Article type
Year
Abstract
Background: It is important to measure and understand the various ways in which Cochrane reviews (CRs) are used. Citation rates are commonly used but only capture some aspects of usage.
Objectives: To compare different citation systems and patterns and CR summary accesses in order to gain a better understanding of how these might reflect CR usage.
Methods: (1) We selected 200 CRs published in 2009 (first 50 from each issue; 40% of all 2009 CRs) and compared internal citations from the ‘cited by’ tab with Web of Science and SCOPUS citations. (2) We compared Cochrane Collaboration websiteaccesses to the top 50 most accessed CR summaries for February–April 2012 with internal cites to these CRs from The Cochrane Library and cites from Google Scholar. (3) We examined what proportion of ‘top 50’ SCOPUS cites were for superseded versions of CRs.
Results: (1) Average ever cites for the sample of 200 CRs (current version at as 2009) were 2.4 from The Cochrane Library, 8.6 from Web of Science and 16.3 from SCOPUS. (2) The ‘Top 50’ showed an average ever cite rate of 13 from The Cochrane Library, 205 from Google Scholar (both for current CR versions) and an average 1256 website accesses over 3 months. (3) There was a mean of 21% (median 11%; range 0–100%) cites of superseded CRs.
Conclusions: The large discrepancy in citation rates from The Cochrane Library, Web of Science (four-fold compared with The Cochrane Library) and SCOPUS (eight-fold) indicate incomplete capture of cites by Cochrane Library in particular. Internal CR cite rates for highly accessed ‘top 50’ CR summaries are six-fold higher than those for 2009 CRs but are very much lower than Google Scholar cites and Cochrane website accesses. We need to further address the reasons for citation discrepancies and continuing citation of superseded CRs and find solutions.
Objectives: To compare different citation systems and patterns and CR summary accesses in order to gain a better understanding of how these might reflect CR usage.
Methods: (1) We selected 200 CRs published in 2009 (first 50 from each issue; 40% of all 2009 CRs) and compared internal citations from the ‘cited by’ tab with Web of Science and SCOPUS citations. (2) We compared Cochrane Collaboration websiteaccesses to the top 50 most accessed CR summaries for February–April 2012 with internal cites to these CRs from The Cochrane Library and cites from Google Scholar. (3) We examined what proportion of ‘top 50’ SCOPUS cites were for superseded versions of CRs.
Results: (1) Average ever cites for the sample of 200 CRs (current version at as 2009) were 2.4 from The Cochrane Library, 8.6 from Web of Science and 16.3 from SCOPUS. (2) The ‘Top 50’ showed an average ever cite rate of 13 from The Cochrane Library, 205 from Google Scholar (both for current CR versions) and an average 1256 website accesses over 3 months. (3) There was a mean of 21% (median 11%; range 0–100%) cites of superseded CRs.
Conclusions: The large discrepancy in citation rates from The Cochrane Library, Web of Science (four-fold compared with The Cochrane Library) and SCOPUS (eight-fold) indicate incomplete capture of cites by Cochrane Library in particular. Internal CR cite rates for highly accessed ‘top 50’ CR summaries are six-fold higher than those for 2009 CRs but are very much lower than Google Scholar cites and Cochrane website accesses. We need to further address the reasons for citation discrepancies and continuing citation of superseded CRs and find solutions.