Scientific misconduct during the pandemic: Retractions in the field of COVID-19

Abstract

The article aims to understand the characteristics of the retracted articles during the COVID-19 pandemic. It aims to identify the countries with the highest number of retractions and the correlation of retractions with funding status, impact factor and collaborations. The study utilised the Retraction Watch Database to identify the retracted articles in the area of COVID-19. The requisite details for each retracted article were recorded, such as title, cause of retraction, date of publication and date of retraction. The impact factor of the journals was ascertained from the Journal Citation Reports (JCR-2021) of Clarivate Analytics, and the causes of retraction were categorised under seven major headings according to Charlesworth Author Services. Further, the study used a chi-squared test to determine the association or relationship between the studied variables. As of December 2022, 264 COVID-19 articles were retracted, of which a large proportion (36, 18.27%) were retracted just after 1 month of publication. The retracted articles were published mainly in journals (224, 84.84%), with 40 (15.15%) articles available on the preprint servers. A significant proportion of retractions were initiated by the authors, editors & authors & editors jointly. However, 6.06% of articles did not mention the retraction authority. Most retractions are due to honest error (131, 49.62). The other causes of retractions include ethical misconduct, ethics violation, conflict of interest and peer-review fraud. Among the countries, the highest number of retractions are credited to the United States (59, 22.34%), China (41, 15.53%) and Malta (30, 11.36%). All the retracted articles were available in the open access mode, with 44 (16.66%) articles funded by different funding organisations. The study reveals that non-funded articles have a higher retraction rate than the funded ones. The study also indicates an inverse relationship between the retraction of articles and journal impact factor, indicating that the higher the impact factor of journals, the lower the retraction rate. There is also a direct relationship between authorship and retractions, i.e. the higher the number of authors, the greater the chances of retraction. Also, the articles having a national collaboration are retracted more than the international ones. The study's main limitation is evaluating a limited set of retractions covered by a single database, which is inherited with limitations compared with other databases. The rush to publish during the pandemic poses threats, which would quickly outdate the study's findings with the outgrowth of retractions. Also, retractions can happen even after a long time, confining the generalisation of results. Retraction of published articles has far-reaching consequences, particularly during the pandemic when a huge influx of publications determines the action-treatment plan for a disease. The study helps to understand the characteristics of retracted articles that may help prevent the dissemination of flawed information during health emergencies. The study highlights the corrective mechanism and its characteristics for scientific misconduct prevalent during COVID-19 pandemic. It provides a thorough understanding of article retractions in the field of COVID-19.