PURPOSE: To document the quality of evidence base for marketing claims in prescription drug advertisements, to facilitate identification of potential targets for quality improvement. METHODS: A sample of 1036 advertisements from four major Finnish medical journals published in 2002. Marketing claims were classified in four groups: unambiguous clinical outcome, vague clinical outcome, emotive or immeasurable outcome and non-clinical outcome. Medline references were traced and classified according to the level of evidence available. The statistical variables used in the advertisements were also documented. RESULTS: The sample included 245 distinct advertisements with 883 marketing claims, 1-10 claims per advertisement. Three hundred thirty seven (38%) of the claims were referenced. Each claim could be supported by one reference or more, so the number of references analysed totalled 381, 1-9 references per advertisement. Nine percent of the claims implied unambiguous clinical outcomes, 68% included vague or emotive statements. Twenty one percent of the references were irrelevant to the claim. There was a fair amount of non-scientific and scientific support for the 73 unambiguous claims, but not a single claim was supported by strong scientific evidence. Vague, emotive and non-clinical claims were significantly more often supported by non-Medline or irrelevant references than unambiguous claims. Statistical parameters were stated only 34 times. CONCLUSION: Referenced marketing claims may appear more scientific, but the use of references does not guarantee the quality of the claims. For the benefit of all stakeholders, both the regulatory control and industry's self-control of drug marketing should adopt more active monitoring roles, and apply sanctions when appropriate. Concerted efforts by several stakeholders might be more effective. Copyright 2004 John Wiley & Sons, Ltd.
PURPOSE: To document the quality of evidence base for marketing claims in prescription drug advertisements, to facilitate identification of potential targets for quality improvement. METHODS: A sample of 1036 advertisements from four major Finnish medical journals published in 2002. Marketing claims were classified in four groups: unambiguous clinical outcome, vague clinical outcome, emotive or immeasurable outcome and non-clinical outcome. Medline references were traced and classified according to the level of evidence available. The statistical variables used in the advertisements were also documented. RESULTS: The sample included 245 distinct advertisements with 883 marketing claims, 1-10 claims per advertisement. Three hundred thirty seven (38%) of the claims were referenced. Each claim could be supported by one reference or more, so the number of references analysed totalled 381, 1-9 references per advertisement. Nine percent of the claims implied unambiguous clinical outcomes, 68% included vague or emotive statements. Twenty one percent of the references were irrelevant to the claim. There was a fair amount of non-scientific and scientific support for the 73 unambiguous claims, but not a single claim was supported by strong scientific evidence. Vague, emotive and non-clinical claims were significantly more often supported by non-Medline or irrelevant references than unambiguous claims. Statistical parameters were stated only 34 times. CONCLUSION: Referenced marketing claims may appear more scientific, but the use of references does not guarantee the quality of the claims. For the benefit of all stakeholders, both the regulatory control and industry's self-control of drug marketing should adopt more active monitoring roles, and apply sanctions when appropriate. Concerted efforts by several stakeholders might be more effective. Copyright 2004 John Wiley & Sons, Ltd.
Authors: Jennifer Gettings; Braden O'Neill; Dave A Chokshi; James A Colbert; Peter Gill; Gerald Lebovic; Joel Lexchin; Navindra Persaud Journal: PLoS One Date: 2014-01-08 Impact factor: 3.240