Fecal swabs were collected from 435 patients in the ICU wards of a tertiary hospital in Shanghai, China, from March 1, 2022, to February 28, 2023, and the above-mentioned samples were inoculated in a Resistant Bacteria Chromogenic Plate (Antu Bio, China). At admission to the ICU wards, fecal swabs were performed by a trained nurse, and the swabs were immediately sent to the clinical laboratory. Meanwhile, to monitor changes in CRE colonization, fecal swabs were repeated twice a week during hospitalization. Each sample was accompanied by a case report form containing the clinical characteristics of the patients, including name, sex, age, admission time, temperature, diagnosis, antibiotic used and laboratory test results. Therefore, active screening of CRE-FC was performed. The meropenem-resistant colonies were subsequently selected, and bacterial identification was performed via MALDI-TOF mass spectrometry (Bruker Daltonics, Billerica, MA, Germany), and antimicrobial susceptibility testing was carried out using a VITEK-2 compact automated microbiology analyser (Biomerieux, Marcy L‘Etoile, France). We removed duplicate strains from the same patient and retained only the first isolated CRE-FC. By quality control, strains were excluded due to low-quality assembly as defined by NCBI (e.g., excessive frameshifted proteins and fragmented assembly), <95% completeness, >5% contamination, or genomes belonging to species other than CRE. A total of 55 isolates of fecal colonization with CRE were identified. Two strains of CRKP, WYZHKP101 and WYZHKP102, were isolated from the ICU environment for comparison with CRKP-FC. The CRE strains were used to test for the presence of the carbapenemase genes bla_KPC, bla_NDM, bla_VIM, bla_IMP, and bla_OXA-48 via PCR with specific primers (bla_KPC: forward: ATGTCACTGTATCGCCGTCT, reverse: TTTTCAGAGCCTTACTGCCC; bla_NDM: forward: GGTTTGGCGATCTGGTTTTC, reverse: CGGAATGGCTCATCACGATC; bla_VIM: forward: GATGGTGTTTGGTCGCATA, reverse: CGAATGCGCAGCACCAG; bla_IMP: forward: GGAATAGAGTGGCTTAAYTCTC, reverse: GGTTTAAYAAAACAACCACC; bla_OXA-48: forward: GCGTGGTTAAGGATGAACAC, reverse: CATCAAGTTCAACCCAACCG).

Antimicrobial susceptibility testing was performed via the broth microdilution method of the Clinical and Laboratory Standards Institute (CLSI). The results were interpreted according to 2022 CLSI breakpoints (23) for all antimicrobial agents except tigecycline, which was interpreted via the interpretation standards of the US Food and Drug Administration (FDA). E. coli ATCC 25922 and K. pneumoniae ATCC 700603 were used as controls for antimicrobial susceptibility testing.

As the predominant CRE-FC was K. pneumoniae (83.6%, 46/55), it is necessary to further analyze the phenotype and genome of 46 CRKP-FC strains. Two CRKP strains, WYZHKP101 and WYZHKP102, which were isolated from the environment, were also further analyzed for phenotype and genome homology. Finally, 48 CRKP isolates were selected and subjected to whole-genome sequencing (WGS). The genomic DNA of the K. pneumoniae strains was extracted via a commercial DNA extraction kit (Qiagen, Germany) following the manufacturer’s instructions. Fragmented genomic DNA was sequenced via the 150-bp paired-end Illumina HiSeq platform (Illumina, San Diego, CA, United States). De novo assembly was conducted via Unicycler version 0.4.8 (24). Then, the genome sequence was acquired after rectification via pilon software (25). Gene annotation was performed via Prokka version 1.12 (26).

Antimicrobial resistance genes were identified via ResFinder 4.1¹. Multilocus sequence typing (MLST), capsule serotypes and virulence genes were identified according to BIGSdb-Pasteur². Core genome multilocus sequence typing (cgMLST) analysis of strains was performed by comparing the 2,358 conserved genome-wide genes of K. pneumoniae using the software Ridom SeqSphere+ version 9.0.8 (Ridom GmbH, Muenster, Germany), with which a phylogenetic tree was constructed. The distance of ≤15 alleles is defined as a cluster. (27, 28).

We used single-nucleotide polymorphisms (SNPs) to investigate the transmission of K. pneumoniae based on genome sequences mining. The phylogenetic analysis included 48 K. pneumoniae genomes from this study and 185 K. pneumoniae genomes obtained from National Center for Biotechnology Information (NCBI). These 233 genome assemblies were subjected to removing recombination regions using Gubbins³ and thereafter to calling core SNPs with Snippy⁴ using strain WYZHKP18 (accession no. GCA_040117795.1) as the reference. We subsequently visualized and annotated a phylogenetic tree using Interactive Tree Of Life (iTOL)⁵.

In accordance with the World Health Organization (WHO) guidelines (29), IPC interventions were carried out for patients with positive CRE screening results. These interventions included contact isolation, the use of disposable or dedicated patient care equipment, and strict enforcement of hand hygiene by medical staff and wards disinfection. The hand hygiene compliance rate is calculated as the ratio of the number of times hand hygiene is performed correctly to the total number of opportunities for hand hygiene, multiplied by 100% to get a percentage value. It reflects the adherence of medical staff to proper hand hygiene practices, which is essential for preventing the spread of infections. Additionally, the wards were subjected to disinfection to maintain environmental hygiene. The environmental hygiene compliance rate is determined by dividing the number of environmental hygiene parameters that meet the set standards by the total number of parameters tested and expressing it as a percentage. This helps in assessing the cleanliness and safety of the ward environment.

The CRE infection rate refers to the proportion of individuals infected with CRE (carbapenem-resistant Enterobacteriaceae) among the specific population or in a specific environment, usually expressed as a percentage. The calculation formula is: CRE infection rate = (the number of individuals infected with CRE/the total number of individuals examined) × 100%. The CRE infection rate can mirror the infection situation of CRE within particular populations or settings, and contribute to assessing the evaluation metrics for the effectiveness of CRE positive screening and IPC interventions. The CRE infection rates in ICU wards were compared between the periods without active CRE screening and IPC interventions (from March 1, 2022, to February 28, 2023) and the periods with active CRE screening and IPC interventions (from March 1, 2022, to February 28, 2023).

The CRE was defined as carbapenem-resistant Enterobacterales including CRKP, carbapenem-resistant E. coli (CRECO) and other bacteria were resistant to at least one of the carbapenems, including imipenem, meropenem, and ertapenem.

The data are presented as the means ± standard deviations (SDs). All the statistical analyses were performed via SPSS version 23.0 (IBM Co., Armonk, NY, United States). Categorical variables were assessed by the chi-square test or Fisher’s exact test. All the statistical tests were 2 tailed, and p < 0.05 was considered statistically significant.

During the study period, we actively screened fecal swabs from 435 patients in ICU wards (Figure 1). The prevalence of CRE-FC in ICU wards was 12.6% (55/435). The predominant CRE-FC was K. pneumoniae (83.6%, 46/55), followed by E. coli (9.1%, 5/55), Klebsiella aerogenes (3.7%, 2/55), Enterobacter cloacae (1.8%, 1/55), and Citrobacter freundii (1.8%, 1/55). We removed duplicate strains from the same patient and retained only the first isolated CRE-FC. The 55 CRE-FC strains were isolated from 55 individual patients (35 males and 20 females), whose mean age was 70.6 ± 13.5 years. Two strains of CRKP, WYZHKP101 and WYZHKP102, were isolated from the ICU environment for comparison with CRKP-FC. As the predominant CRE-FC was K. pneumoniae (83.6%, 46/55), further analysis of the phenotype and genome of 46 CRKP-FC strains, WYZHKP101 and WYZHKP102, is necessary.

The distribution of carbapenemase genes of 55 CRE-FC strains is shown in Table 1. As shown in Table 1, the most prevalent carbapenemase gene was bla_KPC (89.1%, 49/55), which was detected in 45 K. pneumoniae isolates, 1 E. coli, 2 Klebsiella aerogenes and 1 C. freundii, followed by bla_NDM (10.9%, 6/55), which was detected in 4 E. coli isolates, 1 K. pneumoniae isolate and 1 E. cloacae isolate; bla_IMP, bla_VIM, and bla_OXA-48 were not detected (Table 1).

The data presented in Table 2 reveals that in comparison to 2021, the timely isolation rate of CRE-positive patients, the hand hygiene compliance rate, and the qualified rate of environmental disinfection all showed improvements. When it comes to resistance rates, variations were observed among different types of CRE. Specifically, the carbapenem resistance rate of K. pneumoniae in 2022 (39.3%) was significantly lower than that in 2021 (55.9%), with a statistically significant difference. In contrast, the carbapenem resistance rate of E. coli remained relatively stable, being 2.9% in 2022 and 2.6% in 2021, and the difference was not statistically significant. Additionally, the CRE infection rate decreased from 11.4 to 7.1%. Overall, the data indicates that active screening of CRE-FC and the implementation of IPC interventions can effectively reduce the CRE infection rate in ICU wards.

To clarify the antibiotic resistance profile of 46 CRKP strains colonized via feces, we tested their susceptibility to 18 antibiotics (Table 3). All the CRKP strains presented multiple drug resistance phenotypes and were resistant to three or more antibiotic classes. The 46 CRKP-FC strains were resistant to ampicillin, ampicillin/sulbactam, cefoperazone/sulbactam, piperacillin/tazobactam, cefazolin, cefuroxime, ceftazidime, ceftriaxone, cefepime, cefoxitin, imipenem, and meropenem. The antimicrobial resistance rate of ceftazidime/avibactam was the lowest (6.5%), followed by tigecycline (8.7%), trimethoprim-sulfamethoxazole (73.9%), amikacin (76.1%), gentamicin (93.5%) and levofloxacin (97.8%).

A total of 46 CRKP-FC strains were subsequently subjected to whole-genome sequencing. Two strains of CRKP, WYZHKP101 and WYZHKP102, were isolated from the ICU for comparison with 46 strains of CRKP-FC, and whole-genome sequencing was performed. MLST analysis revealed that 46 CRKP-FC strains belonged to 3 different STs, with ST11 being the predominant ST (71.7%, 33/46), followed by ST15 (26.1%, 12/46) and ST290 (2.2%, 1/46) (Figure 2). Capsular serotypes revealed four different K loci; KL64 was the predominant capsule type (54.3%, 25/46), followed by KL19 (26.1%, 12/46), KL47 (4.3%, 2/46), and KL21 (2.2%, 1/46), and six strains could not be typed (Figure 2). These findings suggest that ST11-KL64 CRKP has emerged as the most prevalent CRKP-FC and may contribute to hospital outbreaks of infection.

The presence of carbapenemase genes is concerning, as they can cause resistance to many types of antibiotics, including carbapenems. All the ST11 and ST15 strains harbored bla_KPC-2, and 10 ST15 strains carried two carbapenemase genes (bla_KPC-2, bla_OXA-1) at the same time (Figure 2). One ST290 strain carried two carbapenemase genes (bla_NDM-1 and bla_OXA-9) (Figure 2). Compared with ST11 strains, ST15 strains carried multiple carbapenemase genes, which further increased antibiotic resistance.

Almost all the ST11 strains carried the virulence genes iucABCD-iutA and rmpA2, whereas the ST15 strains lacked these virulence genes (Figure 2). ST11 strains carry more virulence genes than ST15 strains do, which is the opposite of resistance. This result suggests that strains with more resistance may have less virulence potential.

WYZHKP101 and WYZHKP102, which were isolated from the ICU environment, belong to the ST11-KL64 clone. The antibiotic resistance gene profile and virulence gene profile of these strains were similar to those of ST11-KL64 CRKP-FC (Figure 2). These findings suggest that the ST11-KL64 CRKP strains may cause nosocomial infection outbreaks.

To gain a deeper understanding of the homology of these strains, we conducted whole-genome sequencing and core-genome phylogenetic analysis (Figure 3). The phylogenetic tree identified two major clades, each belonging to the same ST; specifically, cluster 1 corresponded to ST11, and cluster 2 corresponded to ST15 (Figures 2, 3).

The results revealed that the colonized ST11-KL64 CRKP strains dominated and formed a single cluster (Cluster 1) in ICU wards, indicating clonal expansion. These ST11 strains contained mainly KL64 capsular serotypes and a few other capsular serotypes (Figure 2). WYZHKP101 and WYZHKP102, which were isolated from the ICU environment, were ST11-KL64 clones belonging to Cluster 1. These findings suggest that ST11-KL64 CRKP has emerged as the most prevalent colonized carbapenem-resistant K. pneumoniae and may contribute to the outbreak of nosocomial infection. ST15 K. pneumoniae was identified as the second most prevalent colonized CRKP clone in ICU wards after ST11 K. pneumoniae. These ST15 strains contained only one capsule serotypes, KL19. These findings suggest that the ST15-KL19 CRKP strains formed a single cluster (Cluster 2), which may also have contributed to the outbreak of nosocomial infection.

To investigate the transmission of ST11-KL64 and ST15-KL19 K. pneumoniae, we conducted single-nucleotide polymorphisms (SNPs) based phylogenetic tree. The phylogenetic analysis included 48 K. pneumoniae genomes from this study and 185 K. pneumoniae genomes (ST11-KL64 and ST15-KL19) obtained from NCBI (Supplementary Table S1). A total of 185 K. pneumoniae (ST11-KL64 and ST15-KL19) from different countries and different isolation times were selected as reference strains.

As shown in Figure 4, the resulting phylogenetic tree had two main branches. ST11-KL64 and ST15-KL19 K. pneumoniae occupied two main branches, respectively. Clusters of SNPs based phylogenetic trees were similar to those of Minimum-spanning trees of cgMLST. Although in the same main branches, K. pneumoniae with the same MLST have obvious genetic distance if they come from different countries or have different isolation times. For the 48 strains in this study, K. pneumoniae with the same MLST had closer genetic distance if the strains were isolated in the same year.