An advantage of our approach is that it incorporates the opinions of a variety of respected members of the immunology community: up-and-coming leaders as well as established ones, investigators from all over the world, and leaders of all sub-subfields, both basic and clinical. The disadvantages are multiple: variable modes of polling, a variable sample size that was usually too small to allow statistical treatment of the data, lack of objective criteria, and nonproportionate sampling of researchers in countries outside the United States.
To identify the most frequently cited authors of immunologic research articles, a “high-impact” immunology database was commissioned from the Institute for Scientific Information (ISI). (See http://www.isnet.com/products/rsg/impact.html for more information on “high-impact” paper methodology.) The ISI database was scanned over the years 1990-1997. For each year, the 200 most-cited papers in journals relevant to the field of immunology were listed. The authors having more than five papers on the list were ranked according to average number of citations per paper (174 authors ranging from 70.2 to 638.5/paper), and the country of the laboratory of each was listed. In addition, panel staff determined whether each author was also cited on the virtual-congress lists.
The contribution of the United States and other nations to immunology citations from 1981-1997 is shown in Figure 2.3. Of the 174 authors identified in the scan of the ISI database, 72% including the top 112 authors identified were in US-based laboratories. The ISI database indicates that US laboratories produced 63% of the papers, which garnered 66% of the citations, in journals relevant to immunology during the period 1981-1997.
The major strengths of this mode of analysis are its relative objectivity and its providing a basis of comparison with the virtual-congress polling data. However, the analysis suffers from some flaws related to the organization of the ISI database. Data from several of the top-level general journals (such as Cell) and immunology journals (such as Journal of Experimental Medicine) were not included. The panel found that the most serious limitation caused by interpreting the results obtained with this type of database is that the cited-paper analysis lists only the first author of each paper. The traditional author order in immunology puts the principal investigator last, so a graduate student or technician could have a better chance of appearing on the list of the most frequently cited papers than a principal investigator. The manner of referencing also males it difficult to know which country the particular research was done in. In addition, the list of authors is truncated after the first 15 names, possibly excluding authors who participated in large clinical trials. An additional limitation imposed by the database was that there was no breakdown according to subfield and sub-subfield of immunology.
The panel identified four leading general journals (Science, Cell, Nature, and Blood) and one of the top journals focused specifically on immunology (Immunity). Panel members scanned the tables of contents of each of the journals for 1995-1997, identifying immunology papers in the general journals and the laboratory nationality of the principal investigator and the subfield in all the journals. In addition, a small sample from the Journal of Experimental Medicine was analyzed.