Evaluating the accuracy of citations in drug promotional brochures

To evaluate the accuracy of statements cited in 3 x 5 inch promotional cards for the 50 most frequently prescribed drugs.

The 50 most frequently prescribed drugs were identified in the April 1992 issue of Pharmacy Times. File cards were requested from the pharmaceutical companies. References in the file card were retrieved and cited statements identified. Criteria used to evaluate reference accuracy were: type of study design, use of peer review journals, retrievability of references in area libraries, and reference documentation as no error, major error, or minor error. Referenced statements were classified as correct, incorrect, misleading, or taken from the abstract, discussion, or conclusion section of a study.

Of the 50 most frequently prescribed products, 21 file cards were obtained. One hundred forty-two cited references were retrieved (average +/- SD per file card 6.9 +/- 8.4). Three hundred thirty-four cited statements were verified (average per file card 15.9 +/- 24.7). Forty-two percent of references were human controlled trials, 17 percent not available, 11 percent review articles, 10 percent manufacturer information, 8 percent human uncontrolled, 4 percent nonhuman controlled, 3 percent retrospective, 2 percent tertiary literature, 1 percent epidemiologic studies, 1 percent prospective, and 1 percent editorial. Thirty-three percent of references were found in peer-review journals. Of all cited references, 73 percent were retrievable. Thirty-nine references were not retrievable: 24 were not held by local libraries and 15 were manufacturers' information. Eighty-five percent of references contained no errors in reference documentation. For accuracy of statements, 46.1 percent were correct, 29 percent not available, 15.3 percent misleading, 4.2 percent incorrect, and 5.4 percent were cited from the abstract, discussion, or conclusion section of the study.

Most references cited in pharmaceutical promotional brochures referenced controlled studies that are retrievable. However, incorporation of more adequately controlled studies, more references from peer review journals, and careful evaluation of referenced statements by the pharmaceutical company, journals' editorial board. Food and Drug Administration, and healthcare professionals may be beneficial.