Background. This study examines changes in the authorship patterns of
chairmen of medicine departments from 1979 through 1990, a time when c
riteria for authorship were widely debated among medical journal edito
rs. Chairmen were selected to be studied because, due to their leaders
hip position and stature, they are likely to ''set the example'' for m
any scholars in the academic medicine community. Method. Computerized
literature searches were conducted for all 233 persons who were chairm
en during at least one year of the 12-year study period (1979-1990) to
determine the average number of publications per chairman per year, t
he average number of coauthors per article published, the location of
the chairman on the published list of authors, and other pertinent dat
a required for a thorough analysis of authorship practices. Statistica
l methods included repeated-measures analysis, sensitivity analyses, a
nd the use of linear models to analyze trends over time. Results. Amon
g all the chairmen, there was no statistically significant decrease in
the average number of articles per year for which the chairmen were l
ast authors, and there was a significant increase in the average numbe
r of coauthors per year on published articles. However, factors such a
s length of chairmanship, frequency of publication, and academic posit
ion (i.e., chair, pre-chair, or post-chair) were found to influence pa
tterns of authorship; for example, the 61 individuals who were chairme
n for ten years or more had a significant decrease in the number of la
st-author articles, whereas the 172 individuals who were chairmen for
fewer than ten years had a significant increase. Conclusion. Changes i
n the definition of authorship by medical journal editors did not seem
to be important influences on the authorship patterns of the chairmen
, since the chairmen did not significantly decrease the number of arti
cles for which they were the last authors, and they significantly incr
eased the number of coauthors of their articles.