Retraction is a mechanism to correct scientific literature. This article aims to demonstrate that retractions have gradually increased, especially in the medical field. Methodologically, it is a literature review article whose statistical data were obtained from studies on retraction articles published in English between 2012 to 2017 in the PubMed and Google Scholar databases. Some of the most frequent causes of retraction found were plagiarism, misconduct, errors, fabrication and duplication. The highest incidence rate was reported in the United States, India, China, Japan and Germany. The highest retraction rate was for low-impact journals. The time for retraction is long but has decreased, which allows these articles to be cited and bad science to arise. It is concluded that uniformity is needed in notes and rules of retraction, retracted articles should be marked appropriately, and time for retraction must be reduced. In relation to bioethics, there is a serious problem in the integrity of scientific literature and a possible impact of retractions on the health of people.