Same dataset, same conclusion, different in detail: Who's doing what from large public database?

摘要

We read with great interest the article by Yu et al. on the association of tuberculosis with lung cancer. This retrospective cohort study indispensably proved the positive relationship, thanks to the nationwide scale and large number of cases they analyzed from the Taiwan National Health Insurance Database. This administrative database has been widely used to produce scientific knowledge. From 2000 until 2009, no less than 383 articles have been published, and 29 (7.5%) of them are related to cancer. However, because it is publicly available, there is a chance that people from different groups work on the same topics from the same dataset at the same period. It coin-cidentally occurred when similar research by Wu et al. was published in the February 2011 issue of Cancer journal (Epublication by September 30, 2010).