[Graduation in Medicine: admission criteria vs. scientific production]

Medical postgraduation in Brazil is divided into two categories: lato sensu (residency and fellowship) and stricto sensu (MSc, PhD), being the former a prerequisite to the latter. The author presents a study comparing two postgraduation courses in which both lato and stricto sensu are available, in order to determine if their selection criteria for the MSc/PhD courses are protective for their own residents (endogenic) and the consequences to their scientific production. The information concerning selection criteria and endogenic proportion was obtained directly from the courses' offices. To evaluate the scientific production, the Ellwein et al. (1989) method, modified, was used: number of papers published in the two major Brazilian journals of the specialty, in which there was, at least, one post-graduation student and the position in the byline as well. Period considered: 1983-1992. For ethical reasons the courses were named by greek letters "beta" and "gamma". "Beta" selects mainly by the curriculum vitae. It had an endogenic proportion of 29.4% (34/116), a mean scientific production of 8.6 +/- 4.3 papers/year that, when considered proportionally (total number of papers/total number postgraduate students) gives us the value of 0.74 (74%). "Gamma" prefers a test in the specialty and the questions come from lectures given in its residency course. Respectively, the "gamma" results were: 93.5% (56/60), 15 +/- 14 and 0.25 (25%). The differences between "beta" and "gamma" were statistically significant (p < 0.001). The author concludes that a high endogenic proportion acts negatively in the scientific production of a stricto sensu (Master and PhD) Brazilian course. Different ways to select are also discussed.