a reported modelled cost per QALY of more than Australian dollars [$A]45 000; p < 0.001). Submissions that include a modelled economic evaluation and have a higher cost per QALY get approved less often than submissions without an economic modelling (p = 0.04). However, after adjusting for economic modelling, there is no statistical difference between cancer and non-cancer drugs in terms of gaining recommendation for PBS listing.

CONCLUSION

The PBAC applies decision criteria equitably to cancer and non-cancer drugs, in that cancer drugs are neither favoured nor disadvantaged but they are more expensive and target a smaller population than non-cancer drugs. Further debate and research is needed to determine society's willingness to pay for a QALY and whether this differs between drugs for cancer and other indications or for interventions that differ on criteria other than cost effectiveness.

The price function of toxicity

The high and accelerating price of new anticancer drugs is giving rise to increased concern. However, monetary price is not the only way to value chemotherapy. Toxic effects can also be seen as a form of payment in which "units" of wellbeing are exchanged for "units" of efficacy. Although this trading analogy is not perfect, a proposal can be made that toxicity is a type of price, and that one of its functions is to signal valuation, similar to the crucial signalling function of monetary price in the real economy. This price function of toxicity, to the extent where there is transparency about the real amounts of toxicity, can have two important and helpful consequences: acting as a brake on the increasing monetary price of new drugs, via a damping effect on demand; and assisting individual patients in the informed contemplation of chemotherapy decisions. However, there are two problems that currently impede the effective dissemination of this highly desirable toxicity information. First, a prediction of toxicity in individual patients is difficult. Second, the vast database of real toxic effects in community practice is rarely made available for public scrutiny. Both of these problems, which together constitute a form of hidden cost, are potentially resolvable at least to some extent. In the absence of accurate information on toxic effects, it is easy for monetary price to progressively diverge from true value. We believe that improved transparency with respect to toxic effects, and better toxicity prediction, offer a better and more genuinely market-orientated solution to the issue of price distortions than the bureaucratic imposition of price controls.