Experience, expertise, or specialty? Uses and misuses of a reference.

OBJECTIVE: To analyze systematically the manner in which the results of a published study are presented in subsequent publications that refer to it. STUDY DESIGN: We identified a convenience sample of 121 scientific papers that referred to an often-cited 1996 study by Kitahata and colleagues. This study reported that greater primary care physician experience with AIDS was associated with lower mortality among their patients with AIDS. OUTCOMES MEASURED: We determined the manner in which the results of the Kitahata and coworkers study were presented, the type of article, and whether its focus was on HIV care. RESULTS: Most of the articles reviewed (78%) appropriately referred to the study as evidence of improved outcomes with increasing provider experience. However, 8% of the articles reviewed referred to the study as evidence of improved outcomes with specialty care and 3% referred to it as evidence of the benefits of expert care. Articles that referred to the study as evidence of improved outcomes with specialty care were more likely to be review articles and articles with a non-HIV focus. CONCLUSIONS: This study demonstrates that misrepresentation of the findings of published studies is not uncommon. More needs to be done to ensure the accuracy of references in scientific publications.