Investigating the Distribution and Antibiotic Resistance of Bacterial Pathogens in Clinical Specimens from a Chinese Maternal and Child Hospital: The Role of Environmental Factors

Introduction

Given that bacterial infections are responsible for numerous diseases with high morbidity and mortality rates, they are a significant threat to public health worldwide. These infections can be acquired from various sources, including contaminated water, food, and surfaces, and may also contract from infected individuals or animals.¹ To develop effective prevention and treatment strategies, it is crucial to understand the sources and distribution of the bacterial pathogens. Surveillance programs are used to identify the source and distribution of bacterial pathogens by monitoring the occurrence and spread of bacterial infections in specific locations or populations. These findings provide valuable information on the prevalence and incidence of bacterial infections as well as changes in bacterial strains and patterns of antibiotic resistance over time.²

Specific drug-resistant bacteria, such as methicillin-resistant Staphylococcus aureus (MRSA), carbapenem-resistant Pseudomonas aeruginosa (CR-PAE), carbapenem-resistant Escherichia coli (CR-ECO), carbapenem-resistant Klebsiella pneumoniae (CR-KPN), and carbapenem-resistant Acinetobacter baumannii (CR-ABA), pose a significant threat to public health.³ These bacteria acquire resistance to multiple classes of antimicrobial agents, leaving fewer treatment options for the infected patients. MRSA is a leading cause of healthcare-associated infection.⁴ CR-PAE, CR-ECO, CR-KPN, and CR-ABA are associated with healthcare-associated infections, particularly in immunocompromised patients, and have caused outbreaks in hospitals and long-term care facilities worldwide.⁵

The spread of infectious diseases in landscapes of changing climate cannot inherently be characterized with complete certainty because of the complexity and diversity of microbial populations. Nevertheless, it is well established that increases in environmental temperatures appreciably augment the multiplication of an array of pathogens.⁶ Some studies have shown a correlation between viral etiology in hospitalized children, including meteorological factors (temperature and relative humidity), and air pollutants (PM 10, PM 2.5, and NO₂);^7,8 however, the correlation between these factors and pathogenic bacteria remains unknown. In addition, there is a notable gap in research investigating the correlations between average volume of precipitation, sunlight hours, and ozone (O₃) levels and the prevalence of various pathogens. Understanding the correlations between environmental changes and individual pathogens over time, both globally and regionally, is essential.⁹ These insights are vital for anticipating and mitigating future bacteriological threats. By identifying how environmental factors influence pathogen behavior, we can devise public health strategies that are more effectively tailored to preempt outbreaks and manage disease prevalence. This understanding is particularly crucial for safeguarding vulnerable groups like children and pregnant women.

The purpose of this study was to retrospectively investigate and statistically analyze the bacterial culture samples of inpatients at the Wuhan Children’s Hospital during 2017–2021 and dynamically analyze the distribution, epidemic trends, and drug sensitivity of isolated pathogens; the findings are expected to help prevent bacterial infections and guide the rational use of drugs in clinical practice. Moreover, the correlations between different pathogens and local environmental factors (climate and air pollution) were explored.

Materials and Methods

Study Design

This retrospective study was conducted to analyze the distribution and antibiotic resistance patterns of bacterial pathogens in clinical specimens and to explore the impact of environmental factors on the prevalence of these pathogens. The study spanned across 5 years (January 2017 to December 2021) and included a comprehensive dataset of clinical and environmental metrics. Clinical and environmental data related to climate and air quality were collected concurrently to investigate potential correlations between the incidence of pathogenic bacteria and environmental variables, including air pollutant levels and climatic changes. In addition, clinical and demographic data of patients, such as age, type of specimens, microbiology results, and antibiotic resistance patterns of the isolates, were obtained from the hospital’s laboratory information system.

Sample Collection

Between January 2017 and December 2021, a total of 42,854 specimens were collected from hospitalized children and women. These specimens included blood, sputum, pus and secretions, urine, and other specimens clinically suspected of harboring an infection, and the specimens were sent to a microbiology laboratory for culture and identification. All sample collections strictly adhered to the “Basic Technical Standard for Clinical Microbiology Laboratory” (People’s Republic of China Health Industry Standard WS/T 805—2022) to ensure the quality and consistency of the data. If multiple cultures from the same patient during hospitalization showed the same bacteria, only one result was recorded.

Bacterial Isolates and Antibiotic Susceptibility Tests

All sample inoculation and culturing were performed strictly in accordance with the “Basic Technical Standard for Clinical Microbiology Laboratory” (People’s Republic of China Health Industry Standard WS/T 805—2022). In particular, various clinical specimens were inoculated onto blood agar, MacConkey agar, and chocolate agar (Guangzhou Detgerm Microbiological Science, China) and incubated at 35°C for 18–48 h to isolate the suspected bacterial colonies. Species identification of the isolates was conducted using standard biochemical methods supported by VITEK 2 compact system (BioMérieux SA, France) and Matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF; Bruker, Germany). The antibiotic susceptibilities of clinical isolates were determined using the VITEK 2 compact system following the instrument specifications, complemented by the Kirby-Bauer (KB) method, and the results were interpreted according to the Clinical and Laboratory Standards Institute (CLSI) criteria (M100, 31st Edition).¹⁰ E. coli ATCC 25922, S. aureus ATCC 25923, and P. aeruginosa ATCC 27853 were used as the quality control strains.

Environmental Data

Wuhan is segmented into 13 administrative districts, which encompass a total of 176 environmental monitoring stations. Considering that most of the samples from hospitalized patients were sourced from these districts, we used the averaged data from all monitoring stations as the city-wide environmental indicators (outliers are excluded based on Grubbs’ test). We obtained monthly environmental data, including climate and air pollution metrics, for the Wuhan region for 2017–2021 from two sources: the China National Meteorological Data Sharing System and the Department of Ecology and Environment of Hubei Province. Average temperature (°C), relative humidity (%), average volume of precipitation (mm), and sunlight hours (h) were chosen as climatic variables, and fine particulate matter (PM2.5), inhalable particulate matter (PM10), nitrogen dioxide (NO2), and ozone (O3) were chosen as air pollution variables. Monthly air pollutant concentrations were calculated as the mean per 24-h period, except for O₃, which was calculated as the flow-weighted mean concentration per 8-h period. The concentrations of air pollutants are presented in μg/m³.

Statistical Analysis

SPSS 24.0 software was used for statistical analysis. Relationships between air pollutant concentrations and climatic variables with the main pathogens using multiple linear stepwise regression analysis. Statistical significance was set at P values < 0.05.

Results

Demographic Characteristics

Between 2017 and 2021, a total of 24,555 bacterial strains were isolated from inpatients, with 755 isolates from women and 23,800 from children in our study. The children were further categorized into age groups as follows: 10,124 infants (0–1 year old), 5469 toddlers (1–3 years old), 4477 preschool-aged children (3–6 years old), and 3730 school-aged children (6–18 years old). Across all age groups, including adults and children, there were 14,926 males and 9629 females.

Distribution of Bacterial Pathogens

In this study, we identified a predominance of gram-negative bacteria (54.70%) and gram-positive bacteria (44.57%), with fungi comprising a markedly smaller proportion (0.73%). The most frequently isolated bacteria were Haemophilus influenzae, Streptococcus pneumoniae, and Staphylococcus aureus, as detailed in Table 1. Sputum was the primary sample source (45.61%), followed by pus and secretions (31.90%) (Figure 1A). A notable decrease in isolates from sputum was observed in 2020, likely due to the COVID-19 pandemic’s impact on sample collection, as detailed in Figure 1B. There were significant changes in the bacterial spectrum over the study period, particularly in 2020, coinciding with the global outbreak of COVID-19. Notably, H. influenzae and S. pneumoniae were the most prevalent pathogens in sputum samples, whereas E. coli was primarily found in urine samples. Figures 2A–D illustrate the annual fluctuations in the prevalence of key pathogens in different sample types.

Table 1 Distribution and Proportion of Pathogenic Bacteria from 2017 to 2021

Figure 1 (A) Percentage of positive culture from various specimen types during study period. (B) Percentage of positive culture from various specimen types per year from 2017–2021.

Figure 2 (A) Distribution of main bacteria in sputum from 2017–2021. (B) Distribution of main bacteria in pus and secretions from 2017–2021. (C) Distribution of main bacteria in blood from 2017–2021. (D) Distribution of main bacteria in urine from 2017–2021.

Age Distribution Prevalence Characteristics of Main Pathogens

Our analysis identified distinct patterns in the prevalence of pathogens across age groups. Notably, H. influenzae and S. pneumoniae were the predominant pathogens in children aged under 6 years. In addition, E. coli was most prevalent among school-aged children (30.72%), whereas S. aureus was the most common in adults (39.47%). Furthermore, the prevalence of both H. influenzae and S. pneumoniae gradually decreased with the age increasing beyond three years, as shown in Table 2.

Table 2 Composition Ratio of Pathogens at Different Ages

Drug Resistance of Main Pathogens to Common Antimicrobials

Throughout the study period from 2017 to 2021, evolving resistance patterns were documented among key pathogens. H. influenzae saw an increase in resistance to ampicillin, escalating from 51.8% to 81.9%, and it showed similar trends for trimethoprim–sulfamethoxazole and azithromycin, with resistance peaking at 83.6% and 59.4%, respectively. Conversely, resistance to cefotaxime remained exceptionally low, consistently below 1% from 2019 to 2021(Supplementary Table 1). Resistance rates to penicillin G and amoxicillin in S. pneumoniae have gradually decreased from 2017 to 2021, as shown in Supplementary Table 2. Interestingly, S. aureus consistently showed resistance to penicillin, exceeding 95% (Supplementary Table 3), and E. coli showed high resistance to ampicillin, approaching 90%; however, it was markedly susceptible to cefotetan, piperacillin-tazobactam, and amikacin, as detailed in Supplementary Table 4. In addition, K. pneumoniae’s resistance to gentamicin significantly decreased from 39.7% to 9.1% from 2017 to 2021. By 2021, the resistance of K. pneumoniae to imipenem and meropenem was ~10%, as shown in Supplementary Table 5. Moraxella catarrhalis also showed a significant reduction in azithromycin resistance, which declined from 52.5% to 22.8% from 2017 to 2021, and furthermore, it showed low resistance to cephalosporins (Supplementary Table 6).

Distribution of Multidrug-Resistant Organisms (MDROs)

Throughout the study period, significant trends in the prevalence of MDROs were documented. The prevalence of MRSA decreased from 34.3% in 2018 to 29.8% in 2021. The detection rates of CR-PAE also dramatically dropped from 17.3% to 3.74% over the same period. In contrast, the detection rate for CR-ABA increased consistently, peaking at 45.45% in 2021. The detection rates for CR-ECO remained stable, whereas those for CR-KPN showed significant fluctuations. Notably, no cases of vancomycin-resistant Enterococcus faecalis (VREM) or penicillin-resistant S. pneumoniae (PRSP) were detected. Our findings, compared with the National Bacterial Resistance Network, show lower rates for CR-ABA and CR-PAE but higher rates for CR-ECO and CR-KPN, as detailed in Figure 3.

Figure 3 The trend in detection rates of multidrug-resistant organisms (MDROs) from 2017 to 2021 in our hospital and nationwide. The bar graph on the Y-axis represents the detection rates of multidrug-resistant organisms in our hospital; the line graph represents the national detection rates of MDROs. The X-axis represents the years.

Seasonal Distribution of Main Pathogens

Overall, during the 5-year study period, H. influenzae showed similar seasonal patterns and clear prevalence peaks every year (except for 2020). Every year, the H. influenzae detection rate peaked during the change of season from winter to spring (February–April), and M. catarrhalis and S. pneumoniae detection rates peaked when summer turned to autumn. H. influenzae, M. catarrhalis, and S. aureus declined significantly by 2020 (Figure 4).

Figure 4 Seasonal distribution of H. influenzae, (S) pneumoniae, (S) aureus, (E) coli, (M) catarrhalis, and K. pneumoniae.

Multiple Regression Analysis of the Association Between Number of Isolates with Main Pathogens and Environmental Factors

Considering the correlation between various environmental factors (co-linearity), a multiple linear stepwise regression analysis was conducted with the number of samples detected for the six main pathogens (H. influenzae, S. pneumoniae, S. aureus, E. coli, M. catarrhalis, and K. pneumoniae) as the dependent variable, and monthly sunshine hours, average temperature, relative humidity, volume of precipitation, and PM2.5, PM10, NO₂, and O₃ concentrations as the independent variables (Table 3). The analysis resulted in six models with adjusted R² values ranging from 0.112 to 0.338, suggesting variable explanatory power across different pathogens. The number of H. influenzae-positive samples was positively correlated with PM10 concentration (coefficient = 0.386, P < 0.05). The number of S. pneumoniae-positive samples was positively correlated with PM2.5 (coefficient = 0.591, P < 0.001). The number of S. aureus- and M. catarrhalis-positive samples were positively correlated with NO₂ concentration (coefficient = 0.357 and 0.398, respectively; P < 0.05). The number of E. coli-positive samples were positively correlated with PM10 and average temperature (coefficient = 0.784 and 1.139, respectively; P < 0.05) and negatively correlated with relative O₃ (coefficient = −0.562, P < 0.05). The number of K. pneumoniae-positive samples were positively correlated with PM2.5 and average temperature (coefficient = 1.071 and 1.028, respectively; P < 0.05) (Supplementary Figures 1–5). Variables such as relative humidity, average volume of precipitation, and sunlight hours were excluded from the final models because of their statistically insignificant influence.

Table 3 Mutiple Linear Stepwise Regression Analysis Model for Correlation Between Air Pollution Concentration, Meteorological Factor and Main Pathogens

Discussion

Our findings revealed that S. pneumoniae was the most common gram-positive bacterium and H. influenzae was the most common gram-negative bacterium; these findings are not consistent with previous reports.^11,12 Among all specimen types, sputum samples had the highest pathogen positivity rate; however, Chanda et al reported in their study that the most common pathogen-positive specimens in hospitalized patients were urine samples, blood samples, and wound swabs, and the most common microorganisms were coliforms.¹³ Furthermore, blood culture analyses revealed a shifting prevalence of predominant bacteria. The observed decline in S. epidermidis detection coincides with improved infection control practices, notably the improved hand hygiene practices following the implementation of evidence-based policies, which significantly reduced false-positive rates in coagulase-negative staphylococcal cultures.¹⁴ These results indicate that the distribution of bacteria and specimen sources differed among different regions and populations.

The variability in bacterial prevalence among different age groups was pronounced, particularly for H. influenzae and S. pneumoniae, which were the most prevalent in children aged under 6 years. This variability may be due to several factors. First, neither the H. influenzae type b (Hib) vaccine nor the thirteen-valent pneumococcal conjugate vaccine (PCV13) is included in China’s national immunization program, leading to low voluntary vaccination rates. In addition, inconsistencies in vaccine timeliness and completion have also contributed to this issue.^15,16 Furthermore, the 76-day COVID-19 blockade significantly disrupted vaccinations, including those for Hib and PCV13, impacting overall vaccination rates. These factors together explain the significant prevalence of these pathogens among younger children.

In China, the rate of ampicillin resistance in H. influenzae strains increased steadily from 12% in 2000–2002 to 58.1% in 2016.¹⁷ The current study revealed an even higher ampicillin resistance rate (88.4%). This significant increase indicated a rapid increase in ampicillin resistance among H. influenzae isolates in China. However, the resistance rate to cefotaxime was <1%, which was much lower than that reported in Iran (33.1%).¹⁸ Therefore, third-generation cephalosporins such as cefotaxime could be considered the preferred treatment option for infections caused by ampicillin-resistant H. influenzae in China. Importantly, the majority of strains included in our study were isolated from respiratory tract samples. In China, azithromycin is commonly used as an empirical medication for respiratory tract infections in children. A multicenter study showed that the resistance rate of H. influenzae to azithromycin increased as follows: 29.89% in 2017, 37.55% in 2018, and 42.04% in 2019.¹⁹ Therefore, the widespread use of azithromycin, coupled with increasing antibiotic resistance rates, may have contributed to the observed high level of azithromycin resistance.

S.pneumoniae is a prominent pathogen responsible for acute respiratory infections worldwide, and its antibiotic resistance rate varies significantly among children from different geographical areas.²⁰ Based on our findings, penicillin G could be considered as the first-line treatment for pneumococcal infections in children, particularly in Wuhan. We found E. coli exhibited a low resistance to cefotetan. Cefotetan, a cephamycin antibiotic developed in the 1980s, is a broad-spectrum antibiotic against gram-negative bacteria with potential as a carbapenem-sparing treatment option.²¹ Notably, K. pneumoniae exhibited higher susceptibility to carbapenems and aminoglycosides such as amikacin and gentamicin, thus corroborating a previous report by Xu et al.²² Our study further supports the notion that aminoglycosides and carbapenems are preferred therapeutic agents for treating K. pneumoniae infections in Wuhan.

Few studies have evaluated the prevalence of MDRO in pediatric patients. The prevalence of MDROs like MRSA has shown a slow decline, which is consistent with the trends reported by the China Antimicrobial Surveillance Network (CHINET), suggesting the effectiveness of enhanced infection control and antibiotic stewardship programs.²³ Despite the relatively lower detection rates of CR-ABA and CR-PAE compared to the national average levels, their presence remains concerning due to their association with high mortality and resistance.²⁴ In addition, although the absence of VREM and PRSP from 2017 to 2021 is encouraging, it is important to note that these pathogens continue to pose a significant threat to global health. Vancomycin-resistant enterococci (VRE), in particular, are considered a priority pathogen by the World Health Organization due to their high level of antimicrobial resistance and ability to cause severe infections.²⁵

The seasonal distribution of key pathogens, particularly H. influenzae, plays a crucial role in shaping prevention and control strategies for bacterial infections. Our findings show that the prevalence of H. influenzae infection typically peaks from February to April every year, a pattern that was also noted by regional studies in Zhejiang, China.²⁶ However, deviations from this pattern have also been noted, such as in Chengdu, where peak detections were noted in May and November,²⁷ and this difference may be attributable to climatic differences across regions. The substantial decline in detection rates of several pathogens, such as H. influenzae, M. catarrhalis, and S. aureus, in 2020 is notably linked to the COVID-19 pandemic’s profound impact on public health practices and social behaviors, including reduced travel and human interactions. These changes have significantly influenced pathogen transmission dynamics. Understanding these seasonal and situational factors is vital for developing effective infection control strategies that are tailored to specific environmental and societal contexts. Future studies should continue to investigate the underlying mechanisms linking seasonal changes to bacterial infection rates and refine strategies to more effectively prevent and manage these infections throughout the year.

To our knowledge, this is the first large-scale epidemiological investigation to comprehensively examine the relationship between meteorological factors and pathogenic infections in hospitalized children and women. This retrospective cohort study spanned five years and aimed to elucidate the potential associations between various meteorological parameters and the detection rates of H. influenzae, S. pneumoniae, S. aureus, M. catarrhalis, and E. coli. Our findings reveal interesting associations between meteorological factors and specific pathogens. We found that higher levels of particulate matter PM10 and PM2.5 in the air were associated with increased detection rates of H. influenzae, S. pneumoniae, and K. pneumoniae. A study in Suzhou, China also reported PM10 as an independent risk factor for the detection of H. influenzae.²⁸ This suggests that particles in air act as carriers that facilitate the transmission of these pathogens into the respiratory system. Furthermore, we observed a correlation between NO₂ concentrations and increased detection of S. aureus and M. catarrhalis. NO₂ may adversely affect the immune system, potentially weakening the body’s defense against these pathogens. Interestingly, higher average temperatures were associated with higher detection rates for E. coli and K. pneumoniae. This could be attributed to the increased temperature, which creates a more favorable environment for the growth and reproduction of these pathogens. Conversely, a negative correlation was observed between the relative concentration of O₃ and detection rate of E. coli. This may be attributed to the antimicrobial properties of ozone because higher O₃ levels can inhibit the growth of E. coli. Although the adjusted R² values in our models are relatively low, they importantly highlight the complex and multifactorial nature of pathogen dynamics. Although this moderate level of explanatory power does not capture all variance, it remains significant given the inherent complexity of the ecological interactions that influence pathogen spread and survival. These values suggest that additional, unidentified factors likely play substantial roles; nevertheless, the environmental variables identified by us still provide crucial insights into pathogen distribution patterns. Such understanding is of vital importance to develop more targeted and effective environmental and public health interventions. Notably, these correlations did not establish definitive causal relationships or detailed mechanisms. Our findings provide valuable insights into the potential influence of air pollution (PM2.5, PM10, NO₂, and O₃) and ambient temperature on the prevalence of specific pathogens in hospitalized pediatric and women patients. Further empirical research is needed to validate these associations and gain a deeper understanding of the complex relationships between pathogens and environmental factors.

Although this study provides insightful data on the epidemiology of bacterial pathogens and their correlation with environmental factors, it has several limitations that should be considered when interpreting the results. First, as a single-center study, the findings may not fully represent the broader population, particularly since the hospital’s catchment area is primarily Wuhan. Second, a minority of patients were from areas near Wuhan, but the environmental data collected did not comprehensively cover these regions. Therefore, the associations identified between pathogen prevalence and environmental factors may contain certain inaccuracies and thus not fully reflect the actual conditions experienced by these specific patients. In the future, multi-center studies are warranted to address these limitations and provide a more comprehensive understanding of the pathogens’ epidemiology and their environmental interactions. This would expand the geographic and demographic diversity of the patient samples, thus enhancing the representativeness and accuracy of the findings. In addition, precise environmental mapping of patients’ locations would be beneficial to draw a more accurate correlation between environmental exposures and pathogen prevalence.

Conclusion and Recommendation

This study underscores the significant variability in the distribution and antibiotic resistance of bacterial pathogens in Wuhan, accentuating the impact of local environmental factors. Importantly, H. influenzae and S. pneumoniae were found to primarily affect children under 6 years of age, highlighting the urgent need for targeted public health measures, including the integration of Hib and PCV13 vaccines into China’s national immunization schedule to potentially reduce the disease burden substantially. Effective first-line treatments identified include ceftriaxone for ampicillin-resistant H. influenzae, penicillin G for S. pneumoniae, and cefotetan for E. coli. The prevalence of multidrug-resistant organisms remains below national averages, suggesting effective local healthcare strategies; however, continued vigilance remains necessary. Seasonal trends show H. influenzae peaking in late winter and early spring, whereas M. catarrhalis and S. pneumoniae peak in late summer to early autumn. Environmental factors, such as PM2.5 and PM10, are positively correlated with the increased detection of H. influenzae, S. pneumoniae, and K. pneumoniae, whereas NO2 is linked to increases in S. aureus and M. catarrhalis. Conversely, a negative correlation was observed between E. coli counts and ozone levels. This research advocates for extensive studies across multiple regions to develop comprehensive strategies that address pathogen behavior under diverse environmental influences, thus enhancing global public health responses, particularly in light of the challenges posed by global events like the COVID-19 pandemic.