Seventy five percent of the reviewed articles were considered valid. This supports findings from a previous study that assessed whether the quality of a study was influenced by sponsorship . The authors found that the majority of the studies sponsored by pharmaceutical companies were of good quality. However, they also found that published studies that were sponsored by pharmaceutical companies were four times more likely to have outcomes favoring the sponsor's product than were studies that had other types of sponsors. One potential solution to reducing the skepticism resulting from industry-funded research is to have pharmaceutical companies improve access to their original data. Another possible solution is for academia to have funding available to independently conduct studies without the influence of the pharmaceutical companies. Although pharmaceutical companies exert control over their clinical trials, clinical outcome studies conducted at academic health science centers may have a role in complementing the post-marketing surveillance of new drugs following FDA approval.
A large majority of the brochures presented data that were based on underlying studies funded by the pharmaceutical company. However, a recent study by Cooper and Schriger found that only 58% of the original research cited in advertisements was sponsored by a pharmaceutical company or had a company affiliated author . This deviation from our results may be due to differences in methods for ascertaining funding source and product mix analyzed (e.g., brochures vs journal advertisements). Alternatively, it may also reflect our study's small sample size. There has been concern that industry-supported manuscripts may selectively report study data that make a medication or therapy appear more efficacious than it actually is [39–41]. The clinician's dilemma becomes, "Are the results from this 'valid study' truly valid?"
Although a majority of the outcomes in the reviewed studies were patient-oriented, a substantial number were not. It is important to present data that reflects the outcome of the patient, such as mortality, quality of life changes, and symptom reduction. Thus, while a CCB may reduce blood pressure just as effectively as a diuretic, patient outcomes (e.g., mortality) may prevail in the long term [11, 42].
While 95% of the brochures presented a visual graphic, only 1 brochure presented an absolute risk reduction and 4 brochures presented relative risk reductions. This "framing effect" may mislead physicians to conclude that a large difference in outcome occurs with the use of the promoted medication or therapy . Moreover, as the consistency of outcomes was being assessed in the underlying study and its subsequent promotional brochure, the reviewers noted that the overwhelming majority of data selected for the brochure were based on a desired visual impact. Although it was not a part of this study's objectives, it appeared that pharmaceutical companies often selected the outcome with the greatest relative risk reduction or percent-change from baseline in lieu of more clinically important findings in the original manuscript. This is an area for further research and scrutiny.
A minority of the studies (eight of 20) compared the medication to another medication or treatment strategy. When a medication has already been approved by the Food and Drug Administration, subsequent studies that compare that medication to placebo have limited value to practicing physicians, as efficacy has already been established. It is of more value to physicians for studies to compare new medications to generic and inexpensive medications that are currently used in practice to determine if a change in disease management is indicated.
Although three of the 20 brochures presented data that were different from the original study, the differences were small and most likely would not have affected the clinician's prescribing behavior.
This study had several limitations. Only two blinded reviewers were available and any discrepancies were resolved by consensus rather than adjudication by a third reviewer. The method in which the reviewers deemed the underlying studies valid has not been validated in previous studies. The sample size may have been too small to detect clinically important findings. Because brochures were collected only in selected family medicine clinics in one city, these findings may not be generalized to other specialties, settings (i.e. hospital), or geographic regions. Only brochures were studied, therefore these findings may not be generalized to other types of promotional products or advertisements, such as journal advertisements. We believe that successful drug marketing to physicians entails more than simply presenting a brochure to the physician and this study only assessed one aspect of drug marketing tactics. There may be seasonal variations in drug marketing that further limit generalizability, although this may have been mitigated to some degree by the wide range of drugs included in the study.