Vital Surveillances: Surveillance of Multidrug-Resistant Bacterial Infections in Non-Adult Patients — Zhejiang Province, China, 2014–2019

Clinical data were obtained from the Annual Review of Hospital Infection Resistance Survey in Zhejiang Province, 2014–2019. Hospitals that participated in the study were distributed across 11 cities in Zhejiang Province: Hangzhou, Jiaxing, Huzhou, Shaoxing, Ningbo, Zhoushan, Taizhou, Jinhua, Quzhou, Lishui, and Wenzhou. Hospitals in China are classified into 3 categories (primary, secondary, and tertiary institutions) based on their medical service capacity. All the hospitals in the study were secondary or tertiary hospitals accredited to perform pathogen identification and anti-microbial susceptibility testing (Supplementary Table S1). The prevalence of CRE, ESBL-E, CRAB, CRPA, and MRSA isolates were determined by analyzing data exported from WHONET software (version 5.6, WHO) with SPSS software (version 23.0, SPSS Inc., Chicago, IL, USA). In group comparisons, Pearson’s chi-square and Fisher’s exact tests were used. In all models, there was statistical significance with P<0.05.

A total of 212,252 non-duplicate strains collected from 2014 to 2019 were analyzed in this study. Among them, 30,163 strains were found to be multidrug-resistant. These included 15,758 ESBL-producing strains of the Enterobacteriaceae family (ESBL-E, accounting for 40.6% of Enterobacteriaceae strains), 1,349 CRE (2.3% of Enterobacteriaceae), 881 CRAB (14.7% of Acinetobacter baumannii), 507 CRPA (9.0% of Pseudomonas aeruginosa), and 11,668 MRSA (27.4% of Staphylococcus aureus). MRSA and ESBL-E were the most common pathogens, accounting for 90.9% of all drug-resistant infections (52.2% for ESBL-E and 38.7% for MRSA infections). Sample characteristics were provided in the Supplementary Table S1.

The prevalences of CRAB, CRE, CRPA, MRSA, and ESBL-E recorded in different years were displayed in Figure 1. The prevalence of CRE decreased from 2.7% (95% CI 2.4%–3.0%) in 2016 to 2.1% (95% CI 1.9%–2.4%) in 2019, and the prevalence of ESBL-E also consistently declined from 42.7% (95% CI 41.2%–44.2%) in 2014 to 39.4% (95% CI 38.2%–40.6%) in 2019. The highest prevalence of CRAB (19.8%, 95% CI 17.6%–22.1%), was recorded in 2015, and a decrease was observed afterwards. It is worth noting that the prevalence of CRPA fluctuated during 2014–2016 and increased significantly from 7.18% (95% CI 5.2%–9.5%) in 2017 to 12.7% (95% CI 10.6%–15.0%) in 2019. MRSA appeared to be another emerging threat. The prevalence of MRSA increased significantly from 24.3% (95% CI 23.1%–25.6%) in 2014 to 29.2% (95% CI 28.2%–30.2%) in 2018 and remained at a high level (27.2%, 95% CI 26.3%–28.2%) after dropping in 2019.

Figure 1. 

The prevalence of CRAB, CRE, CRPA, MRSA, and ESBL-E in non-adult patients — Zhejiang Province, 2014–2019.

Note: The error bars represent 95% CI of the prevalence. Abbreviations: CRAB=carbapenem-resistant Acinetobacter baumannii; CRE=carbapenem-resistant Enterobacteriaceae; CRPA=carbapenem-resistant Pseudomonas aeruginosa; MRSA=methicillin-resistant Staphylococcus aureus; ESBL-E=extended-spectrum β-lactamase-producing Enterobacteriaceae.

The prevalence and risk analysis of CRAB, CRE, CRPA, MRSA, and ESBL-E in ICU and non-ICU groups were described in Table 1. The prevalence and odds ratio (OR) in ICU group were significantly higher than that in non-ICU group. For CRAB, CRE, and CRPA, the prevalence and OR of ICU group were significantly higher than that in non-ICU group (P<0.001). For MRSA, the prevalence was close in ICU and non-ICU groups in 2014 and 2017, and the OR was not statistically significant in these years. For the remaining year, high prevalence and OR of MRSA in ICU group were observed (P<0.05). For ESBL-E, the prevalence in both the non-ICU group and the ICU group showed a high level, 38.0% to 42.6%, respectively. The OR in ICU group was slightly higher than that in non-ICU group.

Prevalence and risk levels of CRAB, CRE, CRPA, MRSA, and ESBL-E in different age groups were shown in Table 2. For CRAB, the age group with the lowest risk was 1 to 5 years old, and the highest risk age group was 15 to 17 years old. For CRE, the lowest risk age group in 2014 was found to be among children less than 1 year old; all the other age groups exhibited significantly higher risk (P=0.014, <0.05). The prevalence of CRPA in different age groups varied like that of CRAB during the study period, with the lowest risk age group being 1 to 5 years old, and the highest risk age group being 15 to 17 years old. For MRSA, the age group <1 year exhibited the highest risk. The age group <1 year also exhibited the highest risk of ESBL-E infection from 2014 onwards, and other age groups showed similar risk level.

Regarding antimicrobial resistance (AMR) in bloodstream infection (BSI) (Table 3), CRAB exhibited no significant difference between BSI and non-BSI consistently for six years. The risk level of CRE in BSI was significantly higher than non-blood samples (except for in 2014 and in 2018, P>0.05). For CRPA, no significant risk was found, but the prevalence decreased from 12.5% (95% CI 1.6%–38.3%) in 2014 to 3.0% (95% CI 0.1%–15.8%) in 2019. The prevalence of MRSA remained stable throughout, and the risk level in blood samples was only significantly higher than that of non-blood samples in 2015 (OR=1.4, 95% CI 1.0–2.1, P=0.037, <0.05). The prevalence of ESBL-E decreased since 2014 and there was no significant change in the risk of ESBL-E in BSI from 2015 to 2019.

The overall prevalences of the five key pathogens were lower than that recorded in the China Antimicrobial Resistance Surveillance System (CARSS) and Infectious Diseases Surveillance of Pediatrics (ISPED). Comparison of results from ISPED 2017–2019 (4-6) indicated that the prevalence of CRE not only remained at a low level (2.0% to 2.7% from our data vs. 8.2% to 10.8% from ISPED), but also exhibited a decreasing trend since 2016 (P=0.004, <0.05). The only exception is carbapenem-resistant Klebsiella pneumonia, being an upward trend observable in ISPED (6). Though the prevalence of CRPA was lower than the data from ISPED (5), the prevalence has risen continuously since 2017. ESBL-E decreased since 2014 but the prevalence remained at a high level. The decrease was also observed in CRAB since 2015 and the prevalence recorded each year was lower than that of CARSS (2). The prevalence of MRSA kept increasing in the first few years (24.3% in 2014 to 29.2% in 2018) but then fell in 2019 (27.2%). The trend and prevalence of MRSA was in general agreement with the report of CARSS (27.5%–29.5%) (2). These data suggested that the prevalence of MRSA and CRPA should be prioritized due to their high prevalence and increasing trends.

Investigating the pattern of pathogens in the ICU environment, especially for MDR organisms, will help develop specific prevention and control strategies. A recent study in a tertiary teaching hospital in western China reported that Acinetobacter baumannii was the leading cause of infection in almost every ICU (7). In this study, ESBL-E was the most prevalent pathogen in ICU non-adult patients in Zhejiang Province (43.4% to 53.8%). High prevalences of CRAB, CRPA, and MRSA were also reported, suggesting a complex ICU environment. Risk analysis identified ICU admission as a risk factor for MDR infection, especially those due to CRAB (OR ranged from 4.4 to 12.0) and CRPA (OR ranged from 4.6 to 9.5). This finding is consistent with previous studies (8-9). Therefore, monitoring the pathogens exposure and incident infections in ICU environment is critical.

Bloodstream infections (BSIs) represent a major cause of mortality and morbidity worldwide. In addition, antimicrobial-resistant organisms, most notably MRSA and ESBL-E, have emerged as the important etiological agents of community-acquired BSI (10). In our study, the prevalence of ESBL-E was the most common pathogen of bloodstream infections, followed by MRSA. The results agreed with the observations from a population-based and large multicenter cohort study in the US and Europe (11). The high prevalence of ESBL-E may be related to inappropriate antibiotic use. One study showed that clinical isolation of ESBL-producing E. coli or ESBL-producing Klebsiella spp. was closely linked to the third-generation cephalosporin treatment (12).

In China, third-generation cephalosporin is the most common antibiotics to treat infections in neonates and older children and therefore may be overused in hospitals (13). MRSA is another critical pathogen associated with significant clinical morbidity and mortality. The prevalence of MRSA in adults is stable in Zhejiang Province and is similar to that in children (from 33.0% in 2014 to 29.8% in 2017) according to the China Antimicrobial Surveillance Network (14). In addition, the MDR pathogens responsible for BSI vary significantly in different regions in China. The predominant pathogen of BSI in our study is ESBL-E, whereas MRSA was the predominant BSI pathogen in Hubei Province (15). Risk analysis indicated that BSI had been a risk factor for CRE infection for many years, but a significant difference was not observed among other bacterial groups.

Surveillance carried out in Zhejiang Province indicated that great attention should be paid to MDR organisms, especially for CRPA and MRSA. Some measures should be taken to alleviate the threat of AMR. On the one hand, for hospital-acquired infections, it is necessary to monitor the ICU environment, where broad-spectrum antibiotic use and the presence of MDR bacteria are common. On the other hand, antimicrobial stewardship programs should be advocated, especially for antibiotic prescription in the community since, in accordance with China Health Care Policy, pediatric patients were referred to community hospitals first, where the misuse and overuse of antibiotics occur frequently. Encouragingly, the government of China has started explorations of AMR. In 2016, National Action Plan for Containing Antibacterial Resistance (2016–2020) was published, aiming at reducing antimicrobial resistance through the synergy between national, regional, and local levels. Surveillance of MDR pathogens in clinical patients is necessary for monitoring AMR.

There were some limitations in this surveillance study. First, due to the differences in medical conditions, data collected from hospitals in rural areas might be lower than the actual value. Second, symptomatic patients were more likely to visit medical institutions compared with asymptomatic ones, which may lead to selection bias. Finally, the lack of available data on antibiotic prescription in the community may influence the analysis of community-sourced infection.

In conclusion, conducting surveillance of multidrug-resistant bacterial infections in non-adult patients to depict the prevalence and variation trends will support better diagnosis and clinical treatment.