Comparison between Scopus, Web of Science, PubMed and publishers for mislabelled review papers.

The present study examined the incidence rate of reviews being mislabelled by Scopus, and compared this rate with Web of Science (WoS), PubMed and official websites of publishers. Top 400 cited publications defined by Scopus as ‘articles’ were examined. Their contents were evaluated to see if any were actually reviews. These publications were cross-checked in WoS, PubMed and publisher websites to identify the assigned document type labels. Out of the 400 Scopus ‘articles’, 117 were reviews (29.3%). The official websites of publishers had 16.0% incidence of mislabelled reviews, which was less than Scopus (29.3%) but more than WoS (14.1%) and PubMed (1.9%). [ABSTRACT FROM AUTHOR]