Drug Advertising that Doesn't Work--Is It Possible?

No less a guru than Dr. Peter Mansfield, the main guy at Australia's Healthy Skepticism website, says so.

Two articles in the Medical Journal of Australia (http://www.mja.com.au/, free access, but you have to register) tell the story. It seems that a few years back, when Australian GPs were (unlike most of their U.S. counterparts) eagerly embracing the electronic health record with computerized prescription writing and other features, they were offered two software packages-- a more expensive one that was advertising-free, and a less expensive system that was supported in part by advertising, primarily pharmaceutical. The ads run in two sizes and are typically changed each month. They are programmed so as to pop up when the physician performs certain tasks--for example, when a doc clicks on an area of the medical chart that records cardiac risk factors, a Lipitor ad might then be triggered.

Joan Henderson and colleagues from Sydney took advantage of an ongoing survey of GP practice patterns to compare the prescribing behavior of GPs who used the ad-containing software with those using the ad-free software. They looked at a number of variables and data adjustments but were unable to find any difference.

Mansfield, in his commentary on their research, agreed that they had rigorously controlled for as many variables as possible, so that their findings were probably reliable. If so, then the drug companies seem to have broken new ground--finding a method of spending their advertising dollars that does not actually change physicians' behavior. Mansfield speculates that the annoyance factor may have actually have swamped out the usual value of ads. He offers the possibility that the ads may prove more effective in future years when the docs have become more habituated to them and so the annoyance variable fades.

I am no advertising expert (thank heaven) but would offer a slightly different hypothesis, or maybe the same hypothesis worded differently. How does pharmaceutical advertising work? Many docs (like me for instance) are quite sure that they completely skip over the drug ads in medical journals, for example, yet most company studies show that the return-on-investment for journal ad dollars is pretty good. I think the secret of a successful ad campaign is that there is no one element working in isolation. The physician hears a consistent message simultaneously from several sources (drug reps, journal ads, patients coming in reporting TV ads in the U.S., and now these software ads Down Under), each reinforcing the other. It's human nature to imagine that what you hear from one place might be false, but what you hear over and over from numerous sources must be true.

Henderson et al. assume quite reasonably that the companies must have gathered data of their own on the apparent effectiveness of their advertising expenditures; and she invited them to make these data public so that her group's results could be compared. Companies making public their proprietary marketing data? As the Russians used to like to remark, when pigs fly.

Henderson J, Miller G, Pan Y, Britt H. The effect of advertising in clinical software on general practitioners' prescribing behaviour. Med J Aust 188:15-20, 7 January 2008.

Mansfield PR. Do advertisements in clinical software influence prescribing? Med J Aust 188:13-14, 7 January 2008.