Does Antibiotic Use Contribute to Biofilm Resistance in Sink Drains? A Case Study from Four German Hospital Wards

<b>Backgound</b>. As biofilms are known to harbour (multi-)resistant species, their presence in health settings must be considered critical. Although there is evidence that bacteria spread from drains to the outside, there is still a lack of research data focusing on drain biofilms from hospitals. <b>Methods</b>. We sampled biofilms from various wards of Helios Hospital Krefeld (Germany), where comprehensive antibiotic consumption data were available. Biofilms were analysed by cell counting, isolation of relevant bacterial groups and genetic and phenotypical resistance parameters. Data were correlated with the prescribed antibiotics of the respective ward. Furthermore, an ex situ biofilm model was employed to investigate the influence of sub-inhibitory antibiotics on the bacterial community and the prevalence of class 1 integrons. <b>Results</b>. Our results show that every ward harboured medically relevant bacterial species. While no significant differences were found in cell counts, the median prevalence of the resistance marker gene <i>intI1</i> correlated with the amount of prescribed antibiotics. In contrast, phenotypical resistances showed no similar tendency. In addition, melting curve analysis data and changes in <i>intI1</i> prevalence show that the composition of the bacterial community shifted depending on the biofilm and antibiotic. <b>Conclusions</b>. To the best of our knowledge, our study is the first considering possible correlations between the consumption data of hospital wards and resistances in drain biofilms the way we did. Based on our results, we conclude that sub-inhibitory concentrations of antibiotics have no general effect on biofilms in terms of bacterial community shift and occurrence of antibiotic-resistant species. Amongst other things, the effect depends on the initial composition of the bacterial community, the antibiotic used and the intrinsic bacterial resistance, e.g., prevalence of class 1 integrons.