Background: Defined by the emergence of antibiotic resistant strains, Staphylococcus aureus is a priority bacterial species with high antibiotic resistance. However, a rise in the prevalence of penicillin-susceptible S. aureus (PSSA) bloodstream infections has recently been observed worldwide, including in Australia, where the proportion of methicillin-susceptible S. aureus causing bacteremia identified phenotypically as penicillin-susceptible has increased by over 35%, from 17.5% in 2013 to 23.7% in 2020. Objectives: To determine the population structure of PSSA causing community- and hospital-onset bacteremia in Australia and to evaluate routine phenotypic antimicrobial susceptibility methods to reliably confirm penicillin resistance on blaZ-positive S. aureus initially classified as penicillin-susceptible by the Vitek® 2 automated microbiology system. Results: Whole genome sequencing on 470 PSSA collected in the 2020 Australian Group on Antimicrobial Resistance Australian Staphylococcus aureus Sepsis Outcome Programme identified 84 multilocus sequence types (STs), of which 79 (463 isolates) were grouped into 22 clonal complexes (CCs). The dominant CCs included CC5 (31.9%), CC97 (10.2%), CC45 (10.0%), CC15 (8.7%), and CC188 (4.9%). Many of the CCs had multiple STs and spa types and, based on the immune evasion cluster type, isolates within a CC could be classified into different strains harboring a range of virulence and resistance genes. Phylogenetic analyses of the isolates showed most CCs were represented by one clade. The blaZ gene was identified in 45 (9.6%) PSSA. Although multiclonal, approximately 50% of blaZ-positive PSSA were from CC15 and were found to be genetically distant from the blaZ-negative CC15 PSSA. The broth microdilution, Etest® and cefinase, performed poorly; however, when the appearance of the zone edge was considered; as per the EUCAST and CLSI criteria, disc diffusion detected 100% of blaZ-positive PSSA. Conclusions: In Australia, PSSA bacteremia is not caused by the expansion of a single clone. Approximately 10% of S. aureus classified as penicillin-susceptible by the Vitek® 2 harbored blaZ. Consequently, we recommend that confirmation of Vitek® 2 PSSA be performed using an alternative method, such as disc diffusion with careful interpretation of the zone edge.
Introduction.Enterococcus faecium clade B has recently been re-classified as Enterococcus lactis. Although E. lactis was previously associated with food products and probiotics, the recent re-classification has prompted the need for the accurate identification of this species and re-interpretation of its disease-causing ability. Since the re-classified E. lactis can currently only be identified by molecular techniques such as whole-genome sequencing, we constructed a MALDI Biotyper® custom database to rapidly identify and differentiate E. lactis causing bacteraemia from E. faecium. Hypothesis/Gap statement. The re-classification of E. faecium clade B as E. lactis warrants the development of rapid and accurate identification methods to distinguish these species, particularly in clinical settings where E. lactis may be misidentified as E. faecium. Aim. The aim of this study was to construct a MALDI Biotyper® custom database to rapidly identify and differentiate E. lactis causing bacteraemia from E. faecium. Methodology. A total of 97 enterococcal isolates, including 38 E. lactis, 51 E. faecium and 8 non-E. faecium non-E. lactis enterococci (E. avium, E. casseliflavus, E. cecorum, E. durans, E. faecalis, E. faecium, E. gallinarum, E. lactis, E. mundtii and E. raffinosus) were investigated. Whole-genome sequence analysis was used to confirm the species of each isolate. A MALDI Biotyper® in-house database was constructed using 29 E. lactis isolates and the ethanol/formic acid/acetonitrile preparation protocol. The in-house database was validated using the 97 enterococcal isolates and the extended direct transfer preparation protocol. Results. Our in-house database correctly identified all isolates at the species level, including the E. lactis isolates, all of which were misidentified as E. faecium by the BioTyper® MBT Compass reference library (2022). Of the 38 E. lactis isolates, 84.2% (n=32) were identified at the high probable species level (score ≥2.300), while the remaining 15.8% (n=6) were identified at the probable species level (score 2.000–2.299). Similarly, all E. faecium isolates (n=51) were accurately identified, including 84.3% (n=43/51) identified at the high probable species level and 15.7% (n=8/51) identified at the probable species level. Conclusion. Our study provides a ready-to-use custom MALDI spectral database that can be implemented in clinical diagnostic and research laboratories to accurately identify E. lactis, which is currently misidentified as E. faecium by the standard spectrum database available on commercial platforms.
Background: Prior research has linked socioeconomic conditions, governance and infrastructure to global variations in antimicrobial resistance (AMR), independently of aggregate antibiotic consumption. Cross-national comparisons of antimicrobial use and resistance, however, are limited by heterogeneous data definitions and data quality. We examined whether associations with socioeconomic, temporal or geographic areas are evident within a single national health system using high-quality longitudinal surveillance data Methods: We analysed community-onset E. coli bloodstream isolates across Australia between 2013 and 2024. Resistance to 11 antibiotics was modelled individually with logistic regression, and panel drug resistance (PDR) across 5-, 6-, and 12-drug panels with ordered probit models. Models included age, sex, area-level socioeconomic and demographic covariates, community antibiotic dispensing, and the temporal pattern was assessed using a three-segment time spline spanning the pre-COVID-19, COVID-19, and post-COVID-19 periods. Findings: 40,102 isolates from 87 Statistical Area Level 4 (SA4) regions were included. Higher household occupancy, higher population density, high immigrant population, and higher manufacturing employment were associated with higher odds of resistance across most antibiotics. Female sex was associated with lower odds of resistance for most agents, except ampicillin and trimethoprim–sulfamethoxazole. Resistance varied non-linearly with age, generally peaking in early to middle adulthood. Associations between area-level antibiotic dispensing and resistance were weak and frequently inverse. Resistance declined during the period of COVID-19 restrictions and increased after restrictions were relaxed. Interpretation: Population-level resistance was strongly patterned by socioeconomic and demographic conditions plausibly related to exposure and transmission, whereas local total antibiotic dispensing showed no consistent positive association with resistance. These findings do not diminish the established role of antibiotic exposure in selecting resistance within individuals but suggest that population patterns cannot be explained by consumption alone. AMR control should therefore complement antimicrobial stewardship with greater attention to the transmission and acquisition of resistant organisms.
ORon behalf of the Australian Group on Antimicrobial Resistance
on behalf of the Australian Group on Antimicrobial Resistance
Background: Antimicrobial resistance (AMR) is widely considered to be driven by antimicrobial consumption through within-host selection. However, whether this mechanism adequately explains population-level patterns of resistance in invasive infections remains uncertain. If antimicrobial use is the dominant determinant, resistance should be highest in demographic groups with the greatest exposure. Methods: We conducted a retrospective analysis of 44,792 community-onset Escherichia coli bloodstream infection episodes identified through national Australian surveillance data (2013–2024). Resistance prevalence across individual antimicrobials and composite multidrug resistance panels was analysed by age and sex. These data were compared with community antimicrobial dispensing derived from the Pharmaceutical Benefits Scheme. Mean resistance was modelled as a function of age and sex. Results: Antimicrobial use was substantially higher in females than males (~23% overall) and increased markedly with age, with individuals aged ≥80 years receiving approximately three times more antimicrobials than those aged 25–30 years. In contrast, resistance was consistently lower in females across most antimicrobials and composite measures. Resistance demonstrated an inverted U-shaped age distribution, peaking at 30–40 years before declining in older age groups. From early adulthood to older age, antimicrobial dispensing increased threefold, whereas mean resistance declined by approximately 20%. These patterns were consistent across antimicrobial classes, years, and jurisdictions. Conclusions: These findings show that demographic patterns of antimicrobial resistance in community-onset E. coli bloodstream infections are not well explained by a simple population-level consumption model. These findings should be interpreted as important hypothesis-generating insights. Although antimicrobial exposure remains important for individual-level selection, the observed discordance between prescribing and resistance suggests that other factors, including differences in transmission pathways, healthcare contact, disease prevalence, community sanitation and socioeconomic circumstances may also significantly shape resistance patterns.