Clinically relevant antibiotic-resistant bacteria in aquatic environments – An optimized culture-based approach

Abstract The emergence of antibiotic-resistant clinically relevant facultative pathogenic bacteria in the environment has become one of the most important global health challenges. Antibiotic-resistant bacteria (ARB) have been found in surface waters and wastewater treatment plants. Drinking water guidelines and the EU bathing water directive 2006/7/EC include the surveillance of defined microbiological parameters on species level, while the monitoring of ARB is missing in all existing guidelines. However, standardized methods for the detection of ARB exist for clinical investigations of human materials only. They are based on cultivation on selective agar plates. These methods cannot be used directly for environmental samples, because of the high amount and diversity of bacterial background flora which interferes with the detection of human-relevant ARB. The aim of this study was to introduce a proposal for future normative standard operation procedures, with international relevance, for the culture-based detection of clinically-relevant antibiotic resistant bacteria in aquatic environmental samples like wastewater and surface water: gram-negative bacteria resistant against 3rd generation cephalosporins (ESBL) and against carbapenems (CARBA), gram-positive vancomycin-resistant enterococci (VRE) and methicillin-resistant Staphylococcus aureus (MRSA). The final adaptation of standardized cultivation methods included increasing the standard incubation temperature from 36 °C to 42 °C, which effectively inhibits the environmental background flora on agar plates while the desired target species survive. This enables the detection of target species in suitable sample volumes. Putative target colonies which belong to the remaining background flora had to be excluded by morphological and physiological differentiation. Therefore, a time and cost optimized testing scheme with good performance was developed, which allows an effective exclusion of non-target isolates in samples. Depending on the target species and sample type, sensitivity of up to 100% is achieved, and specificity ranges from 91.1% to 99.7%, while the positive predictive value, negative predicted value and accuracy rate are always >90%.