Carbapenem-resistant Enterobacterales (CRE) are recognized as among the most critical priority antimicrobial resistant pathogens and represent a major global threat to human health and antimicrobial resistance. Their emergence is particularly concerning in inpatient settings, where vulnerable patients and high antimicrobial use facilitate selection and transmission, underscoring the need for robust infection prevention and control (IPC) strategies (World Health, O, 2017). CRE prevalence varies considerably across healthcare environments, ranging from 1% to 30.4% in long-term care facilities in the United States and from 13% to 22.7% across Asian hospitals (Chen et al., 2021). In a neonatal intensive care unit (NICU) in Shanghai, China, the CRE colonization rate reached 11.7% during 2017–2018 (Yin et al., 2021).

In Vietnam, carbapenems have been used clinically since 2008 (Nguyen et al., 2013), and increasing reports of carbapenem-resistant organisms—including strains producing NDM-1—reflect the growing burden of resistance nationwide (Hoang et al., 2013; Tran et al., 2015). A multicenter survey involving 12 hospitals showed that 52% of screened patients were colonized with CRE, with prevalence rising from 13% at admission to 89% by hospital day 15; in one provincial pediatric hospital, carriage increased from 29% to 53% over a 16-month period (Tran et al., 2019). These data highlight both high background colonization pressure and rapid in-hospital acquisition.

A national survey of Vietnamese intensive care units reported HAIs in 30% of approximately 5,000 admitted patients (Le et al., 2016), and mortality among neonates with culture-confirmed HAIs caused by carbapenem-resistant Gram-negative bacteria exceeded 50% (Peters et al., 2019). CRE colonization is a prerequisite to subsequent invasive infection. Clearance of colonization is protracted, with median persistence of roughly one year, and prolonged carriage was associated with therapeutic antimicrobial exposure (Kramer et al., 2013).

High CRE colonization rates at discharge raise concerns about onward dissemination to household members, the community, and the environment. Such transmission may further complicate the management of community-onset infections, including urinary tract infections that may fail standard empirical therapy, even when treated with intravenous agents such as amikacin. Evidence from a longitudinal study of extended-spectrum β-lactamase (ESBL)-producing Enterobacterales demonstrated high within-household transmission probabilities - 67% from index patient to household contact and 37% in the reverse direction - and prolonged carriage associated with antibiotic exposure (Haverkate et al., 2017). Although focused on ESBL organisms with resistance to third generation cephalosporins, these findings underscore the potential for similar dynamics for CRE.

Early and accurate detection of CRE colonization is critical for IPC, outbreak containment, and guiding clinical management, particularly in high-risk units such as NICUs and ICUs (Zaidah et al., 2017). This study aimed to determine the prevalence and risk factors for CRE colonization among children hospitalized in a provincial pediatric hospital in Vietnam. The findings are intended to inform the development and implementation of targeted strategies to mitigate CRE transmission within healthcare facilities and reduce dissemination to the community.

Among the 376 hospitalized children included in the survey, the overall prevalence of CRE colonization was 28.2% (Table 1). Children ≤6 years accounted for most of the study population (344/376, 91.5%) and had a higher prevalence of colonization (29.7%) compared with those >6 years (12.5%). Similarly, children ≤3 years exhibited a CRE prevalence of 30.3%. CRE colonization did not differ markedly by sex, with colonization observed in 28.9% of males and 26.7% of females.

Most patients had been admitted directly from home (314/376), and this group showed a colonization prevalence of 29.9%. In contrast, children transferred from other hospitals or lower-level medical facilities had smaller sample sizes but also demonstrated notable colonization rates. (please see Table 1).

CRE colonization was 17.1% of among children screened within 48 hours of admission, and 30.8% after 48 hours. The highest prevalence was observed in patients with stays ≥7 days, 40.8% of whom were colonized. In the neonatal subgroup, CRE prevalence reached 38.9%. Overall, these findings indicate substantial colonization pressure across age groups and admission pathways, with markedly increased prevalence associated with prolonged hospitalization.

The prevalence of CRE colonization varied markedly across hospital departments (Table 2). The highest prevalence was observed in the Intensive Care Unit, where 77.8% of patients were colonized. High rates of colonization rates were also seen in the 3 Specialities Department (43.7%), the Neonatal Department (38.9%), and the Respiratory and Infectious Diseases Departments (both 34.4%). In contrast, substantially lower rates were found in the Trauma (9.5%), Nutrition (10.0%), and Emergency (13.3%) Departments. Intermediate prevalence levels were reported in Gastroenterology (24.1%), Cardiology (23.5%), and Nephrology–Urology (16.3%). These findings indicate substantial differences in CRE burden across clinical areas, with intensive care and neonatal services representing the highest-risk settings.

A total of 128 CRE isolates were recovered from 106 colonized patients, corresponding to an average of 1.2 isolates per patient (Table 3). Escherichia coli was the most frequently identified species (40.6%), followed by K. pneumoniae (34.4%) and E. cloacae (16.4%). Species distribution varied across departments. In the ICU, E. coli and K. pneumoniae predominated, together accounting for 75% of isolates, while Pseudomonas spp. constituted an additional 15%. In the Neonatal Department, K. pneumoniae and E. cloacae were equally common (both 35.3%). Respiratory wards showed a similar dominance of E. coli and K. pneumoniae (each 42.9%). Less commonly isolated species included K. oxytoca, Citrobacter freundii, Aeromonas sobria, Acinetobacter spp., and Brevundimonas diminuta. These findings indicate that the CRE reservoir in the hospital is primarily composed of E. coli and K. pneumoniae, with notable departmental variation suggesting multiple ecological niches for CRE persistence and transmission.

Antimicrobial susceptibility patterns showed uniformly high resistance across CRE species (Table 4). All E. coli, K. pneumoniae, and E. cloacae isolates demonstrated 100% resistance to carbapenems and near-panresistance to 1st- and 2nd-generation cephalosporins, 3rd- and 4th-generation cephalosporins, cephamycins, and piperacillin–tazobactam. Fluoroquinolone resistance was also high, ranging from 90.1% in E. coli to 97.7% in K. pneumoniae. Aminoglycoside resistance was common, particularly among K. pneumoniae (84.1%). Fosfomycin resistance varied by species, from low levels in E. coli (3.8%) to high levels in E. cloacae (71.4%). Multidrug resistance (MDR), defined as combined resistance to 3rd-generation cephalosporins, fluoroquinolones, and aminoglycosides, was most prevalent in K. pneumoniae (79.5%) and E. cloacae (76.2%). Overall, the isolates exhibited extensive resistance across nearly all tested antimicrobial classes, leaving very limited therapeutic options.

Multivariable analysis identified several significant risk factors for CRE colonization (Table 5). Children screened more than 48 hours after admission had more than twice the odds of being colonized compared with those screened within 48 hours (aOR 2.27, 95% CI 1.10–4.67, p = 0.026), indicating high in-hospital acquisition. Age ≤6 years was associated with higher colonization in crude analysis (cOR 2.93, p = 0.049), although this association did not remain statistically significant after adjustment. Sepsis at the time of the survey was strongly associated with CRE colonization, with affected children having more than sixfold increased odds to be CRE colonized compared with those with other diagnoses (aOR 6.54, 95% CI 1.55–27.64, p = 0.011). Unfortunately, data obtained did not allow us to determine whether patients had sepsis due to CRE or were septic due to other causes and colonized with CRE. Other diagnostic categories, including respiratory infections and diarrhea, did not show significant associations to CRE colonization. Overall, prolonged hospitalization and severe systemic illness emerged as the strongest predictors of CRE colonization in this cohort.

In this point prevalence survey (PPS), we found a CRE colonization rate of 28.2% among hospitalized children in a provincial pediatric hospital. Colonization was particularly common in high-risk units, with prevalence of 77.8% in the ICU, 43.7% in the 3 Specialities Department, 38.9% in the Neonatal Department, and 34.4% in the Respiratory Department, whereas the lowest rates were observed in the Nutrition (10%) and Trauma (9.5%) Departments. These findings are comparable to, and in some instances higher than, earlier reports from Viet Nam and other settings. At the Vietnam National Children’s Hospital, a cross-sectional survey conducted from March to June 2018 reported a CRE carriage rate of 35% among ICU inpatients (Garpvall et al., 2021). Most of those patients had previously been treated at lower-level healthcare facilities, including provincial pediatric or obstetric-paediatric hospitals (Garpvall et al., 2021). In an adult cohort from the Intensive Care–Toxicology Unit of Thai Binh Provincial General Hospital, the CRE carriage rate at admission was 41% (McConville et al., 2017). In contrast, rectal screening on admission to adult medical or surgical ICUs in the United States found a CRE colonization rate of 11% (Armin et al., 2023). At a university hospital in Thailand, the prevalence of CRE colonization at admission was 15.5% in 2022, and a cross-sectional study from Iran reported a CRE colonization prevalence of 37% among hospitalized children (van Loon et al., 2018). Taken together, our data add to the evidence that CRE carriage in healthcare facilities in low- and middle-income countries remains high and underline the urgent need for strengthened and targeted interventions.

Most confirmed CRE cases in our study (94/106; 88.7%) were detected more than 48 hours after admission, and hospitalization beyond 48 hours was associated with a 2.27-fold higher risk of CRE colonization compared with screening within 48 hours (p = 0.026). Thus, in-hospital acquisition was a major driver of CRE colonization in this setting. Length of hospital stay is a well-recognized risk factor for CRE colonization and infection, as patients are increasingly exposed to transmission via colonized or infected patients, contaminated environmental surfaces, and medical equipment (van Loon et al., 2018; Tran et al., 2019; Yen et al., 2023). The risk is amplified when hospitalization is prolonged because of severe disease or the need for invasive procedures—such as mechanical ventilation, central venous catheters, urinary catheters, or surgery—particularly in ICUs and neonatal units (van Loon et al., 2018; Tran et al., 2019; Yen et al., 2023). The high colonization pressure observed in our ICU and Neonatal Department is consistent with this pattern and reflects a combination of patient vulnerability and a challenging transmission environment.

In our study, most CRE-positive patients were admitted directly from home (travel by oneself) and were distributed across all districts and the provincial city (Table 1; Appendix 1). Our data from admission screening showed significant CRE colonization which increased during the hospital stay and which may further contribute to spread within the community after discharge. It also raises concern for increasing rates of community-onset infections—such as urinary tract caused by CRE, with consequent difficulties in selecting appropriate empirical antibiotics.

In our study, most CRE-positive patients were admitted directly from home and resided across all districts and the provincial city (Table 1; Appendix 1). High CRE prevalence was also observed among children diagnosed with sepsis at the time of the survey, and sepsis was associated with substantially higher odds of CRE colonization in multivariable analysis (adjusted OR = 6.54, 95% CI 1.55–27.64, p = 0.011). This distribution suggests that CRE colonization is not confined to a few geographic “hot spots” and raises concern both for household transmission after discharge and for spread to other healthcare facilities upon referral or readmission. It also points to a potential increase in community-onset infections, such as urinary tract caused by CRE, which may be difficult to treat empirically with standard regimens.

In this PPS, all identified CRE species showed high levels of resistance to multiple antibiotic classes (Appendix 2). Escherichia coli was the most common species (40.6%), followed by Klebsiella pneumoniae (34.4%) and Enterobacter cloacae (16.4%) (Table 3). The resistance profiles showed very high rates of non-susceptibility to carbapenems, extended-spectrum cephalosporins, fluoroquinolones, and β-lactam/β-lactamase inhibitor combinations (Table 4). CRE colonization is a recognized prerequisite to healthcare-associated infections (HAIs) caused by CRE and therefore represents a latent clinical risk: if infection occurs, therapeutic options are limited and the probability of treatment failure increases (Kramer et al., 2013; Tischendorf et al., 2016) These findings emphasize the importance of identifying colonized patients early and preventing progression to infection whenever possible.

To date, Viet Nam still lacks comprehensive national data on CRE carriage, particularly due with respect to low screening capacity and systematic sharing of data on CRE colonization status between facilities. Notification of CRE carriage at inter-facility transfer is not uniformly implemented across the country, which hampers coordinated infection prevention and control (IPC) measures. Early and accurate detection of CRE colonization is essential for identifying outbreaks and guiding IPC and treatment strategies (Zaidah et al., 2017). The cost of CRE screening depends on consumables, reagents, staff time, and local resource constraints and therefore differs between high-income and low- and middle-income countries (World Health Organization, 2019). For example, the average cost of a single CRE screening test was estimated at $8.65 in the United States (Lin et al., 2022). In contrast, in our study the material cost for one CRE screening plate was approximately $2, which could be used for two patients, thus a laboratory cost of about $1 per patient. Nonetheless, CRE screening is currently not reimbursed by health insurance and is not routinely performed in Vietnamese healthcare facilities, limiting its uptake despite its relatively low direct cost in this setting.

CRE screening and cohort care are widely recognized as key components of effective CRE control programs. At Vietnam National Children’s Hospital, admission screening of more than 900 children to three ICUs, followed by cohorting according to CRE status, reduced CRE acquisition from 90% to 48% (Garpvall et al., 2021). In Israel, a nationwide intervention combining CRE screening with cohort and isolation care significantly reduced CRE transmission (Schwaber and Carmeli, 2014). Our findings, together with these previous results, support the implementation of targeted screening strategies and cohort care in high-risk pediatric departments, particularly ICUs and neonatal units, as part of a broader IPC package.

This study has several limitations. First, it is based on a single PPS conducted in one provincial pediatric hospital, and thus not representative for Viet Nam. Nevertheless, the PPS provides a detailed snapshot of the wards studied and it may be relevant for similar provincial pediatric hospitals. Second, genotyping data are not presented here; however, a companion publication using the same isolate collection reported carbapenemase genes in 63.8% of isolates, with NDM being the predominant carbapenemase in 86.7% of CRE isolates (Pham et al., 2024). For the other 36.2% of isolates please see pham et al (Pham et al., 2024). Third, due to low reported HAI incidence and limited routine culture before starting antibiotics, we were unable to systematically link colonizing and clinical isolates and therefore could not quantify the progression from colonization to infection. Previous studies, however, have shown a strong association between CRE colonization and subsequent CRE HAI (Tischendorf et al., 2016). Finally, as a PPS, this study captures a single time point and does not account for temporal fluctuations in colonization pressure or the incidence of new acquisitions over time.

In summary, we report a high prevalence of CRE colonization among pediatric inpatients in a Vietnamese provincial hospital, with evidence of substantial in-hospital transmission and a predominance of multidrug-resistant E. coli, K. pneumoniae, and E. cloacae. The particularly high colonization rates in ICU and neonatal wards, and among children with prolonged hospital stay or sepsis, highlight priority targets for IPC interventions, CRE screening, and cohort care. These findings form an important baseline for the ongoing randomized controlled trial on interventions to reduce CRE colonization and transmission in hospitals, households, communities, and domesticated animals, and underscore the urgent need to strengthen CRE control strategies in pediatric healthcare settings in Vietnam.