Refining the Computation of Topic Based Impact Factor :
Some Suggestions

Occupational Medicine (Oxford) 49:(8) 571, November 1999

Dear Sir:

The proposal of "an alternative to journal-based impact factors" by Takahashi et al1 to compute the average impact of articles on a particular topic is an interesting approach. It has the potential to solve the problem of estimating the impact factors (IF) for articles in relatively small fields. Although there have been many articles proposing modified journal IFs, I cannot recall any that translated the data into article-based topic-based Ifs.

In contrast to the journal IF, the proposed topic-based IF attempts to group articles rather than journals using medical subject headings (MeSH) in MEDLINE. It, therefore, relies on the assumption of objectivity in the application of MeSH, but this unfortunately is not always true. The MEDLINE indexing system involves human indexing, which has both advantages and disadvantages. The latter includes inconsistency between indexers in applying MeSH and changing terminology.

As an alternative, the Science Citation Index (SCI) database can be used to create topical databases as Melino and I did recently in an analysis of apoptosis literature 2. Co-citation clustering would also bring in articles that would not be found in a MEDLINE search. While it is always best to obtain actual citation counts for each article from an SCI search, I believe their method will produce useful results, even though the universe of articles retrieved may not include all the articles that might be relevant to the topic chosen. I would like to see specific examples of their proposed method based on real data as a development of the hypothetical example in their letter. Their method will produce an expected citation impact based on averaging citation counts for the entire database, but only actual citation data for each paper can demonstrate the variations from the most cited to the least cited papers.


1.    back to text Takahashi K, Aw TC, Koh D, An alternative to journal-based impact factors (letter).   Occup. Med. 1999; 49:57-58.

2.    back to text Garfield E, Melino G. The growth of the cell death field: An analysis from the ISI-Science Citation Index. Cell Death Differ 1997, 4: 352-361.


Eugene Garfield
Chairman Emeritus
Institute for Scientific Information
3501 Market Street, Philadelphia, PA 19104