Clinical characteristics, bacterial profiles and antimicrobial resistance of neonates with bloodstream infection caused by ESKAPEEc pathogens in Southwest China Authors: Xiaoshan Peng, Biao Chen, Chunmei Jing, Hong Wei, Hongmei Xu, Ruiqiu Zhao Published: 2025 ABSTRACT AbstractBackgroundData focusing on neonatal bloodstream infections (BSIs) caused by ESKAPEEc pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, Enterobacter spp., and Escherichia coli ) was scarce. This study was conducted to explore the clinical and microbiological aspects of ESKAPEEc BSI in neonates.MethodsA retrospective study was performed in neonates with culture-confirmed BSI at the Children’s Hospital of Chongqing Medical University, between January 2015 and December 2020. Clinical data, causative organisms, and antimicrobial susceptibility profiles were comprehensively collected and analyzed.ResultsIn total, 625 patients were enrolled; 346 had infections caused by ESKAPEEc pathogens (prevalence, 55.4%; 95% confidence interval, 51.5–59.3%). Klebsiella pneumoniae (144/625; 23.0%) was the predominant ESKAPEEc pathogen causing neonatal BSI, followed by Escherichia coli (104/625; 16.6%), and Enterococcus faecium (55/625; 8.8%). Klebsiella pneumoniae (138/556; 24.8%) was the dominant ESKAPEEc in neonates presenting with late-onset sepsis (LOS), while Escherichia coli (14/69; 20.3%) was the most frequent ESKAPEEc isolated among patients with early-onset sepsis (EOS). Moreover, the resistance rates of Gram-negative ESKAPEEc to third-generation cephalosporins, piperacillin/tazobactam, aminoglycosides, and fluoroquinolones were statistically higher than the values of the Gram-negative non-ESKAPEEc (all p < 0.05). Notably, 9.0%–31.8% of Gram-negative ESKAPEEc and 0%–25.9% of Gram-negative non-ESKAPEEc species were resistant to carbapenem, whereas the carbapenem resistance rates were not statistically different between the 2 groups (all p > 0.05). Gram-positive ESKAPEEc and Gram-positive non-ESKAPEEc isolates showed relatively high resistance levels to multiple antibiotics. 1(1/331; 0.3%) vancomycin-resistant coagulase-negative Staphylococcus (CoNS) was observed. In addition, low birthweight and preterm birth were significantly associated with both ESKAPEEc infection and ESKAPEEc-related mortality (all P < 0.05). Neonates with ESKAPEEc BSI had higher rates of concurrent meningitis (17.9% vs. 9.3%, p = 0.002) and mortality (6.4% vs. 2.5%, p = 0.023).ConclusionsOver half of neonates presenting with BSI were ascribed to ESKAPEEc pathogens. Klebsiella pneumoniae and Escherichia coli were the dominant ESKAPEEc pathogens. The distribution of ESKAPEEc organisms in neonatal BSI differed by timing of infection onset. ESKAPEEc infections and ESKAPEEc-related mortality occurred more frequently among preterm neonates or those with lower birthweight. The high resistance rates of Gram-negative ESKAPEEc were of serious concern.Keywords: Bloodstream infection, ESKAPEEc pathogens, Neonate, Antimicrobial resistance Background Neonatal bloodstream infection (BSI) is a serious infectious disease, complicated by high morbidity and mortality. It has been reported that among the surviving newborns, the incidence rate ranged from 4.5‰ to 9.7‰, and the case fatality rate was up to 11%–19% [1, 2]. The barriers for diagnosing BSI in neonates include variable routes of infection and atypical clinical manifestations, which pose huge diagnostic challenges. Furthermore, the progress of neonatal BSI is rapid; if not handled in a timely manner, shock, multiple organ dysfunction, and even death may occur within a few hours. Therefore, early diagnosis and prompt initiation of antimicrobial treatment are of great significance in protecting against disease progression and improving the outcome of neonatal BSI. The unreasonable application of broad-spectrum antimicrobial agents is a fatal and critical issue, resulting in the generation of multidrug-resistant bacteria and the escalation of antimicrobial resistance (AMR). Antimicrobial-resistant ESKAPEEc pathogens (Enterococcus faecium (E. faecium), Staphylococcus aureus (S. aureus), Klebsiella pneumoniae (K.pneumoniae), Acinetobacter baumannii (A. baumannii), Pseudomonas aeruginosa (P. aeruginosa), Enterobacter spp, and Escherichia coli (E. coli) ) are thought to be a serious worldwide threat to human health, and their prevalence in BSIs is rising [3, 4]. Previous studies have indicated that the ESKAPEEc organisms contributed to 58.6%-76.3% of BSIs [5, 6] and were associated with increased costs of care and higher mortality [4]. Moreover, the high AMR rate of ESKAPEEc pathogens has brought about the limited selection of antimicrobial agents. Usually, choosing the proper empirical antibiotic for patients with bacteremia was based on the prevalence of causative bacteria and their AMR patterns, whereas the pathogens’ distribution and AMR profiles varied greatly according to the region and study population [7–11]. Hence, promptly investigating and analyzing the spectrum of ESKAPEEc pathogens leading to neonatal BSI, along with their susceptibility data, is essential to better characterize the clinical antibiotic choices for defending against these organisms. Recently, most of the existing research papers have focused on the epidemiology and AMR profiles of pediatric BSI [6, 10, 11], and a paucity of data concerning clinical aspects of ESKAPEEc BSI was accessible among hospitalized children [5]. To date, there is a lack of comprehensive research on ESKAPEEc-BSI in neonates. Therefore, we sought to explore the clinical characteristics, bacterial profiles, and AMR of neonatal BSI, and then compared them between the ESKAPEEc and non-ESKAPEEc groups. Methods Study population A retrospective study was conducted from January 2015 to December 2020, including all hospitalized neonates (aged ≤ 28 days) with laboratory-confirmed BSI at the Children’s Hospital of Chongqing Medical University. This hospital is a National Children’s Medical Center with 2,200 beds and the largest pediatric medical institution in southwest China, providing clinical services for approximately 3.07 million outpatient visits and 99,300 inpatient admissions in 2022. The confirmed diagnosis of BSI was based on the presence of BSI symptoms and positive blood cultures [12]. All eligible neonates were treated with empirical antibiotics according to the professional judgment of the clinicians. We excluded polymicrobial infections or duplicate isolates from the same patient. If the records or resistance data of the enrolled patients were incomplete, these patients were excluded from the study. The data for our research were obtained directly from medical and laboratory records, such as age at admission, gender, gestational age at birth, birth weight, mode of delivery, premature rupture of membranes (PROM), meconium-stained amniotic fluid (MSAF), outcome, diagnosis, symptoms, microbiology laboratory data (causative organisms and antimicrobial resistance results), and so on. Neonatal BSI was divided into early-onset sepsis (EOS) and late-onset sepsis (LOS) according to the onset time. EOS occurred within 3 days after birth, whereas LOS referred to infection occurring more than 3 days after birth. Furthermore, according to the gestational age, neonates who were born < 37 weeks of gestation were categorized as preterm, while neonates were categorized as term if they were born after 37 weeks of gestation. ESKAPEEc referred to a cluster of pathogens defined in previous studies [3–6]. The term “sluggish reflexes” in neonates with BSI was manifested by symptoms including lethargy, irritability, a reduction in feeding/crying/movement, or progressively worsening responsiveness. Mortality included two situations in this study. One situation meant the neonates died directly during hospitalization, and the other referred to some patients who abandoned treatment due to being critically ill and died during follow-up. Microbiological methods For infants with suspected sepsis, blood was collected aseptically from two separate sites prior to antimicrobial therapy and inoculated into paired culture bottles (≥‌ 2 bottles per episode). The volume drawn was determined by patient weight (2–4 mL per bottle). Specimens were processed in the clinical laboratory for pathogen identification. Blood cultures were incubated for up to 5 days (extendable upon clinical request). Positive blood cultures were analyzed and interpreted through a collaborative effort between microbiology laboratory specialists and clinical physicians. Bacterial identification was performed using the VITEK 2 system (bioMérieux, France) or the BD Phoenix system (Becton Dickinson, USA). Antimicrobial susceptibility testing results were interpreted based on the Clinical and Laboratory Standards Institute (CLSI) guidelines [13]. Statistical analysis Continuous variables were characterized as median (interquartile range [IQR]), while categorical variables were reported as frequency and percentage. Using Mann-Whitney U to compare continuous variables and the χ2 test or Fisher’s exact test to compare categorical variables. Related statistical software were used to analyses all the data. A P value < 0.05 was considered statistically significant. Results Study population During the 6-year period, a total of 625 neonates with laboratory-confirmed BSI were included in the final analysis based on the exclusion criteria. The cohort consisted of 380 males (60.8%) and 245 females (39.2%). The median age at admission was 6.1 days (IQR 0.2–15.3). The median gestational age and birth weight of the neonates were 37.3 weeks (IQR, 32.4–39.3) and 2.8 kg (IQR, 1.6–3.3), respectively. 556 (89.0%) cases were classified as LOS, and 69 (11.0%) as EOS. Of the infants, 346 were infected with ESKAPEEc BSI (Table 1), corresponding to a prevalence of 55.4% (95% confidence interval, 51.5–59.3%). Table 1Comparison clinical characteristics of the ESKAPEEc and non-ESKAPEEc BSI groupsVariableTotal(n = 625)ESKAPEEc(n = 346)Non-ESKAPEEc (n = 279)P ValueGender Male, No. (%)380(60.8)213(61.6)167(59.9)0.664 female, No. (%)245(39.2)133(38.4)112(40.1)Age at admission (days), Median(IQR)6.1(0.2–15.3)4.5(0.1–13.4)7.3(1.7–17.5) < 0.001 Birthweight (Kg), Median(IQR)2.8 (1.6–3.3)2.6 (1.4–3.2)3.0 (2.0-3.4) < 0.001 Low birth weight < 2.5Kg, n (%)255(40.8)168(48.6)87(31.2) < 0.001 Gestational age (weeks), Median(IQR)37.3(32.4–39.3)36.6(31.1–39.1)38.3(34.0-39.6) < 0.001 Preterm, < 37 weeks’ gestation, No. (%)288(46.1)182(52.6)106(38.0) < 0.001 Natural birth, No. (%)319(51.0)167(48.3)152(54.5)0.122PROM, No. (%)151(24.2)84(24.3)67(24.0)0.939MSAF, No. (%)77(12.3)40(11.6)37(13.3)0.520Clinical manifestation Fever, No. (%)173(27.7)99(28.6)74(26.5)0.562 Aggravated jaundice, No. (%)204(32.6)112(32.4)92(33.0)0.873 Sluggish reflexes, No. (%)329(52.6)200(57.8)129(46.2) 0.004 Apnea or shortness of breath, No. (%)279(44.6)162(46.8)117(41.9)0.222 Ventosity or vomit, No. (%)120(19.2)68(19.7)52(18.6)0.749 Convulsion, No. (%)31(5.0)15(4.3)16(5.7)0.423EOS, No. (%)69(11.0)31(9.0)38(13.6)0.065Concurrent meningitis, No. (%)88(14.1)62(17.9)26(9.3) 0.002 mortality, No. (%)29(4.6)22(6.4)7(2.5) 0.023 Abbreviations: BSI, bloodstream infection; PROM, Premature rupture of membranes; MSAF, meconium-stained amniotic fluid; EOS, early-onset sepsis; IQR, interquartile range. Boldface type means P < 0.05 Comparison clinical characteristics of the ESKAPEEc and non-ESKAPEEc BSI groups Abbreviations: BSI, bloodstream infection; PROM, Premature rupture of membranes; MSAF, meconium-stained amniotic fluid; EOS, early-onset sepsis; IQR, interquartile range. Boldface type means P < 0.05 Pathogen composition Of the 625 isolates, 331 (53.0%) were Gram-positive, and 294 (47.0%) were Gram-negative. Gram-positive ESKAPEEc, Gram-positive non-ESKAPEEc, Gram-negative ESKAPEEc, and Gram-negative non-ESKAPEEc were 79, 252, 267, and 27, respectively. As shown in Table 2, the predominant ESKAPEEc pathogens in our study were K. pneumoniae (144/625; 23.0%), E. coli (104/625; 16.6%), E. faecium (55/625; 8.8%), followed by S. aureus (24/625; 3.8%), Enterobacter spp. (14/625; 2.2%), A. baumannii (5/625; 0.8%), and P. aeruginosa (0; 0%). Among non-ESKAPEEc pathogens, Staphylococcus epidermidis (S. epidermidis) (143/625; 22.9%) and Enterococcus faecalis (32/625; 5.1%) were the most frequent. In addition, a significant increase in the percentage of ESKAPEEc was observed from 2015–2017 to 2018–2020 (47.7% vs. 68.2%, P < 0.001) (Fig. 1). Table 2Common bacteria isolated from neonatal bloodstream infection patients in China, 2015–2020OrganismTotalEOSLOSN (%)N (%)N (%) ESKAPEEc Enterococcus faecium 55(8.8)2(2.9)53(9.5) Staphylococcus aureus 24(3.8)5(7.3)19(3.4) Klebsiella pneumoniae 144(23.0)6(8.7)138(24.8) Acinetobacter baumannii 5(0.8)1(1.4)4(0.7) Pseudomonas aeruginosa 0(0)0(0)0(0)Enterobacter spp14(2.2)3(4.3)11(2.0) Escherichia coli 104(16.6)14(20.3)90(16.2) Non-ESKAPEEc Staphylococcus epidermidis 143(22.9)15(21.7)128(23.0) Staphylococcus hominis 15(2.4)0(0)15(2.7) Staphylococcus haemolyticus 31(5.0)5(7.3)26(4.7) Enterococcus faecalis 32(5.1)5(7.3)27(4.9)Group B Streptococcus23(3.7)11(15.9)12(2.2) Serratia marcescens 14(2.2)0(0)14(2.5)Other non-ESKAPEEc organisms21(3.4)2(2.9)19(3.4)Total625(100)69(11.0)556(89.0)Abbreviations: EOS, early-onset sepsis; LOS, late-onset sepsis Common bacteria isolated from neonatal bloodstream infection patients in China, 2015–2020 Abbreviations: EOS, early-onset sepsis; LOS, late-onset sepsis The proportion of ESKAPEEc pathogens among neonates with bloodstream infection from 2015 to 2020 The distribution of ESKAPEEc pathogens differed significantly between EOS and LOS. Among 69 EOS pathogens, ESKAPEEc accounted for 44.9% (31/69), and non-ESKAPEEc organisms comprised 55.1% (38/69). In neonates with EOS, the leading pathogens of ESKAPEEc and non-ESKAPEEc were E. coli (14/69; 20.3%) and S. epidermidis (15/69; 21.7%), respectively. In contrast, ESKAPEEc pathogens were significantly more prevalent in LOS (315/556; 56.6%), where K. pneumoniae (138/556; 24.8%) and E. coli (90/556; 16.2%) were the predominant agents; S. epidermidis (128/556; 23.0%) remained the most frequent non-ESKAPEEc organism. Specifically, the prevalence of K. pneumoniae was markedly higher in LOS than in EOS (24.8% vs. 8.7%, P = 0.003). Conversely, Group B Streptococcus (GBS) infection was far more common in EOS (15.9% vs. 2.2%, P < 0.001) (Table 2). Comparison of antibiotic resistance The proportion of species isolated with antimicrobial resistance is shown in Table 3. We next compared the resistance rates of Gram-positive species between ESKAPEEc and non-ESKAPEEc. Gram-positive ESKAPEEc and Gram-positive non-ESKAPEEc isolates showed relatively high resistance levels to penicillin (74.7% vs. 75.8%), amoxicillin/clavulanic acid (48.1% vs. 57.5%), erythromycin (57.0% vs. 72.2%), clindamycin (67.1% vs. 58.7%), and ciprofloxacin (43.0% vs. 34.9%). Whereas they showed opposite resistance patterns to oxacillin, moxifloxacin, levofloxacin, and trimethoprim/sulfamethoxazole. The resistant rates of Gram-positive ESKAPEEc to moxifloxacin (31.6% vs. 7.5%, < 0.001), levofloxacin (31.6% vs. 19.8%, P = 0.029) were significantly higher than the values of Gram-positive non-ESKAPEEc, while the resistant rates to oxacillin (19.0% vs. 61.1%, < 0.001), erythromycin (57.0% vs. 72.2%, P = 0.011), trimethoprim/sulfamethoxazole (12.7% vs. 33.7%, < 0.001) were significantly lower in Gram-positive ESKAPEEc group than those in Gram-positive non-ESKAPEEc group. In addition, only 1(0.3%) of 331 vancomycin-resistant coagulase-negative Staphylococcus (CoNS) was observed in this study (Table 3; Fig. 2). Table 3Resistance rates (%) of ESKAPEEc and non-ESKAPEEc in neonatal bloodstream infection, 2015–2020Antimicrobial agentAll G + species N(%)All G- species N(%)TotalG + ESKAPEEcG + non-ESKAPEEcχ2 P TotalG- ESKAPEEcG- non-ESKAPEEcχ2 P Penicillin250(75.5)59(74.7)191(75.8)0.0400.841–––Oxacillin169(51.1)15(19.0)154(61.1)42.708 < 0.001 –––Ampicillin–––205(69.7)194(72.7)11(40.7)11.835 < 0.001 Piperacillin–––118(40.1)115(43.1)3(11.1)10.424 0.001 Cefazolin–––213(72.4)192(71.9)21(77.8)0.4230.515Ceftriaxone–––146(49.7)141(52.8)5(18.5)11.533 < 0.001 Cefotaxime–––170(57.8)162(60.7)8(29.6)9.690 0.002 Ceftazidime–––118(40.1)112(41.9)6(22.2)19.615 < 0.001 Cefepime–––118(40.1)111(41.6)7(25.9)2.4990.114Amoxicillin/clavulanic acid183(55.3)38(48.1)145(57.5)2.1670.141Cefoperazone/sulbactam–––39(13.3)38(14.2)1(3.7)1.5360.215Piperacillin/tazobactam–––86(29.3)85(31.8)1(3.7)9.377 0.002 Erythromycin227(68.6)45(57.0)182(72.2)6.500 0.011 –––Clindamycin201(60.7)53(67.1)148(58.7)1.7620.184–––Gentamicin71(21.5)17(21.5)54(21.4)0.0000.98671(24.1)68(25.5)3(11.1)8.781 0.003 Moxifloxacin44(13.3)25(31.6)19(7.5)30.323 < 0.001 –––Ciprofloxacin122(36.7)34(43.0)88(34.9)1.7030.19268(23.1)68(25.5)0(0)16.927 < 0.001 Levofloxacin75(22.7)25(31.6)50(19.8)4.782 0.029 57(19.4)57(21.3)0(0)7.150 0.007 Trimethoprim/sulfamethoxazole95(28.7)10(12.7)85(33.7)13.051 < 0.001 94(32.0)94(35.2)0(0)13.973 < 0.001 Aztreonam–––115(39.1)109(40.8)6(22.2)3.5630.059Vancomycin1(0.3)01(0.4)0.0001.000–––Meropenem–––39(13.3)36(13.5)3(11.1)0.0020.961Ertapenem–––24(8.2)24(9.0)0(0)1.5800.209Imipenem–––92(31.3)85(31.8)7(25.9)0.3980.528Abbreviations: G-, Gram-negative. Boldface type means P < 0.05. N (%), number resistant (percentage resistant) Resistance rates (%) of ESKAPEEc and non-ESKAPEEc in neonatal bloodstream infection, 2015–2020 Abbreviations: G-, Gram-negative. Boldface type means P < 0.05. N (%), number resistant (percentage resistant) Antimicrobial resistance profiles of ESKAPEEc and non-ESKAPEEc pathogens in neonatal BSI. (a) Comparison resistance rates of the Gram-positive ESKAPEEc and non-ESKAPEEc; (b) Comparison resistance rates of the Gram-negative ESKAPEEc and non-ESKAPEEc Gram-negative species exhibited resistance rates higher than 40.0% to ampicillin, piperacillin, and cephalosporins, but the susceptibility rates to other antibiotics all exceeded 60.0%. The overall resistance rates of Gram-negative ESKAPEEc to multiple antibiotics except cefazolin were higher than the values of the Gram-negative non-ESKAPEEc, whereas no significant differences were observed in the resistance of isolates to carbapenems between the 2 groups (all p > 0.05). The resistance rates of ampicillin, piperacillin, ceftriaxone, cefotaxime, ceftazidime, piperacillin/tazobactam, aminoglycosides, fluoroquinolones, and trimethoprim/sulfamethoxazole were statistically different between Gram-negative ESKAPEEc and Gram-negative non-ESKAPEEc groups (all p < 0.05) (Table 3; Fig. 2). Comparison of the ESKAPEEc and non-ESKAPEEc BSI groups Demographic and clinical manifestations are summarized in Table 1. Compared with non-ESKAPEEc BSI, neonates with BSI due to ESKAPEEc were significantly younger (median, 4.5 [0.1–13.4] vs. 7.3 [1.7–17.5], P < 0.001), had lower birthweight (median, 2.6 [1.4–3.2] vs. 3.0 [2.0-3.4], P < 0.001) and had a higher rate of preterm birth (52.6% vs. 38.0%, P < 0.001). In contrast, the proportion of EOS did not differ significantly between the groups (9.0% vs. 13.6%, P = 0.065). Furthermore, there were no significant differences between the two groups regarding clinical manifestations, except that the incidence of sluggish reflexes was significantly higher in the ESKAPEEc BSI group (57.8% vs. 46.2%, P = 0.004) (Table 1). In terms of outcome of the study population, we found that 29 of the 625 neonates (4.6%) died before or after being discharged from the hospital: 22 (22/346; 6.4%) were in the ESKAPEEc BSI patient group, and 7 (7/279; 2.5%) were in the non-ESKAPEEc BSI group. K. pneumoniae infection (12/29; 41.4%) was the leading cause of mortality, followed by E. coli (5/29; 17.2%), E. faecium (4/29; 13.8%), and other pathogens (8/29; 27.6%). Neonates with ESKAPEEc BSI had a higher percentage of concurrent meningitis (17.9% vs. 9.3%, p = 0.002) and mortality rate (6.4% vs. 2.5%, p = 0.023) compared with those with non-ESKAPEEc BSI (Table 1). Risk factors for mortality in neonates with ESKAPEEc BSI Potential risk factors for mortality among neonates with ESKAPEEc BSI are presented in Table 4. Twenty-two (6.4%) of the 346 neonates with ESKAPEEc BSI died in our study. Compared with survivors, deceased neonates had a higher prevalence of preterm birth (81.8% vs. 50.6%, P = 0.005) and apnea or shortness of breath (68.2% vs. 45.4%, P = 0.038). They also had significantly lower gestational age (median, 31.7 [28.6–35.7] vs. 36.8 [31.4–39.1], P < 0.001) and lower birthweight (median, 1.6 [1.2–2.4] vs. 2.6 [1.4–3.2], P = 0.003). In contrast, no significant differences were revealed in terms of gender, age at admission, mode of delivery, PROM, MSAF, EOS, or type of pathogenic infection between deceased and surviving groups (all p > 0.05). Table 4Analysis of risk factors for death in neonates with ESKAPEEc BSIVariableDied (n = 22)Alive (n = 324)P ValueGender Male, No. (%)14(63.6)199(61.4)0.836 female, No. (%)8(36.4)125(38.6)Age at admission (days), Median(IQR)1.0(0-8.3)4.8(0.1–13.8)0.137Birthweight (Kg), Median(IQR)1.6(1.2–2.4)2.6(1.4–3.2) 0.003 Low birth weight < 2.5Kg, n (%)17(77.3)151(46.6) 0.005 Gestational age (weeks), Median(IQR)31.7(28.6–35.7)36.8(31.4–39.1) < 0.001 Preterm, < 37 weeks’ gestation, No. (%)18(81.8)164(50.6) 0.005 Natural birth, No. (%)11(50.0)156(48.1)0.866PROM, No. (%)6(27.3)78(24.1)0.735MSAF, No. (%)3(13.6)37(11.4)1.000Clinical manifestation Fever, No. (%)6(27.3)93(28.7)0.886 Aggravated jaundice, No. (%)9(40.9)103(31.8)0.376 Sluggish reflexes, No. (%)16(72.7)184(56.8)0.143Apnea or shortness of breath, No. (%)15(68.2)147(45.4) 0.038 Ventosity or vomit, No. (%)7(31.8)61(18.8)0.228 Convulsion, No. (%)1(4.5)14(4.3)1.000EOS, No. (%)3(13.6)28(8.6)0.683Concurrent meningitis, No. (%)1(4.5)61(18.8)0.161ESKAPEEc Enterococcus faecium 4(18.2)51(15.7)0.999 Staphylococcus aureus 0(0)24(7.4)0.374 Klebsiella pneumoniae 12(54.5)132(40.7)0.204 Acinetobacter baumannii 0(0)5(1.5)1.000Enterobacter spp1(4.5)13(4.0)1.000 Escherichia coli 5(22.7)99(30.6)0.438Abbreviations: PROM, Premature rupture of membranes; MSAF, meconium-stained amniotic fluid; EOS, early-onset sepsis; IQR, interquartile range. Boldface type means P < 0.05 Analysis of risk factors for death in neonates with ESKAPEEc BSI Abbreviations: PROM, Premature rupture of membranes; MSAF, meconium-stained amniotic fluid; EOS, early-onset sepsis; IQR, interquartile range. Boldface type means P < 0.05 Discussion BSI has been recognized as one of the biggest threats to children’s health worldwide, especially in neonates. A majority of BSIs were ascribed to the infection of ESKAPEEc pathogens, which posed substantial therapeutic challenges and an economic burden. Information on neonatal BSI caused by ESKAPEEc was indeed limited. Our 6-year study on neonatal BSI in China showed a dominance of ESKAPEEc pathogens. K. pneumoniae and E. coli were the predominant ESKAPEEc causing neonatal BSI. The distribution of ESKAPEEc bacteria differed by timing of infection onset. The resistance rates of ESKAPEEc to multiple antibiotics were of serious concern. In addition, it also revealed that ESKAPEEc infections occurred more frequently among preterm neonates or those with lower birthweight. Neonates with ESKAPEEc BSI also had higher rates of concurrent meningitis and mortality. In this study, ESKAPEEc accounted for 55.4% (346/625) of BSI, which was in accordance with previous data (58.7–61.7%) [6, 14], but not in line with studies from Sichuan, China (76%) [5] and Romania (97%) [15]. In addition, we also noticed that the predominant ESKAPEEc pathogens causing neonatal BSI were K. pneumoniae (144/625; 23.0%), E. coli (104/625; 16.6%). The finding was closely aligned with those of previous Chinese studies, which revealed that E. coli and K. pneumoniae were the leading ESKAPEEc causative agents in both adult BSIs from Southwest China [6] and in neonatal sepsis from East China [7]. However, it differed from surveillance data from South Africa [16], which demonstrated that K. pneumoniae (233/907; 25.7%), A. baumannii (174/907; 19.2%), and S. aureus (82/907; 9.0%) were the most common ESKAPEEc organisms causing neonatal BSI. Collectively, these discrepancies highlight that the distribution of pathogens causing BSI in neonates varied across regions and the study population. Therefore, more attention should be paid to monitoring the epidemiology of ESKAPEEc pathogens, especially K. pneumoniae. We found a distinct difference in the distribution of ESKAPEEc pathogens based on the timing of infection onset. Non-ESKAPEEc pathogens (38/69; 55.1%) dominated in EOS, whereas ESKAPEEc organisms (315/556; 56.6%) were relatively more frequent in LOS. The predominance of ESKAPEEc in LOS is likely because advances in medical technology have improved the survival of newborns (especially preterm and low-birthweight infants), resulting in prolonged hospital stays. Furthermore, the increasing use of invasive procedures and broad-spectrum antibiotics has not only elevated the risk of healthcare-associated infections but also promoted the emergence of multidrug-resistant bacteria. In EOS, E. coli (14/69; 20.3%) was the leading ESKAPEEc pathogen, while S. epidermidis (15/69; 21.7%) and GBS (11/69; 15.9%) were predominant non-ESKAPEEc organisms. Comparable findings were reported in another study from China [17]. The different available evidence was found in surveys from Tanzania [9] and South Africa [16, 18]. Although GBS infection was significantly more common in EOS than in LOS (15.9% vs. 2.2%, P < 0.001) (Table 2), its overall importance in EOS may be declining in some low- and middle-income settings due to improved preventive measures and microbiology monitoring [18]. In addition, our study also found the dominant pathogen in patients with LOS was K. pneumoniae (138/556; 24.8%). This finding was similar to those of previous studies conducted in mainland China [19], Egypt [20], and South Africa [16], but differed from those in Korea [21] and Tanzania [9]. Given these findings, empirical antibiotic therapy for EOS and LOS must be tailored to local pathogen profiles, accounting for the high prevalence and diversity of ESKAPEEc pathogens. Both Gram-positive ESKAPEEc and Gram-positive non-ESKAPEEc isolates exhibited high resistance to penicillin (74.7% vs. 75.8%), amoxicillin/clavulanic acid (48.1% vs. 57.5%), erythromycin (57.0% vs. 72.2%), and clindamycin (67.1% vs. 58.7%). In a Tanzanian study, the resistance rate to erythromycin (86.0%) among Gram-positive species was quite close to the surveillance data in our study, but the rates to penicillin (92.9%) and clindamycin resistance (39.9%) did not coincide with the proportions we reported [22]. According to a multicenter surveillance study in China, Gram-positive ESKAPEEc resistance to erythromycin (approximately 60%) was comparable to our finding, whereas the resistance rates to penicillin (>80%), clindamycin (approximately 20%) were not consistent with our data [23]. The resistance rates of Gram-positive ESKAPEEc to fluoroquinolones were higher than those of the Gram-positive non-ESKAPEEc bacteria in our study. Similar to this result, Marturano et al. [4] found that 39.2% of Gram-positive ESKAPEEc and 26.8% of Gram-positive non-ESKAPEEc isolates were resistant to levofloxacin in the US from 2014 to 2018. Of particular concern was the detection of one vancomycin-resistant CoNS isolate (0.3%) in our cohort. A study from six lower-tier South African hospitals during 2019–2020 also demonstrated that the resistance rates to linezolid and vancomycin among Gram-positive organisms were 0.5% and 1.2%, respectively [16]. Similarly, a survey in North India documented a sharp rise in linezolid-resistant Enterococcus spp. (from 0.0% to 11.0%) and a high prevalence of vancomycin-resistant Enterococcus spp. (increasing from 31.0% to 40.0%) between 2018 and 2022 [24]. Though vancomycin is still regarded as one of the final lines of defense against Gram-positive isolates, infection control measures are urgently needed to control the spread of these organisms. The high prevalence of antimicrobial resistance in Gram-negative species, particularly multidrug-resistant bacteria, has aroused global concern [25, 26]. In our study, resistance rates among Gram-negative species were 40.1%–57.8% for third- and fourth-generation cephalosporins, 29.3% for piperacillin/tazobactam, and 8.2%–31.3% for carbapenems. Notably, a multicenter South African study found high resistance rates in Gram-negative organisms to third-generation cephalosporins (66.2%), piperacillin/tazobactam (31.6%), and carbapenems (38.9%), all of which were higher than our corresponding findings [16]. In an Indian study, the resistance rates among Gram-negative species to third-generation cephalosporins (41.6–80.5%), piperacillin/tazobactam (13.7–66.2%), and carbapenems (20.4–63.2%) differed from those in our study [24]. Separately, research from Beijing children’s hospitals in China reported higher carbapenem resistance (10.4–54.5%) than our study, despite similar rates of third-generation cephalosporin resistance (52.5–55.8%) [11]. Additionally, our research found that the resistance rates of Gram-negative ESKAPEEc to third-generation, piperacillin/tazobactam, aminoglycosides, and fluoroquinolones were statistically higher than those of the Gram-negative non-ESKAPEEc (all p < 0.05). This contrasted with the findings of Marturano et al. [4]. The discrepancy might be because ESKAPEEc infections are associated with a poorer prognosis [4, 27], potentially prompting more frequent use of these antibiotics and thereby exerting selective pressure that drives higher resistance. Although a higher resistance rate to carbapenems was observed in Gram-negative ESKAPEEc compared to Gram-negative non-ESKAPEEc, this difference was not statistically significant (P > 0.05). This lack of significance may be partially attributable to the study population, single-center design, and limited sample size. To our knowledge, comprehensive data on the clinical aspects of ESKAPEEc BSI in neonates are lacking. This study found that low birthweight and preterm birth were significantly associated with ESKAPEEc infection and ESKAPEEc-related mortality ( all P < 0.05). There was only one study among the pediatric population (aged under 14 years old) comparing clinical characteristics of the ESKAPEEc and non-ESKAPEEc BSI [5]. Similar to research performed by Peng X, et al., preterm birth and/or low birth weight were risk factors for ESKAPEEc BSI acquisition [5]. This association might be explained by the weaker immune defenses in these infants, coupled with their frequent exposure to invasive procedures, broad-spectrum antibiotics, and extended hospitalization. Moreover, 4.6% (29/625) of neonates with BSI died; neonates with ESKAPEEc BSI had higher rates of concurrent meningitis and mortality in our study. This finding diverged from a previous Chinese study, which reported a mortality rate of 14.0% in pediatric BSI patients and no significant difference in mortality between the ESKAPEEc and non-ESKAPEEc BSI groups [5]. Conversely, another study demonstrated that patients who had ESKAPEEc infection were more likely to die [4]. Previous studies also showed that case-fatality rates in adult populations with underlying diseases were more frequent with ESKAPEEc BSI [28] or multidrug-resistant ESKAPEEc infection [29, 30]. Our study focused on BSI in neonates caused by ESKAPEEc pathogens, which is unique. We collected clinical and microbiological data of ESKAPEEc BSI in neonates. These results provided a foundation for empirical antibiotic selection. However, our study had several limitations. First, only neonatal BSI was involved in our study, and we could not assess age-stratified epidemiological differences in ESKAPEEc pathogens. Moreover, the retrospective design, which relied on medical records, led to inevitable data incompleteness and precluded a robust analysis of the impact of prior antibiotic exposure on ESKAPEEc infections. Future prospective studies are needed to accurately quantify prior antibiotic use to better elucidate its role in neonatal ESKAPEEc BSI. Conclusions Our study indicated that over half of neonates presenting with BSI were attributed to ESKAPEEc pathogens. K. pneumoniae and E. coli were the dominant ESKAPEEc pathogens responsible for neonatal BSI. The distribution of ESKAPEEc organisms in neonates with BSI differed by timing of infection onset. We also observed that the Gram-negative ESKAPEEc exhibited relatively high resistance rates to multiple antibiotics. Implementing the surveillance and timely reporting of AMR could enable earlier and targeted antibiotic selection against ESKAPE and thus alleviate the disease burden. In addition, it also revealed that ESKAPEEc infections and ESKAPEEc-related mortality occurred more frequently among preterm neonates or those with lower birthweight. Acknowledgements Not applicable. Abbreviations Bloodstream infection Late-onset sepsis Early-onset sepsis Coagulase-negative Staphylococcus Antimicrobial resistance The Clinical and Laboratory Standards Institute Interquartile range Premature rupture of membranes Meconium-stained amniotic fluid Group B Streptococcus Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, Enterobacter spp and Escherichia coli Author contributions Xiaoshan Peng, Hongmei Xu and Ruiqiu Zhao designed this research. Biao chen, Xiaoshan Peng, Chunmei Jing and Hong Wei collected the data. Xiaoshan Peng interpreted the data and wrote the original draft. Hongmei Xu and Ruiqiu Zhao revised and edited the manuscript. Funding This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Data availability The datasets that approved the results in the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate The study protocol was in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the Children’s Hospital of Chongqing Medical University [(2023) Annual Luncheon Review (Research) No.91]. Because of the retrospective design of the current study, the data did not involve personally identifiable information and informed consent was waived. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets that approved the results in the current study are available from the corresponding author on reasonable request.