Antibiotic and antimicrobial resistance

Carbapenemase-producing Enterobacterales (CPE) cause infections, particularly nosocomially, with limited treatment options. NDM-1-producing Klebsiella pneumoniae cases have substantially increased since 2022, associated with the Ukraine war.

We aimed to investigate transmission patterns using Germany’s Integrated Genomic Surveillance (IGS), combining notifications and sequence data.

We selected NDM-1-producing K. pneumoniae cases, confirmed by isolates between 1 January 2022 and 28 February 2023. Isolates were Illumina whole genome-sequenced and linked to notifications. Clusters were defined as ≤ 12 allelic differences in core genome-wide single nucleotide variant-based genotyping. Cluster categories were: ‘no exposure abroad’, ‘exposure in Ukraine’ or ‘other exposure abroad’ if ≥ one case stayed in Ukraine or elsewhere. Follow-up of 13 clusters examined further exposure information.

Among 424 cases of most frequent sequence types, 326 (77%) belonged to 61 clusters. Seventeen (28%) clusters were associated with no exposure abroad, 33 (54%) with exposure in Ukraine, seven (11%) with other exposure abroad, and four (7%) had insufficient data. Cases in clusters with exposure in Ukraine were more dispersed, younger, and more often wound-infected than in other exposure location categories (p < 0.01). Cluster follow-up revealed one cluster with all cases from Ukraine or Russia, another with nosocomial transmission following case importation, and a third with all cases from one German hospital without exposure abroad.

Most cases were in clusters, suggesting preventable chains of transmission. Three patterns emerged: transmission abroad, transmission in German hospitals from imported cases or local outbreaks. IGS can identify where transmission could be interrupted. International cooperation needs strengthening to prevent CPE spread.

To enhance antibiotic stewardship and effectively address antimicrobial resistance (AMR), better understanding of subnational antibiotic consumption patterns is essential.

We aimed to assess antibiotic consumption in Germany using data from 2022 and integrated from two surveillance systems, focusing on regional differences by examining non-university acute care hospitals.

We used pharmacy dispensing data from 525 regional or local hospitals and 35 university hospitals, covering 46.5 million patient days (PD), nearly half of all occupied bed days nationwide, to calculate antibiotic use densities (AUD) for systemic antibiotics, expressed as World Health Organization (WHO) ATC/DDD (Anatomical Therapeutic Chemical/Defined Daily Dose) per 100 patient days (DDD/100 PD). The analysis primarily focused on consumption patterns in non-university hospitals, assessing key antibiotic groups through mixed-effects regression. For sensitivity analyses, we employed hospital-adapted daily dose definitions.

Pooled AUD for participating non-university hospitals was 51.8 DDD/100 PD, with aminopenicillins/beta-lactamase inhibitors being the most prescribed group. Regression analyses, adjusted for hospital size and ward type/admitting specialty, indicated notable regional variation. We identified statistically significant differences in antibiotic consumption, particularly for beta-lactam antibiotics, fluoroquinolones and tetracyclines. For example, several regions exhibited up to 1.4-fold higher use of first- and second-generation cephalosporins compared with the western reference region.

This study highlights substantial regional variation in antibiotic use in German acute care hospitals, underlining the importance of further investigation into influencing factors such as regional guidelines and resistance rates. The methodological approach applied here may serve as a model for other countries interested in analysing regional differences in antibiotic consumption.

Fifteen- and 20-valent pneumococcal conjugate vaccines (PCVs) offer broader protection against invasive pneumococcal disease (IPD) than PCV13. Adopting these vaccines may result in decreasing IPD incidence, antibiotic use and antimicrobial resistance (AMR) rates. If the additional serotypes in PCV15 and PCV20 are associated with AMR, AMR rate reduction could be greater than expected from reduced antibiotic consumption alone.

This retrospective analysis assessed the association between AMR and non-PCV13 serotypes in PCV15 and PCV20.

Laboratory-based surveillance data on 8,123 IPD isolates were obtained retrospectively from the Belgian Reference Centre for invasive Streptococcus pneumoniae. Isolates (n = 8,088) were serotyped and tested for AMR. Associations between vaccine serotype groups and AMR were evaluated by multinomial logistic regression. Where associations varied with patients’ age, ranges of odds ratios (ORs) are presented.

PCV15-non-PCV13 and PCV20-non-PCV13 serotypes accounted for 7.4% (n = 597) and 37% (n = 2,992) of IPD isolates respectively. Of non-PCV20 serotypes, 24% (508/2,125) were penicillin resistant. Compared with non-PCV20 serotypes, PCV15-non-PCV13 serotypes were more often associated with erythromycin (ORs: 3.59–9.43) and tetracycline (OR: 2.00) resistance, and with trimethoprim/sulfamethoxazole (OR: 0.11) susceptibility. PCV20-non-PCV15 serotypes were more often associated with amoxicillin (OR: 9.45) and cefotaxime (ORs: 5.06–82.38) resistance, and with erythromycin (ORs: 0.13–0.18), tetracycline (OR: 0.71) and penicillin (ORs: 0.05–0.46) susceptibility.

PCV20 may lead to a larger decrease in overall IPD incidence than PCV15. Although the PCV20 vaccination impact on AMR may be limited, some resistant or difficult to treat infections could be avoided. Serotype replacement might lead to infections with low level penicillin resistance increasing, but most of these should remain treatable.

Mandatory reporting of healthcare-associated infections (HCAI) in England is conducted locally by acute hospital groups and can be a large burden on healthcare staff.

We aimed to determine the case ascertainment of a potential centrally-implemented, automated HCAI surveillance system in England using preexisting data feeds at the UK Health Security Agency.

We compared monthly case numbers submitted between 1 April 2016 and 31 March 2023 by acute hospital groups (locally-implemented surveillance) to routinely-collected laboratory and hospital encounter records (centrally-implemented surveillance) for all infections under mandatory surveillance in England. Since laboratories can serve multiple hospitals, we compared several methods of assigning laboratory-confirmed cases to hospital groups.

Locally-implemented vs centrally-implemented surveillance identified: meticillin-resistant Staphylococcus aureus bacteraemias 5,453 vs 5,859 (ratio 1.07), meticillin-susceptible S. aureus bacteraemias 84,680 vs 83,326 (0.98), Escherichia coli bacteraemias 281,100 vs 275,133 (0.98), Klebsiella species bacteraemias 65,877 vs 67,301 (1.02), Pseudomonas aeruginosa bacteraemias 25,862 vs 25,715 (0.99), Clostridioides difficile infections (CDI) 94,054 v 90,942 (0. 97) respectively. Assigning hospital groups by linking laboratory records to hospital encounters produced lower monthly mean absolute difference (MAD) vs locally-implemented surveillance than using laboratory records alone. MAD was 0.65 cases/month for bacteraemias, 2.99 for CDI; differences occurred in both directions. MAD decreased over time for bacteraemias but increased from April 2021 onwards for CDI.

Centrally-implemented surveillance could be feasible for bacteraemias in England due to comparable case numbers with local surveillance. However, more research is needed around understanding and managing data quality of automated feeds, particularly for CDI.

Data for healthcare-associated infections (HAI) and antibiotic use in long-term care facilities (LTCF) in Switzerland are lacking but are necessary to take actions.

We aimed to estimate HAI prevalence and antibiotic use and to record existing structure and process indicators in the area of infection prevention/antibiotic use in Swiss LTCF.

We invited all Swiss LTCF for this PPS in September 2024 using the adapted Healthcare-Associated Infections in European Long-Term Care Facilities (HALT)-4 protocol. The proportion of residents with HAI and systemic antibiotic treatment was calculated for a representative sample, stratified by language region and size. We assessed resident-level and institutional risk factors for HAI in all participating institutions, using random-effects logistic regression.

We included 94 LTCF (7,244 residents), whereof 49 LTCFs (3,375 residents) belonged to the representative sample. Median age of residents in the representative sample was 87 years (range: 36–107) and 2,334 (69.2%) were female. Prevalence of HAI was 2.2% (95% confidence interval (CI): 1.7–2.7); 2.7% (95% CI: 2.2–3.3) were receiving antibiotic treatment, with highest use in LTCF in French-speaking cantons (5.9%; 95% CI: 4.2–7.5). Urinary tract (46%) and respiratory infections (20%) were most common, aminopenicillins (26%) and nitrofurantoin (19%) the most commonly used antimicrobials. The strongest independent risk factor for HAI was presence of urinary catheters (adjusted odds ratio = 2.65; 95% CI: 1.71–4.11).

Prevalence of HAI and antibiotic use in Swiss LTCFs were comparable to the European average from 2023/24. There are regional differences in antibiotic consumption. Urinary catheterisation, potentially modifiable, was the most important risk factor for HAI.

Carbapenemase-producing Enterobacterales (CPE) frequently cause nosocomial outbreaks. To investigate these, tracing focused on patients with related CPE strains and spatiotemporal contact (e.g. contact with each other in a room or on a ward during overlapping periods) has limitations. Moreover, as widely available molecular typing methods cannot detect plasmid-related transmissions, carbapenemase gene transfer across enteric bacteria through plasmids in hospitals remains poorly understood.

Because whole-genome sequencing (WGS), particularly long-read sequencing, can offer insights into bacterial relationships both at core-genome and plasmid levels, we tested its utility, using VIM-CPE as example, to investigate plasmid and CPE spread in a hospital beyond outbreaks.

We included inpatient episodes from 2018 to 2021 involving blaVIM-bearing CPE isolates. Short- and long-read WGS data were combined with patient movement information to identify genomically related hospital-acquired VIM-CPE and putative transmission routes.

Among 43 included inpatient episodes, 27 isolates were hospital-acquired, with 23 genomically related based on core-genome or plasmid analyses. For 14 of these 23 isolates, patient movement data supported suspected transmission events. Plasmid and core-genome level analyses revealed that most transmission events did not temporally concur, occurring over up to 33 months. Thus, conventional infection tracing methods focusing on concurrent spatiotemporal contact missed a substantial proportion of transmission events.

With our findings, we advocate for broader epidemiological investigations of temporal connections if genomic data suggest relatedness. We emphasise considering plasmid transfer alongside analyses of core-genome relatedness of bacteria beyond patient contact events to study CPE and resistance spread, and guide infection control strategies.

Staphylococcus haemolyticus (SH) is an opportunistic pathogen associated with nosocomial infections, particularly bacteraemia in neonates. Epidemiological trends and genetic diversity of these infections worldwide are largely unknown.

To investigate an increase in SH vascular catheter-related bacteraemia in neonates and describe the molecular epidemiology in France between 2019 and 2023.

We analysed clinical and microbiological surveillance data from the French national surveillance network for central catheter-related (venous and umbilical) infections between 2019 and 2023. We also performed genomic and phylogenetic analyses of 496 strains isolated both inside (n = 383 from neonates, staff and environmental samples) and outside (n = 113 from adults) the neonatal intensive care unit (NICU) settings.

The proportion of SH among the 474 reported cases of nosocomial bacteraemia increased from about 20% to 30% over 5 years, mainly affecting very low birth weight preterm neonates (≤ 1,500 g). The ST29 sequence type (ST) not prevalent in previous studies was predominant, accounting for 74% of NICU strains. ST29 was characterised by phenotypic multidrug resistance to at least six classes of antibiotics (oxacillin, quinolones, gentamicin, cotrimoxazole, clindamycin and rifampicin), which distinguished it with good sensitivity and specificity from other prevalent multidrug-resistant STs identified (ST1 and ST25). ST29 strains more frequently harboured the drfG, vga-LC and mupA genes and a triple point mutation (D471E, I527M and S532N) in the rpoB gene.

The present study highlights the success of a highly resistant ST29 lineage in French NICUs mainly affecting very low birth weight premature neonates.

Community-associated Clostridioides difficile infections (CA-CDI) have increased worldwide. Patients with CDI-related symptoms occurring < 48 hours after hospitalisation and no inpatient stay 12 weeks prior are classified as CA-CDI, regardless of hospital day attendances 3 months before CDI onset. Healthcare-associated (HA) CDIs include those with symptom onset ≥ 48 hours post hospitalisation.

To consider an incubation period more reflective of CDI, and changing healthcare utilisation, we measured how varying surveillance specifications to categorise patients according to their CDI origin resulted in changes in patients’ distribution among CDI origin categories.

New CDI cases between 2012–2021 from our hospital were reviewed. For patients with CA-CDI, hospital day attendances in the 3 months prior were recorded. CA-CDI patients with hospital day attendances and recently discharged CDI patients (RD-CDI; CDI onset 4–12 weeks after discharge) were combined into a new ‘healthcare-exposure’ category (HE-CDI). Time from hospitalisation to disease onset was varied and the midpoint between optimal and balanced cut-offs was used instead of 48 hours to categorise HA-CDI.

Of 1,047 patients, 801 (76%) were HA-CDI, 205 (20%) CA-CDI and 41 (4%) were RD-CDI. Of the CA-CDI cohort, 45 (22%) met recent HE-CDI criteria and, when reassigned, reduced CA-CDI to 15%. Sensitivity analysis indicated a day 4 cut-off for assigning HA-CDI. Applying this led to 46 HA-CDI reassigned as CA-CDI. Applying both HE and day 4 criteria led to 72% HA-CDI, 20% CA-CDI, and 8% HE-CDI (previously RD-CDI).

CDI surveillance specifications reflecting healthcare exposure and an incubation period more characteristic of C. difficile may improve targeted CDI prevention interventions.

Previous United Kingdom campaigns targeting antimicrobial resistance (AMR) recommended running multimedia campaigns over an increased timeframe. The 3-year-long Keep Antibiotics Working (KAW) campaign was a mass media campaign in England targeting the public and general practitioners (GPs).

Every year, pre- and post-campaign questionnaire data were collected from the public, whereas post-campaign interview data were obtained from GPs. Data were weighted to allow pre- and post-campaign comparisons between independent samples. Significant changes in nominal and ordinal data were determined using Pearson’s chi-squared (X²) and Mann–Whitney U tests, respectively.

Prompted campaign recognition was high, increasing by 6% from 2018 to 2019 (2017: data unavailable; 2018: 68% (680/1,000); 2019: 74% (740/1,000); X² = 8.742, p = 0.003). Knowledge regarding declining antibiotic effectiveness when taken inappropriately improved following the campaign (net true: pre-2017 = 69.1% (691/1,000); post-2019 = 77.6%; (776/1,000); X² = 5.753, p = 0.016). The proportion of individuals reporting concern for themselves or for children (≤ 16 years) about AMR increased by 11.2% (Z = −5.091, p < 0.001) and 6.0% (Z = −3.616, p < 0.001) respectively, pre- to post-campaign. Finally, in 2017, reported confidence to say no to patients requesting antibiotics differed significantly between GPs who were and were not aware of the campaign (net agree: 98.9% (182/184) vs 92.4% (97/105) respectively; X² = 4.000, p = 0.045).

A high level of prompted campaign recognition was achieved. The KAW campaign improved aspects of AMR knowledge and certain attitudes towards appropriate antimicrobial use. It increased awareness of and concern about AMR, supporting GP confidence to appropriately prescribe antibiotics. Future determination of measurable behaviour changes resulting from AMR campaigns is important.

Escherichia coli is the leading cause of urinary tract infections (UTI) and bloodstream infections (BSI), and the emergence of antimicrobial resistance (AMR) in E. coli causes concern.

To investigate changes in the proportion of extended-spectrum β-lactamase (ESBL) producing isolates among E. coli isolated from urine and blood in Finland during 2008–2019.

Susceptibility testing of 1,568,488 urine (90% female, 10% male) and 47,927 blood E. coli isolates (61% female, 39% male) from all Finnish clinical microbiology laboratories during 2008–2019 was performed according to guidelines from the Clinical and Laboratory Standard Institute during 2008–2010 and the European Committee on Antimicrobial Susceptibility Testing during 2011–2019. A binomial regression model with log link compared observed trends over time and by age group and sex.

The annual proportion of ESBL-producing E. coli isolates among E. coli from blood cultures increased from 2.4% (23/966) to 8.6% (190/2,197) among males (average annual increase 7.7%; 95% CI: 4.4–11.0%, p < 0.01) and from 1.6% (28/1,806) to 6.4% (207/3,218) among females (9.3%; 95% CI: 4.8–14.0%, p < 0.01). In urine cultures, the proportion of ESBL-producing E. coli isolates increased from 2.2% (239/10,806) to 7.2% (1,098/15,297) among males (8.8%; 95% CI: 6.5–11.3%, p < 0.01) and from 1.0% (1,045/108,390) to 3.1% (3,717/120,671) among females (8.6%; 95% CI: 6.3–11.0%, p < 0.01). A significant increase was observed within most age groups.

Considering the ageing population and their risk of E. coli BSI and UTI, the increase in the annual proportions of ESBL-producing E. coli is concerning, and these increasing trends should be carefully monitored.

In China, the blaNDM gene has been recovered from human bacterial isolates since 2011. After 2014, detections of this gene in animal and food bacterial isolates have increasingly been reported.

We aimed to understand how blaNDM-bearing bacteria could spread between humans, animals, and animal-derived food.

A total of 288 non-duplicate Escherichia coli strains, including 130 blaNDM-carrying and 158 blaNDM-negative strains were collected from clinical (humans), food-producing animals (pigs) and food (retail pork) sources between 2015 and 2017. The strains were whole genome sequenced. Core-genome-multilocus-sequence-typing was conducted. To investigate if sequence types (STs) found in human, animal or food samples could have a prior origin in a clinical, animal or food-borne animal reservoir, discriminant analysis of principal components (DAPC) was used. Plasmids bearing blaNDM were characterised.

The 130 blaNDM-carrying E. coli strains comprised a total of 60 STs, with ST167 (10/51), ST77 (6/33) and ST48 (6/46) being most prevalent in clinical, animal and food sources, respectively. Some ST10 and ST167 strains were respectively found among all three sources sampled, suggesting they might enable transfer of blaNDM between sources. DAPC analysis indicated possible transmissions of ST167 from humans to animals and ST10 from animals to human. In 114 of 130 blaNDM-carrying isolates, blaNDM was located on an IncX3 plasmid.

This study in a Chinese context suggests that cross-species transmission of certain STs of E. coli harbouring blaNDM on mobile elements, may facilitate the spread of carbapenem-resistant Enterobacteriaceae. Stringent monitoring of blaNDM-bearing E. coli in ecosystems is important.

National and regional carbapenemase-producing Enterobacterales (CPE) surveillance is essential to understand the burden of antimicrobial resistance, elucidate outbreaks, and develop infection-control or antimicrobial-treatment recommendations.

This study aimed to describe CPE and their epidemiology in Norway from 2015 to 2021.

A nationwide, population-based observational study of all verified clinical and carriage CPE isolates submitted to the national reference laboratory was conducted. Isolates were characterised by antimicrobial susceptibility testing, whole genome sequencing (WGS) and basic metadata. Annual CPE incidences were also estimated.

A total of 389 CPE isolates were identified from 332 patients of 63 years median age (range: 0–98). These corresponded to 341 cases, 184 (54%) being male. Between 2015 and 2021, the annual incidence of CPE cases increased from 0.6 to 1.1 per 100,000 person-years. For CPE-isolates with available data on colonisation/infection, 58% (226/389) were associated with colonisation and 38% (149/389) with clinical infections. WGS revealed a predominance of OXA-48-like (51%; 198/389) and NDM (34%; 134/389) carbapenemases in a diversified population of Escherichia coli and Klebsiella pneumoniae, including high-risk clones also detected globally. Most CPE isolates were travel-related (63%; 245/389). Although local outbreaks and healthcare-associated transmission occurred, no interregional spread was detected. Nevertheless, 18% (70/389) of isolates not directly related to import points towards potentially unidentified transmission routes. A decline in travel-associated cases was observed during the COVID-19 pandemic.

The close-to-doubling of CPE case incidence between 2015 and 2021 was associated with foreign travel and genomic diversity. To limit further transmission and outbreaks, continued screening and monitoring is essential.

Cassini et al. (2019) estimated that, in 2015, infections with 16 different antibiotic-resistant bacteria resulted in ca 170 disability-adjusted life-years (DALYs) per 100,000 population in the European Union and European Economic area (EU/EEA). The corresponding estimate for Switzerland was about half of this (87.8 DALYs per 100,000 population) but still higher than that of several EU/EEA countries (e.g. neighbouring Austria (77.2)).

In this study, the burden caused by the same infections due to antibiotic-resistant bacteria (‘AMR burden’) in Switzerland from 2010 to 2019 was estimated and the effect of the factors ‘linguistic region’ and ‘hospital type’ on this estimate was examined.

Number of infections, DALYs and deaths were estimated according to Cassini et al. (2019) whereas separate models were built for each linguistic region/hospital type combination.

DALYs increased significantly from 3,995 (95% uncertainty interval (UI): 3;327–4,805) in 2010 to 6,805 (95% UI: 5,820–7,949) in 2019. Linguistic region and hospital type stratifications significantly affected the absolute values and the slope of the total AMR burden estimates. DALYs per population were higher in the Latin part of Switzerland (98 DALYs per 100,000 population; 95% UI: 83–115) compared with the German part (57 DALYs per 100,000 population; 95% UI: 49–66) and in university hospitals (165 DALYs per 100,000 hospitalisation days; 95% UI: 140–194) compared with non-university hospitals (62 DALYs per 100,000 hospitalisation days; 95% UI: 53–72).

The AMR burden estimate in Switzerland has increased significantly between 2010 and 2019. Considerable differences depending on the linguistic region and the hospital type were identified – a finding which affects the nationwide burden estimation.

Carbapenemase-producing Enterobacterales (CPE) are rapidly increasing worldwide, also in Europe. Although prevalence of CPE in Germany is comparatively low, the National Reference Centre for Multidrug-resistant Gram-negative Bacteria noted annually increasing numbers of NDM-5-producing Escherichia coli isolates.

As part of our ongoing surveillance programme, we characterised NDM-5-producing E. coli isolates received between 2013 and 2019 using whole genome sequencing (WGS).

From 329 identified NDM-5-producing E. coli, 224 isolates from known geographical locations were subjected to Illumina WGS. Analyses of 222 sequenced isolates included multilocus sequence typing (MLST), core genome (cg)MLST and single-nucleotide polymorphism (SNP)-based analyses.

Results of cgMLST revealed genetically distinct clusters for many of the 43 detected sequence types (ST), of which ST167, ST410, ST405 and ST361 predominated. The SNP-based phylogenetic analyses combined with geographical information identified sporadic cases of nosocomial transmission on a small spatial scale. However, we identified large clusters corresponding to clonal dissemination of ST167, ST410, ST405 and ST361 strains in consecutive years in different regions in Germany.

Occurrence of NDM-5-producing E. coli rose in Germany, which was to a large extent due to the increased prevalence of isolates belonging to the international high-risk clones ST167, ST410, ST405 and ST361. Of particular concern is the supra-regional dissemination of these epidemic clones. Available information suggest community spread of NDM-5-producing E. coli in Germany, highlighting the importance of epidemiological investigation and an integrated surveillance system in the One Health framework.

The emergence of colistin resistance is a One Health antimicrobial resistance challenge worldwide. The close contact between companion animals and humans creates opportunities for transmission and dissemination of colistin-resistant bacteria.

To detect potential animal reservoirs of colistin-resistant Escherichia coli and investigate the possible sharing of these bacteria between dogs, cats and their cohabiting humans in the community in Lisbon, Portugal.

A prospective longitudinal study was performed from 2018 to 2020. Faecal samples from dogs and cats either healthy or diagnosed with a skin and soft tissue or urinary tract infection, and their cohabiting humans were screened for the presence of colistin-resistant E. coli. All isolates were tested by broth microdilution against colistin and 12 other antimicrobials. Colistin-resistant isolates were screened for 30 resistance genes, including plasmid-mediated colistin resistance genes (mcr-1 to mcr-9), and typed by multilocus sequence typing. Genetic relatedness between animal and human isolates was analysed by whole genome sequencing.

Colistin-resistant E. coli strains harbouring the mcr-1 gene were recovered from faecal samples of companion animals (8/102; 7.8%) and humans (4/125; 3.2%). No difference between control and infection group was detected. Indistinguishable multidrug-resistant E. coli ST744 strains harbouring the mcr-1 gene were found in humans and their dogs in two households.

The identification of identical E. coli strains containing the plasmid-mediated mcr-1 gene in companion animals and humans in daily close contact is of concern. These results demonstrate the importance of the animal–human unit as possible disseminators of clinically important resistance genes in the community setting.