Clinical features and risk factors of Klebsiella pneumoniae infection in premature infants: a retrospective cohort study (2024)

Background

With the continuous advancement of modern medical technology, the survival rate of premature infants has significantly increased. Klebsiella pneumoniae (K. pneumoniae) is one of the most common pathogens causing neonatal infections, particularly posing a serious risk to premature infants. This study aimed to analyze the clinical characteristics, antibiotic susceptibility profiles, and treatment outcomes of K. pneumoniae infections in these infants.

Methods

We retrospectively compared cases of K. pneumoniae infection in premature and term infants admitted in a tertiary hospital from January 2017 to December 2022 in China. Clinical and microbiological characteristics were evaluated. Data analysis was performed using the Statistical Package for the Social Sciences (SPSS), with statistical significance defined as P < 0.05.

Results

We enrolled 166 premature infants and 68 term infants. In premature infants, fetal distress, patent ductus arteriosus, patent foramen ovale, enteritis, anemia, hypoproteinemia, bloodstream infections, abdominal infection, mechanical ventilation, nasogastric feeding, drainage tube, parenteral nutrition, and prior exposure to carbapenem antibiotics were identified as significant risk factors for K. pneumoniae infections in univariate analysis. Furthermore, septic shock, bloodstream infections, abdominal infections, indwelling catheters, drainage tubes, parenteral nutrition, and previous exposure to glycopeptide antibiotics were significantly correlated with mortality. Independent risk factors for K. pneumoniae infections in premature infants included fetal distress (OR: 3.702, [95% CI: 1.056–12.986], P = 0.041), enteritis (OR: 4.434, [95% CI: 1.066–18.451], P = 0.041), anemia (OR: 4.028, [95% CI: 1.550-10.466], P = 0.004), bloodstream infections (OR: 1.221, [95% CI: 0.061–1.802], P = 0.022), mechanical ventilation (OR: 4.974, [95% CI: 1.685–14.685], P = 0.004) and prior exposure to carbapenem antibiotic (OR: 14.738, [95% CI: 2.393–90.767], P = 0.004). Additionally, abdominal infections (OR: 8.598, [95% CI: 2.000-36.957], P = 0.004) and indwelling catheters (OR: 7.698, [95% CI: 0.998–59.370], P = 0.050) were positive predictors of mortality.

Conclusion

K. pneumoniae isolates exhibit a notable prevalence of infection, poor treatment outcomes, and elevated resistance in preterm neonates. These findings enhance our understanding of K. pneumoniae infections and their association with clinical outcomes among premature infants.

K. pneumoniae is one of the common pathogens for neonatal infections, mainly associated with multidrug resistance in cases of bloodstream infections, meningitis, urinary tract infections, and pneumonia [1,2,3]. Premature infants, in particular, are at serious risk of infection, threatening their lives and health [4]. In recent years, with the abuse and overuse of antibiotics, the resistance of K. pneumoniae to commonly used antibiotics has been increasing. According to the China Antimicrobial Surveillance Network (https://www.carss.cn), the mean resistance rates of third-generation cephalosporins and carbapenems in K. pneumoniae infections were 33.0% and 9.0% in 2017, and 27.7% and 10.0% in 2022, respectively. In outbreaks among neonatal units, K. pneumoniae was the most frequently implicated pathogen, often exhibiting resistance to commonly used antibiotics [5, 6].

With the constantly advancing modern medical technology, the survival rate of premature infants has shown a significant increase. Nonetheless, premature infants continue to encounter an array of health risks, with infection ranking as a primary causative factor of mortality in this population [7]. The immature immune system of newborns is closely associated with a higher rate of infections among neonates [8]. Premature infants exhibited more immature immune responses and had a higher risk of infection compared to full-term infants [9]. Of the 2.761million children who died during the neonatal period, preterm birth complications were responsible for 35.0% of the deaths, and neonatal sepsis accounted for 15.3% [10]. Potential risk factors for infections caused by Klebsiella pneumoniae-resistant strains in infants include prolonged length of stay, previous exposure to cephalosporin, haematological malignancies, and use of indwelling devices [11,12,13]. Nonetheless, there is a dearth of literature regarding the predisposing factors for K. pneumoniae infection in preterm neonates.

Our study aimed to provide a comprehensive report on the clinical features, risk factors, antibiotic susceptibility, and treatment outcomes associated with K. pneumoniae infections in premature infants. Therefore, to gather accurate and up-to-date information on the clinical characteristics, resistance profiles, and treatment outcomes of premature infants infected with K. pneumoniae, we conducted a retrospective cohort study in the neonatal ward of a provincial children’s hospital from 2017 to 2022.

Study setting and patients

This retrospective study was conducted in the neonatal ward of the Maternal and Child Health Hospital of Guangxi Zhuang Autonomous Region from January 2017 to December 2022. Infants were diagnosed with a bacterial infection caused by K. pneumoniae. The inclusion criteria were defined as follows: the first incidence of K. pneumoniae infection falling within the study period, infants meeting the diagnostic criteria for infection, solely considering the initial occurrence of K. pneumoniae infection, and the exclusion of any cases of recurrent infections. Conversely, cases with incomplete or unavailable medical records were systematically excluded from the analysis.( Fig.1).

To identify the risk factors for K. pneumoniae infection in premature infants, the infants were divided into two groups: a premature infant group and a term infant group. Additionally, to explore the risk factors for the treatment outcomes of K. pneumoniae infection in premature infants, the latter were further divided into a mortality group and a survival group.

Data collection and definitions

The medical data were collected from the hospital’s electronic health records system. The collected data encompassed essential demographic information (gender and days of age), previous hospitalizations, cesarean section, fetal distress, premature rupture of membrane, contamination of amniotic fluid, infants of diabetic mother, breast-fed, underlying diseases, infection type, invasive procedures and devices, antibiotic exposure, and outcomes.

The definition of a premature infant is an infant delivered prior to the completion of 37 weeks of gestation. The definition of a term infant is one born following a gestation period of at least 37 weeks. An infant was defined as infected if the clinical manifestations, laboratory results, and imaging findings were compatible with a clinical infection associated with isolation of K. pneumoniae in a pertinent biological specimen (e.g., Sputum, blood, sterile body fluid). A history of hospitalization refers to a person’s record of receiving treatment and being admitted to a medical facility, including transfers from other NICUs. Contamination of amniotic fluid refers to the presence of bacteria, viruses, or other pathogenic microorganisms in the amniotic fluid. The definition of enteritis is inflammation of the small intestine and colon, caused by viral, bacterial, fungal, and parasitic infections, accompanied by symptoms such as diarrhea, vomiting, fever, and dehydration. The definition of sepsis refers to a systemic condition of bacterial, viral, or fungal origin that is associated with hemodynamic changes and other clinical manifestations. Klebsiella sepsis refers to the presence of K. pneumoniae in blood cultures, causing systemic diseases that include symptoms related to changes in hemodynamic circulation and other clinical manifestations. The results were categorized as such: based on the medical documentation, patients who were discharged in a cured or improved state were identified as survival, whereas those who were either deceased or in a grave condition upon discharge were considered to have experienced a mortality.

Microbiological methods

In this study, all isolates were identified by the VITEK2 Compact system (BioMérieux, Marcy l’Etoile, France) or the Zhuhai DL-96II automatic bacterial identification drug sensitivity apparatus (Zhuhai DL Biotech Co., Ltd.,China). The antibiotic susceptibility assays were performed on the isolates utilizing either the disk-diffusion methodology or the VITEK 2 Compact system or the Zhuhai DL-96II automated bacterial identification drug sensitivity apparatus, with each protocol tailored to the specific antibiotics employed. The results were interpreted following the guidelines outlined by the Clinical and Laboratory Standards Institute (CLSI, version 2021). The quality control strains employed in this study were P. aeruginosa ATCC27853 and E. coli ATCC25922 (National Center for Clinical Laboratories, Beijing, China). In our study, Minimum inhibitory concentrations (MICs) of common clinical antibiotics have been determined included ceftazidime, cefepime, ampicillin-salbactam, ceftriaxone, meropenem, piperacillin-tazobactam, imipenem, amikacin, gentamicin, minocycline, levofloxacin, sulfamethoxazole, ciprofloxacin, ticarcillin-clavulnic acid, chloramphenicol, amoxicillin, cefuroxime, cefazolin, and cefoxitin.

Statistical analysis

Data analysis was performed using IBM Statistical Product and Service Solutions (SPSS) version 25.0. Measurement data were reported as mean ± standard deviation, and count data were reported as percentages. Continuous variables were analyzed using Student’s t-test or Mann-Whitney U-test, while categorical variables were analyzed using χ2 or Fisher’s exact test. Statistical significance was considered at P < 0.05. Univariate logistic regression analyses were conducted to identify risk factors. To further investigate independent risk factors, all statistically significant variables were included in the multivariate model.

Characteristics of patients

During the study period, a total of 234 unique cases of neonatal infections with K. pneumoniae were identified. According to the definition of premature infants, 166 cases (70.9%) were preterm, while 68 cases (29.1%) were term infants. There were 71 infants (30.3%) less than 14 days old, with the proportion in the preterm infant group significantly lower than in the term infant group (22.8% vs. 48.5%, P = 0.00). Of the total infants, 156 (66.6%) were male, and 78 (33.4%) were female. A history of hospitalization was reported in 119 infants (50.8%), and 114 (48.7%) were delivered by cesarean section, while 106 (45.2%) were breastfed. Additionally, there were 69 cases (29.4%) of fetal distress, 75 cases (32.0%) of premature rupture of membranes, and 46 cases (19.6%) of amniotic fluid contamination. Most infants had at least one underlying condition, including jaundice (137 cases, 58.5%), acleistocardia (130 cases, 55.5%), anemia (95 cases, 40.5%), neonatal subependymal hemorrhage (53 cases, 22.6%), patent ductus arteriosus (52 cases, 22.2%), enteritis (48 cases, 20.5%), and hypoproteinemia (27 cases, 11.5%). The primary types of infection included pneumonia (128 cases, 54.7%) and bacteremia (57 cases, 24.3%), while the most common invasive procedures and devices were mechanical ventilation (161 cases, 68.9%) and parenteral nutrition (148 cases, 63.2%). The predominant antibiotic exposures involved carbapenems (60 cases, 25.6%) and β-lactamase inhibitors (151 cases, 64.5%). Furthermore, 202 infants (86.3%) were discharged after recovery. The demographic and clinical data of infants with K. pneumoniae infections are shown in Table1.

Microbiological characteristics of K. pneumoniae strains

A total of 234 isolates were identified as K. pneumoniae. All strains have completed antimicrobial susceptibility testing, and the drug resistance profiles were shown in Table2. The results indicated that the resistance rates for amoxicillin (82.4%), cefuroxime (58.5%), cefazolin (62.8%), ampicillin-sulbactam (55.1%), and ceftriaxone (51.7%) exceeded 50%, while resistance rates for other antibiotics remained below 50%. Additionally, our results revealed that the resistance rates for ceftazidime(p = 0.002), cefepime(p = 0.000), ampicillin-sulbactam(p = 0.000), ceftriaxone(p = 0.000), piperacillin-tazobactam(p = 0.048), levofloxacin(p = 0.011), sulfamethoxazole(p = 0.010), ticarcillin-clavulanic acid(p = 0.005), amoxicillin(p = 0.021), cefuroxime(p = 0.002), cefazolin(p = 0.000), cefoxitin(p = 0.017) were significantly higher in isolates from premature infants compared to those from term infants.

Furthermore, among the 166 strains of bacteria isolated from premature infants, the antibiotics with a drug resistance rate greater than 50% were ceftazidime (50.0%), cefepime (53.0%), ampicillin-salbactam (62.6%), ceftriaxone (59.0%), ticarcillin-clavulnic acid (51.2%), amoxicillin (86.1%), cefuroxime (65.0%), cefazolin (69.8%). Additionally, the results also unveiled that the resistance rates of ceftazidime (p = 0.025), meropenem(p = 0.002), piperacillin-tazobactam(p = 0.012), imipenem(p = 0.002) were significantly higher in the isolates from premature infants exhibiting mortality compared to those from premature infants with survival (Table3).

Risk factors for K. pneumoniae infection in premature infants

To identify risk factors associated with K. pneumoniae infection in premature infants, we conducted a case-control investigation. Univariate analysis revealed that the factors most significantly associated with K. pneumoniae infections in premature infants were fetal distress, patent ductus arteriosus, patent foramen ovale, enteritis, anemia, hypoproteinemia, bloodstream infections, abdominal infection, mechanical ventilation, nasogastric feeding, drainage tubes, parenteral nutrition, and prior exposure to carbapenem antibiotics. In contrast, the least relevant factors were age (< 14 days) and breastfeeding (Table1).

The results of the multivariate analysis were depicted in Table4. The independent risk factors for K. pneumoniae infections in premature infants were fetal distress (OR: 3.702, [95% CI: 1.056–12.986], P = 0.041), enteritis (OR: 4.434, [95% CI: 1.066–18.451], P = 0.041), anemia (OR: 4.028, [95% CI: 1.550-10.466], P = 0.004), bloodstream infections (OR: 1.221, [95% CI: 0.061–1.802], P = 0.022), mechanical ventilation (OR: 4.974, [95% CI: 1.685–14.685], P = 0.004), and prior exposure to carbapenem antibiotics (OR: 14.738, [95% CI: 2.393–90.767], P = 0.004).

Risk factors for poor prognosis of premature infants infected with K. pneumoniae

The rates of mortality and survival among premature infants infected with Klebsiella pneumoniae were 17.47% (29/166) and 82.53% (137/166), respectively. Univariate analysis indicated that, among these premature infants, septic shock, bacteremia, abdominal infection, indwelling catheters, drainage tubes, parenteral nutrition, and prior exposure to glycopeptide antibiotics were significantly correlated with mortality (Table5).

In the multivariate analysis, septic shock (OR: 5.557, [95% CI: 1.411–21.890], P = 0.014), abdominal infection (OR: 8.598, [95% CI: 2.000-36.957], P = 0.004), and indwelling catheters (OR: 7.698, [95% CI: 0.998–59.370], P = 0.050) were identified as significant predictors of negative outcomes in premature infants infected with Klebsiella pneumoniae (Table6).

Klebsiella pneumoniae is common in neonates across several countries, whether as an infection or colonization, posing a significant threat to affected infants. K. pneumoniae is associated with longer hospitalization and higher morbidity and mortality [7, 14, 15]. Although studies have addressed neonatal bloodstream infections or sepsis caused by K. pneumoniae or multidrug-resistant strains, infections in premature infants caused by K. pneumoniae remain relatively rare [16,17,18]. Understanding the clinical features and risk factors of K. pneumoniae infection in premature infants is crucial for effective management and prevention.

In this retrospective cohort study, we investigated the clinical features of K. pneumoniae infection in premature infants and identified associated risk factors. In our study, it has been discovered that preterm neonates constitute the predominant cohort of hospitalized infants afflicted with K. pneumoniae. Compared to term infants, premature infants had a higher probability of K. pneumoniae, a higher rate of antibacterial resistance, and worse treatment outcomes. Multivariate analysis indicated that fetal distress, enteritis, anemia, bloodstream infections, mechanical ventilation, and prior exposure to carbapenem antibiotics were independent risk factors for K. pneumoniae infections in premature infants. Additionally, septic shock, abdominal infection, and indwelling catheters were identified as positive predictors ofmortality in premature infants infected with K. pneumoniae.

In our research, we observed that the resistance rates of amoxicillin, cefuroxime, cefazolin, ampicillin-sulbactam, and ceftriaxone were relatively high, which were consistent with other reports [19, 20]. The proportion of CRKP strains in our study was 20.0%, much lower than the proportions for adults in the CHINET report (26.3%) [21]. Our results also revealed that the resistance rates for ceftazidime, cefepime, ampicillin-sulbactam, ceftriaxone, piperacillin-tazobactam, levofloxacin, sulfamethoxazole, ticarcillin-clavulanic acid, amoxicillin, cefuroxime, cefazolin and cefoxitin were notably higher in premature infants’ isolates compared to those from term infants. Based on our preliminary statistics, except for levofloxacin, these antibiotics were commonly used for neonatal infection prevention in our hospital. Interestingly, the resistance rates of ceftazidime, meropenem, piperacillin-tazobactam, and imipenem were significantly higher in the isolates from premature infants exhibitingmortality compared to those from premature infants with survivals. This suggests that preterm neonates are more vulnerable to K. pneumoniae, which shows high resistance to commonly used antibiotics, increasing the risk of unfavorable therapeutic outcomes. Premature infants, who have not fully developed during pregnancy, experience incomplete organ function and an immature immune system, making them more susceptible to infections.

In infants, the risk factors associated with K. pneumoniae or multidrug-resistant K. pneumoniae infection may vary slightly across different countries. Common factors include previous and/or ongoing antibiotic therapy, previous and/or long hospitalization, invasive procedures such as intubation, cesarean section, and catheterization, as well as low birth weight, preterm birth, total parenteral nutrition, infection type, and mechanical ventilation [5, 22,23,24]. The risk factor of preterm birth appeared in several studies, but no relevant literature was found regarding risk factors for K. pneumoniae infection in preterm infants specifically. In our study, the risk factors associated with K. pneumoniae infections in premature infants included fetal distress, patent ductus arteriosus, patent foramen ovale, enteritis, anemia, hypoproteinemia, bloodstream infections, abdominal infection, mechanical ventilation, nasogastric feeding, drainage tubes, parenteral nutrition, and prior exposure to carbapenem antibiotics. Notably, fetal distress, enteritis, anemia, bloodstream infections, mechanical ventilation, and prior exposure to carbapenem antibiotics were independent risk factors for K. pneumoniae infections in premature infants. Apart from fetal distress, patent ductus arteriosus, patent foramen ovale, enteritis, anemia, and hypoproteinemia, the other risk factors resemble the common factors associated with neonatal K. pneumoniae or multidrug-resistant K. pneumoniae infection. Fetal distress is a factor that can contribute to premature birth. It was also a risk factor for early-onset neonatal sepsis [25]. It increases the likelihood of bacterial colonization or exposure to pathogens during delivery, leading to infections such as pneumonia, sepsis, or meningitis. Patent ductus arteriosus and acleistocardia can result in abnormal cardiovascular function and poor blood circulation, impacting the body’s ability to combat infections. The intestine, as one of the largest immune organs, when compromised by enteritis, increases the risk of bacterial invasion. Anemia impairs oxygen delivery and immune function, while hypoproteinemia reduces the synthesis of antibodies and other essential proteins, decreasing resistance to infections. Interestingly, Ventilator Associated Pneumonia is the most common type of infection in this study, and mechanical ventilation is the most common invasive procedure in this research. Mechanical ventilation was also an independent risk factor for premature infants infected with K. pneumoniae. Therefore, reducing the use of mechanical ventilation is an effective way to reduce the occurrence of Ventilator Associated Pneumonia.

Additionally, univariate analysis indicated that septic shock, bloodstream infections, abdominal infection, indwelling catheters, drainage tubes, parenteral nutrition, and previous exposure to glycopeptide antibiotics were significantly correlated withmortality in premature infants infected with K. pneumoniae. Importantly, septic shock, abdominal infection, and indwelling catheters were identified as positive predictors ofmortality in premature infants infected with K. pneumoniae. We have observed that bloodstream infections, abdominal infection, indwelling catheters, drainage tubes, parenteral nutrition are not only risk factors for infections but also for negative treatment outcomes. Septic shock is a risk factor for poor treatment outcomes, consistent with previous research results [7]. Therefore, it is crucial to focus on infection prevention, early identification, and timely treatment of potential infection risks in premature infants.

Based on the results of multivariate analysis, clinical doctors should be vigilant about the possibility of K. pneumoniae infection when premature infants exhibit fetal distress, enteritis, anemia, bacteremia, require mechanical ventilation, and have been previously exposed to carbapenem antibiotics. Especially in cases where premature infants infected with K. pneumoniae develop septic shock, abdominal infections, and have indwelling catheters, early intervention measures should be taken to prevent further deterioration and avoid treatment failure.

Importantly, K. pneumoniae causing neonatal sepsis is part of our research. Neonatal sepsis is further divided into early-onset sepsis and late-onset sepsis [7]. The clinical manifestations of early-onset infections typically manifest within the first 72h of life. Early-onset infections are typically acquired before or during delivery, often through vertical mother-to-infant transmission. Late-onset infections develop after delivery, typically occurring beyond 3 to 7 days of age, and are usually caused by organisms acquired from interactions within the hospital environment or the community. Early-onset sepsis and late-onset sepsis can have different causes, risk factors, and outcomes. Stratifying sepsis by onset time helps in understanding the timing of infection and potentially different management strategies required for each. While a classification analysis was omitted in this study, the significance of such categorization in elucidating our findings cannot be understated. Moreover, it presents a promising avenue for future investigations in the realm of classification research.

The current study had some limitations. The most significant limitation was that it was a retrospective observational analysis conducted at a single center, which made it prone to selection bias. In future research, we aim to replicate these findings in larger, more diverse centers to confirm the broader generalizability of our results. Second, our research only focused on infections caused by K. pneumoniae and did not involve infections caused by fungi, viruses, and other pathogens. For example, previous studies have indicated that preterm infants were vulnerable to a multitude of pathogens and predisposed to polymicrobial infections [10, 26]. Third, we have only conducted research on drug resistance phenotypes, not on the molecular level, such as drug resistance mechanisms and molecular epidemiology studies. The association between the molecular level of strains and clinical characteristics is key to preventing and controlling infections, which is the focus of our future work. Fourth, we did not conduct a subgroup analysis of late premature infants and very premature infants, which may lead to incomplete research results. Finally, this study focused on the differentiation of K. pneumoniae infection in premature and term infants, but neglected to differentiate between early and late sepsis caused by K. pneumoniae, which is crucial for infection prevention, early identification, and timely treatment of potential infection risks for premature infants.

Authors and Affiliations

1. Medical Science Laboratory, Children’s Hospital, Maternal and Child Health Hospital of Guangxi Zhuang Autonomous Region, Nanning, 530003, People’s Republic of China

   Xiaofen Wei,Jiahui Liang,Huan Zhang,Chenglan Yan,Xiangjun Lu,Yan Chen&Linlin Li

Contributions

Conceptualization, L L; Methodology, L L; Software, L L; Validation, L L, H Z and J L, Formal Analysis, L L; Investigation, X W; Resources, C Y and H Z; Data Curation, X W，X L and Y C; Writing – Original Draft Preparation, L L; Writing – Review & Editing, L L; Visualization, L L; Supervision, L L; Project Administration, L L; Funding Acquisition, L L and J L.

Corresponding author

Correspondence to Linlin Li.

Ethics approval and consent to participate

The study was approved by the ethics committee (Institutional Review Board of the Maternal and Child Health Hospital of Guangxi Zhuang Autonomous Region) with approval No. 2023-3-42 and as our research was retrospective and primarily utilized existing medical records data, the data were anonymized during analysis, protecting the privacy of the participants and avoiding imposing additional risks or burdens on them. Therefore, the ethics committee (Institutional Review Board of the Maternal and Child Health Hospital of Guangxi Zhuang Autonomous Region) waived the requirement for informed consent.

Consent for publication

Not Applicable.

Competing interests

The authors declare no competing interests.

Clinical trial number

Not Applicable.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions