Retractions are linked by the Web of Science to original reports

         23 March 2006     Annals of  Internal  Medicine  2006; 0: 0000605-200604180-00123-7
         Eugene Garfield, Ph.D.  (
         Marie McVeigh ( and Marion Muff (
         Thomson ISI, Philadelphia, PA 19104

In response to: Harold C. Sox, and Drummond Rennie “Research Misconduct, Retraction, and Cleansing the Medical Literature: Lessons from the Poehlman CaseAnnals of Internal Medicine (18 April 2006 Volume 144 Issue 8) 

Dear Editor: 

The recent article about the importance of integrating retraction notices with their original reports noted their treatment in PubMed, but failed to take into account the procedures followed in the Science Citation Index (SCI) the electronic version of which is included in the Web of Science. 

Ever since the SCI was launched in the sixties, published retractions have been indexed by SCI. In each case a citation link was established between the retraction, that is “correction,” and the original source article. To find retractions, like all other corrections, all one had to do was conduct a cited reference search based on the author, journal and year of the retracted paper. You would then see a list of all items that cited the original work including the retractions, which like all other corrections would be coded as such. However, since 1996 the treatment of retractions was amplified by including the notation for the retraction together with the bibliographic citation for the source item. If one does a search on a subject or an author and finds a paper which has been retracted, the retraction can be seen immediately adjacent to the source entry. 

Thus, the retraction entry for WS Hwang’s paper on “Patient-specific embryonic stem cells derived from human SCNT blastocyst (Retraction of vol 308, pg 1777, 2005) is followed by SCIENCE 311 (5759): 335-335 January 20 2006. 

When you conduct a cited reference search on the original paper at Hwang HS, Science, 2005, you immediately see the statement that “this article was retracted see Science 311, 335, Jan. 20, 2006” 

In previous generations authors often unwittingly cited retracted research because they did not or could check citation indexes. Today there is no excuse since access to PubMed and Web of Science is widely available. 

Eugene Garfield, Chairman Emeritus 
Marie McVeigh, Senior Manager, JCR & Bibliographic Policy 
Marion Muff, Bibliographic Policy Manager 
Institute for Scientific Information Thomson Scientific 3501 Market Street Philadelphia, PA 19104 

Conflict of Interest:
The authors are all connected to the purveyor of the Science Citation Index and Web of Science.