Effects of article retraction on citation and practice in medicine.

At times, there are reasons for authors to make a formal statement of retraction of work they publish in biomedical journals. This study examines 235 retracted articles and looks at the reasons for these retractions and citations to the articles subsequent to retraction. The primary reasons for retraction are error of various kinds (such as problems with method or sample, including contamination of samples) and misconduct. The 235 articles are cited a total of 2,034 times after retraction. This set of citations can be divided into two groups: citations that appear in journals included in the Abridged Index Medicus and those that appear in other journals included in MEDLINE. While most of the citations in these two groups of journals can be categorized as "implicitly positive," 275 make explicitly positive mention of retracted articles. The implications for continued citation for biomedical research and clinical practice are discussed.