Sunday, November 28, 2010

Cost-effectiveness and decision making

NPR (and I'm sure other news sources) recently had a story about a drug called Truvada that has been shown to reduce the risk of contracting HIV.  This would be incredible news if the drug were cheap.  The simple fact, however, is that right now the drug costs $13,000 per year in the United States while the drug may cost less (because it is already available in generic forms) overseas.

One researcher says that given the high lifetime cost of having HIV (given the drug that are available to control it) Truvada (clearly an expensive drug) may still be as cost-effective as things that are done to prevent heart disease and cancer.

One person at the NIH has a comment attributed to him suggesting that cost-effectiveness analysis should guide private and government insurance coverage.  This is interesting as cost-effectiveness is not generally a criterion used in many government programs for the decision of whether to cover.  It is also interesting because a lot of cost-effectiveness studies focus on quality adjusted life years and there are many interpretations of whether the use of quality adjusted life years was explicitly excluded in the health care reform bill.  The case of Truvada could be thought to illustrate why government should have this tool at its disposal--particularly if there is a coverage policy that can be used to help with targeting the drug.

It is also critical to note the importance of sensitivity analyses. The article pointed out the difference in effectiveness of Truvada for those with very high adherence to the recommended regimen in contrast with those who only took the drug some of the time.  Another important sensitivity analysis would be the cost-effectiveness when the patent expires and generics are available even in the US.  Is it not clear what the policy should be when it is possible that there will be long-term cost-effectiveness even if the drug is not cost-effectiveness at the higher price level when the patent still applies.

And, while I may see this as evidence in favor of cost-effectiveness tools, others may see it as exactly the reason that either (a) we should totally leave even the ability to get the drug to the market (i.e. if you can afford it that is fine but it is too expensive for governments to cover) or (b) the drug should be covered but it should be the responsibility of the patient to be as forthcoming as possible about the many factors that influence adherence, the risk of contracting HIV, and feelings about side effect, and the responsibility of the provider to assess these and to determine whether Truvada or some other preventive method is best for each individual patient.  In fact, even with coverage, a provider should be making relatively similar decisions.

The key question for any coverage of Truvada is whether any decision maker can be given a strong incentive to consider the monetary costs in comparison with the benefits rather than just comparing the potential for clinical harms versus clinical improvements.