Nicole van Leuven (Kleve / DE; Bonn / DE), Anna Dicks (Kleve / DE), Ralf Lucassen (Kleve / DE), Patrick Braß (Krefeld / DE), Combat Consortium (Cardiff / GB), André Lipski (Bonn / DE), Dirk Bockmühl (Kleve / DE)
Introduction & Aim
Since biofilms are known to contain resistant bacterial species and many hospital-acquired infections are believed to be biofilm-associated, their occurrence in hospitals is critical. As drains get in contact with antimicrobial residues like antibiotics, these may act as resistance drivers. Class 1 integrons as a multi-resistance marker allow for monitoring changes in resistances, checking for integrase gene intI1.
Our research aimed to investigate the impact of antibiotics on the composition of hospital drain biofilms and their genetic and phenotypic resistance profile based on consumption data of different wards and experimental approaches.
Methods
20 bathroom sink biofilms were sampled from 4 wards of a hospital. The cell count of viable cells, Enterobacteriaceae and yeasts/moulds was determined. Additionally, relevant pathogens were isolated, identified and analyzed for resistances with VITEK®2 compact system and by Epsilometer-testing. intI1-prevalence was measured by qPCR. Further testing took place by adding different antibiotics to the media during biofilm growth. After DNA extraction, changes of DNA in a high resolution melting analysis (HRMA) and the intI1 prevalence were evaluated.
Results
Our data show that cell counts of the wards were not significantly different. MIC-values for antibiotics obtained by VITEK®2 and Epsilometer-testing showed no correlation to the administration of recommended daily doses (RDD)/100 bed days. In contrast, the median intI prevalence correlated with the RDD. While adding sub-inhibitory concentrations of antibiotics during growth, differences in HRMA were observed depending on the biofilm and antibiotic (Fig.1). Moreover, shifts in HRMA data correlated with changes in intI1 prevalences.
Summary
Our approach gave an insight into correlations between resistance values and the RDD of hospital wards. Expected pathogens were found. Genetic resistance markers correlated with the amount of prescribed antibiotics, although phenotypes didn´t. Thus, a selective pressure by sub-inhibitory antibiotic concentrations were found to cause community shifts accompanied by changes in the intI1 prevalence.