LAUSR

INTRODUCTION Research activity evaluation in French hospitals is based on the number of publications, author position (first, second, third, second-to-last, last, investigator list, and "Other") and journal category (A being the highest category followed by B, C, D, E, and NC). METHODS The profile of publications over the 2004-2014 period in terms of these indicators was evaluated. Hospitals were classified into six groups according to administrative status. Time trends were analysed by three models. One-way ANOVA followed by Tukey's test was performed. RESULTS A total of 192886 publications were analysed. The increase in the number of publications ranged from 628% for for-profit private hospitals to 141% for public teaching hospitals. The most frequent category was B for cancer centres (25%), whereas this was E in public teaching (22%) and non-teaching hospitals (28%), in not-for-profit private hospitals (25%), in the military hospital (30%), and in for-profit private hospitals (24%). The first position was the most frequent for public teaching hospitals (38%) and the military hospital (44%), whereas the "Other" position was the most frequent in cancer centres (26%), in public non-teaching hospitals (28%), in not-for-profit private hospitals (27%), and in for-profit private hospitals (29%). DISCUSSION Different patterns were identified. The author position indicated that all types of hospital are involved in research projects. This study also found that public non-teaching hospitals, not-for-profit private hospitals, for-profit private hospitals, and cancer centres collaborated with other institutions which were often distinguished by publishing in high-category journals.