- © 2004 Canadian Medical Association or its licensors
DTCA is illegal in Canada, which serves as a measure to protect those who are ill from undue marketing influences and from the harm that might result from medically unjustified use of medications. We trust that John Graham is not suggesting that the burden of proof be on health authorities to provide ironclad evidence of harm in order to maintain such safeguards.
Graham's claim that DTCA has net benefits if it elicits no greater ambivalence than requests for nonadvertised drugs assumes that the latter are beneficial. Antibiotics, anxiolytics–hypnotics, stimulants and narcotic analgesics were among the nonadvertised drugs requested in our study.1 Advertising is not the only factor associated with pressure to prescribe, but if it adds to existing pressures, the net effect would be greater harm.
Graham quotes an FDA survey of US physicians, only 18% of whom felt that DTCA had created problems with a patient encounter.2 However, 47% reported some pressure to prescribe, and 17% reported that the pressure was moderate to strong. In our study, physicians reported pressure to prescribe in only 48 (3%) of 1431 consultations overall, but in 15 (20%) of 74 consultations involving DTCA drug requests.
Marc Lacroix questions physicians' ambivalence about prescriptions for requested DTCA drugs, focusing on treatment choices that would have been “unlikely” for other similar patients. Our decision to separate “possible” and “unlikely” choices from “very likely” choices was made a priori, to distinguish prescriptions with any degree of ambivalence from those without ambivalence in a manner that was sensitive to physicians' legal and moral responsibility for prescribing decisions. However, if we instead compare “unlikely” with “possible” and “very likely” choices, as Lacroix suggests, neither the direction nor the magnitude of effect changes. Physicians were still significantly more likely (5/60 or 8.3%) to judge requested DTCA drugs to be “unlikely” choices for other similar patients than nonrequested drugs (7/500 or 1.4%) (unadjusted odds ratio [OR] 6.8, 95% confidence interval [CI] 1.8– 26.2; adjusted OR 9.4, 95% CI 2.8– 32.0; unit of analysis is each newly initiated prescription; analysis performed by a general estimation equation with adjustment for age, sex, health status, income, education, drug payment, and physician's sex and graduation year).
Finally, we could find no evidence to support Lacroix's suggestion that DTCA leads to savings in health care costs.
Barbara Mintzes Kenneth L. Bassett Morris L. Barer Centre for Health Services and Policy Research University of British Columbia Vancouver, BC