Pitfalls and problems - duplicate publications

Some studies result in more than one publication. Authors may publish the methods of the study, present preliminary data at a conference resulting in an abstract, then publish some results, and later publish longer term follow-up. There's nothing wrong with this, as long as you can tell it's all about the same study.

Sometimes, studies may be published more than once for other reasons - more publicity, more papers on the author's CV, or to allow different authors to be first authors. Sometimes, it's not easy to tell whether they are reports of the same study. This can cause problems for reviewers because we might count a study more than once and so give extra weight to it in our review.

This phenomenon was studied by Martin Tramèr and colleagues (Tramèr MR et al BMJ 1997;315:635-640).

They found that 17% of reports of trials of a drug were duplicates and 28% of the data were duplicated. If this had not been spotted and a review had been done counting studies more than once, the drug would have appeared more effective than it actually was. So, we need to be alert to the possibility of duplicates. Look for

• same authors in different orders
• similar study inclusion and exclusion criteria
• many reports of a study done in the same place at the same time
• results tables that look familiar

Now you understand the process of selecting studies for your review, return to Module 7 and we will work on designing this Section of your data collection form