Journal impact factors, h indices, and citation analyses in toxicology.

Academic departments, institutions, and funding sources are increasingly interested in quantifying the academic's productive output and quality of individual researchers. Since Gross and Gross first published a detailed analysis of a single journal's bibliography in 1927 [1], there has been a progressive increase in the scientific methods of journal citation quantification. The most commonly used databases are Journal Citation Reports (JCR) and the Science Citation Index (SCI) [2] produced by the Institute for Scientific Information (ISI). Originally introduced in 1961 as a means for retrieving bibliographic data, the SCI has undergone numerous changes, both in its content and methods, as well as how the information from the database is used. The recent expansion of biomedical knowledge and increasingly sophisticated scientific techniques has led to a proliferation of biomedical journals. The JCR is an annual publication by the ISI, which reports the impact factor and other bibliometric data for thousands of journals-the JCR cited 6,164 journals in 2006, compared with just 4,625 in 1995 [3]. In a remarkably prescient article published over 70 years ago, Bradford (4) found that a small percentage of journals accounts for a large percentage of what is published and that an even smaller percentage of journals accounts for what is most often cited. That is, there are diminishing returns in trying to comprehensively cover the world's body of literature. Careful selection thus is an effective way to avoid "documentary chaos," a phrase coined by Samuel C. Bradford referring to the angst felt when trying to keep up to date with the information explosion.