Scientific versus commercial sources of influence on the prescribing behavior of physicians

Abstract

W hen a physician prescribes a medication for a patient, the act is often shaped, in a large part, by forces unrelated to the biochemical properties of the drug-a phenomenon which has been called “the non-pharmacological basis of therapeutics” [ 11. A number of papers have appeared which analyze how physicians arrive at therapeutic decisions, and generally, the scientific literature is accorded a prominent position in the list of influential sources. These studies have been the subject of several comprehensive reviews [2-51. However, such research has often relied heavily on self-report as a major source of data, introducing a strong potential bias. In contrast, surveys of actual prescribing practices indicate that irrational drug choices are made frequently, despite the availability of ample empirical evidence counseling otherwise. To study the relative contributions of scientific and commercial sources of information, we chose two index drugs whose pharmacologic effects have been shown by controlled studies to be minimal or not significantly different from those of nonprescription (over-thecounter) preparations, but which are heavily advertised as being effective. We then interviewed a random sample of primary care physicians in the greater Boston area, and sought to determine their use of the drugs, their perception of the drugs’ properties, and their beliefs concerning what influenced their choice of drugs. The nature of drug advertising is such that physicians often deny the relative importance of commercial sources in influencing their prescribing-either because they are unaware of it or because they are reluctant to admit to being influenced by nonscientific sources. This limits the value of self-report as a means of. determining how physicians make prescribing decisions. We attempted to overcome this problem by choosing drugs about which the content of information provided varied systematically from source to source. For both drug groups, commercial “channels” (advertisements, detail personnel) presented a message of efficacy and reliability, whereas scientific channels (published reports of clinical trials or review articles) presented a message of minimal efficacy or total uselessness. The content of physicians’ beliefs about these drugs could thus be used as a “marker,” indicating from which sources the information in fact came. For example, a physician may respond that he bases his prescribing on the scientific literature alone and never reads pharmaceutical advertisements or pays attention to drug salespeople. However, his answers in another part of the questionnaire may reveal that he believes that cerebral vasodilators are effective in the treatment of senile dementia. Such information is unlikely to have come from the scientific literature since it does not present this point of view.