Background: Molecular techniques for microbial identification have highlighted the relevance of polymicrobial infections in humans, such as those affecting the urinary tract. Although in vitro investigations have demonstrated connections between co-infections and microbial interaction, their role is unclear in clinics, given the overlap with host conditions. Objective: We aimed to separate the roles of organisms and patient conditions in human polymicrobial urinary samples by performing a relevant epidemiological analysis. Methods: We analyzed retrospective results from urine cultures performed during one year in a 1200 beds Italian hospital. Patients were grouped as uncompromised and compromised and positive urine cultures were grouped as monomicrobial and polymicrobial. We assessed associations between single microorganisms and the groups of positive samples and between single microorganisms and the group of patients through a multivariate logistic regression model, adjusting by the confounding effect of seven variables. Results: We enrolled 24,067 urine samples, among which 7208 were positive, 75% monomicrobial and 25% polymicrobial. We found that the polymicrobial samples had a microbial scenario wider than the monomicrobial ones and the organisms most sampled had the highest number of different pairwise associations. Certain organisms shown having absolute numerical advantages in the polymicrobial urine cultures with respect to the monomicrobial ones, independently of host's conditions. Conclusions: The numerical advantage shown by certain organisms in polymicrobial urine samples over monomicrobial samples supports the hypothesis of microbial synergies favouring the occurrence of certain co-infections.
Keywords: UTI; compromised host; epidemiology; microbial interaction; microbial synergy; polymicrobial infections; uncompromised host.