Support DTCA Because It Leads to Bad Medicine?

I'm grateful to my pals Rick Bukata and Jerry Hoffman, of Primary Care Medical Abstracts, for calling my attention to this study, which I missed when it first came out.

Adam E. Block of Harvard set out to do a cost-benefit analysis of direct-to-consumer advertising (DTCA) of prescription drugs and chose depression as his test case. He concluded that DTCA is a wonderful thing. (He also explicitly denied having any conflicts of interest or receiving any industry funding.)

Here's the basic scoop--Block used a variety of surveys and studies to gather his data, and relied heavily on the Kravtiz et al. study that sent actors into phyysician's offices with different stories about DTCA, to see which ones got prescriptions. He figured out that DTCA would lead to massive overuse of antidepressants, with only about 1 out of 15 of those getting prescriptions actually being depressed. However, he also calculated that each depressed patient who takes a drug and gets benefit from it achieves such a gain in quality of life, that the improvements in that small group outweigh by far the costs of the drugs for the large group that does not need them.

Most understand that the problem with this sort of cost-benefit analysis is in the assumptions, and so Block quite responsibly subjects his findings to a sensitivity analysis. He claims that even fairly substantial shifts in his numbers would not change the final outcome; i.e. his results are reasonably robust.

Despite the sensitivity analysis, it is easy to spot serious holes in Block's model. For example, he assumes that antidepressants would be prescribed only for 60 days and so he calculates the cost of the drug on that basis. It is much more likely that an antidepressant, once started, would be prescribed for a year, if not indeed forever. Block also presumes that antidepressants have no side effects and so no patient's quality of life ever deteriorates as a result of taking one. I was not able to plug new numbers into his equation to see how changing these two assumptions would alter his results. As a final concern, he appears to assume that all patients who now have depression and yet are untreated have serious depression and so stand to gain substantially from therapy.

But for now, let's assume that Block's methods were impeccable and that his results hold water. On what basis would we recommend a policy of DTCA? It would seem, at least in the depression case, that we advocate DTCA because it leads to 15 patients with no depression being treated with a drug that they don't need, for every one patient with depression that gets the drug that truly is indicated. In short, DTCA should be recommended as good public policy because it results in abysmal medical practice by just about anyone's standard. (Block, to his credit, seems to have some sense of this and discusses the implications a bit at the end of his article; but overall he softpedals this major message.)

The obvious public policy question is how one could invest the resources now devoted to DTCA in such a manner that, instead of producing abysmal medical practice, one actually improved medical practice with regard to depression. It is hard to imagine that there is not a better approach than DTCA.

Block AE. Costs and benefits of direct-to-consumer advertising: the case of depression. Pharmacoeconomics 25:511-21, June 2007.

Kravitz R, Epstein R, Feldman M, et al. Influence of patients' requests for direct-to-consumer advertised antidepressants. JAMA 293:1995-2002, 2005.