Abstract
OBJECTIVES: We aimed to determine publication trends in leading clinical research journals (impact factor >20) during the rise of the coronavirus disease 2019 (COVID-19) pandemic and to check for an increase in publication times of non-COVID-19 original research articles. METHODS: We collected publication data from five print-based medical journals and one online journal--JAMA: The Journal of the American Medical Association, The Lancet (Lancet), The New England Journal of Medicine, Annals of Internal Medicine, The BMJ (BMJ), and BMC Medicine (BMC Med)--for the December 2019 through May 2020 period. We categorized each article as either "COVID-19-related" or "non-COVID-19-related". When available, we further extracted data on submission-to-acceptance dates and acceptance-to-publication dates for original research articles for the January through July 2019 and January through July 2020 periods. We compared the time from submission to publication for non-COVID-19 original research articles during the two periods and tested for statistical significance with a one-tailed Wilcoxon rank-sum test. RESULTS: We found that non-COVID-19-related articles began decreasing in volume as COVID-19-related articles increased. In BMJ and Lancet, the COVID-19-related articles began overtaking the non-COVID-19-related articles in number during April and May 2020. However, COVID-19-related primary research articles only began consistently appearing in journal issues during May 2020. Only BMJ and BMC Med publicly recorded complete data regarding their publication timelines. After removing outliers, we found that the mean time from submission to publication for articles published in BMJ from January through July 2019 was 204 days (median: 194 days; IQR: 163-236), and from January through July 2020 was 223 days (median: 218 days; IQR: 177-261) (p=0.04). In BMC Med, the mean time from submission to publication from February through July 2019 was 153 days (median: 150 days; IQR: 123-181), and from February through July 2020 was 163 days (median: 157 days; IQR: 132-191) (p=0.06). Conclusion: We discovered a steadily increasing trend in the percentage of COVID-19-related articles and a concomitant decreasing trend in the percentage of non-COVID-19-related articles published in high-impact print journals during the period from December 2019 through May 2020. For non-COVID-19-related articles published in BMJ, we found a statistically significant increase upon comparing the submission-to-publication times for the period from January through July 2020 with the submission-to-publication times for the period from January through July 2019.