More on the Effects of Samples in the Doctor's Office

It's nice to be able to come back to a topic we have not addressed in a while, the good ol' down home matter of the doctor's office sample cupboard. Dr. Richard G. Pinckney and colleagues from the University of Vermont (subscription required to access article) noted that previous attitude surveys about how samples impacted on prescribing practices had usually been restricted to just a few office sites and so set out to do a broader survey, taking advantage of a Vermont statewide primary care database. Now you could immediately object that if you had to survey docs in any state of the US, Vermont might well be the greatest outlier, especially since they are now considering something akin to a single-payer system. Be that as it may, in Vermont they had their database and in Vermont they did their study, which got in the end a 35% response rate, fairly typical for today's world.

Along they way they cited a Kaiser Family Foundation poll that showed that 92% of US physicians reported that at least once in their careers, they'd received samples from a drug rep. I interpret that to mean that 8% of physicians are either liars or have rotten memories.

Pinckney and friends found in this particular instance that about 3/4 of their responding docs had samples in their offices, and they proceeded to compare that group with the remainder who did not. They provided two vignettes involving patients with no insurance and with hypertension and depression, respectively. The majority of docs got the right answer--use a cheap and effective thiazide diuretic for the first-pass treatment of the hypertension, and if the depressed patient needs drug therapy, pick a cheap generic. But there were significant differences between their two groups. Only 70% of the sample-docs picked the thiazide while 91% of the no-sample-docs did, and while 91% of the sample-docs went for the generic antidepressant, every single one of the no-sample-docs did.

They also asked those folks about attitudes toward samples, and again not surprisingly, the sample-docs were significantly more likely to believe that samples expedite treatment, make patients happy, help the indigent, and allow the docs to check out which meds work best. The no-sample-docs were more likely to believe that samples distort treatment plans, increase the costs of care, and lead to overuse. Pinckney et al. noted that actually a majority in both groups agreed that samples increase the costs of care and can affect physicians' prescribing habits.

What's cart and what's horse? It could be of course that docs who have certain attitudes are more likely to elect to dispense with the traditional sample cupboard, or that docs who have samples around all the time end up getting certain attitudes, or a bit of each. From such a study one can only hypothesize associates and not causes. Still this is just a bit more support for the already-fairly-well-documented idea that "free" samples in the office do more harm than good.

(Hat tip to Primary Care Medical Abstracts for the citation.)

Pinckney RG, Helminski AS, Kennedy AG, et al. The effect of medication samples on self-reported prescribing practices: a statewide, cross-sectional survey. Journal of General Internal Medicine 26:40-44, January 2011.