Phenotypic and molecular characterization of Klebsiella pneumoniae in neonates and pediatric patients: antibiotic resistance, β-Lactamase genes, and sociodemographic risk factors Authors: Hakeem Zada, Murad A. Mubaraki, Muhammad Qasim, Mubbashir Hussain, Fozia Fozia, Hassan Naveed, Abdul Rehman, Ijaz Ahmad, Nawab Ali Published: 2025 ABSTRACT BackgroundNeonates, infants, and pediatric patients are usually prone to Klebsiella pneumoniae infections. Antimicrobial resistance is a major cause and needs effective surveillance because of the emergence of multidrug-resistant (MDR) and extensively drug-resistant (XDR) strains. The current study was conducted to evaluate K. pneumoniae prevalence, antimicrobial resistance patterns, and detect β-Lactamase and carbapenemase genes among clinical samples from selected major cities of Khyber Pakhtunkhwa (KP), Pakistan.MethodsClinical samples (2100) from hospitalized patients were collected at different selected tertiary care and district-level hospitals of Khyber Pakhtunkhwa (KP), Pakistan. Antibiotic susceptibility testing and phenotypic assays were performed for the determination of resistance profiles and detection of ESBL and carbapenemase. Further molecular characterization was performed by PCR of ESBL and carbapenemase genes. An analysis of clinical and sociodemographic data of neonates, infants, and pediatric patients was conducted for the identification of potential major risk factors contributing to K. pneumoniae infections.ResultsThe overall prevalence of K. pneumoniae among neonates, infants, and pediatric patients across hospitals in Khyber Pakhtunkhwa was 17.6% (371/2100). Blood was the most frequent infection source (10%), followed by urine (3.3%) and sputum (2.2%). The highest prevalence occurred in neonates (22%), compared to infants (15.9%) and pediatric patients (14.9%). Among the isolates, MDR-ESBL phenotypes were 29%, MDR-carbapenemase were 22.6%, and XDR (6.1%). The highest susceptibility was to amikacin (91% in non-MDR, 42% in MDR, and 26% in XDR isolates. Molecular analysis revealed blaCTX-M-14 (17.2%) and blaSHV (15.6%) as the most prevalent ESBL genes, and blaOXA-48 (5.1%) as the predominant carbapenemase gene. Significant risk factors included neonatal age (AOR: 2.10, 95% CI: 1.05–4.23, p = 0.033), prior antibiotic use (AOR: 2.95, 95% CI: 1.87–4.68, p = 0.0002), and hospital stay ≥7 days (AOR: 2.00, 95% CI: 1.35–3.03, p = 0.002). Breastfeeding had a protective effect, significantly reducing infection risk (OR: 0.50; 95% CI: 0.28–0.89; p = 0.022).ConclusionThe rising prevalence of MDR and XDR K. pneumoniae poses a critical threat to neonatal and pediatric care in Khyber Pakhtunkhwa. Strengthening antibiotic stewardship, enforcing strict infection control practices in pediatric units, and integrating continuous surveillance into public health policy are essential to curb the spread of resistant strains.Supplementary informationThe online version contains supplementary material available at 10.1186/s12879-025-12157-x. Introduction Klebsiella pneumoniae is a major pathogen associated with urinary tract infections, respiratory tract infections, surgical site infections, septicemia, and pneumonia [1]. K. pneumoniae belongs to the Enterobacteriaceae family and is among pathogens known for evading multiple antibiotic therapies. K. pneumoniae resistant to third-generation cephalosporin and carbapenem-resistant strains are categorized by the World Health Organization (WHO) as a global priority [2]. Its prevalence is a major concern after categorization, specifically in neonates, infants, and pediatric patients who are usually more vulnerable. Susceptibility to K. pneumoniae infections in neonates and infants is usually due to factors including invasive procedures, prolonged hospital stays, an underdeveloped immune system, and frequent antibiotic exposure [3]. K. pneumoniae is a major contributor to meningitis, sepsis, pneumonia, severe morbidity, and mortality in pediatric and neonatal care units [4]. In Pakistan, a high prevalence of K. pneumoniae has been reported in neonatal intensive care units (NICUs), where multidrug-resistant (MDR) strains are attributed to significant mortality rates [5]. Globally, the trends are consistently observed with approximately 5–15% of all bloodstream infections by K. pneumoniae in infants and neonates in resource-limited areas [6–9]. Resistance to multiple antibiotic classes by MDR strains, including β-lactams, aminoglycosides, and fluoroquinolones, causes challenges in treatments [10]. It is further worsened by strains that compromise the efficacy of frontline antibiotics, possessing extended-spectrum β-lactamases (ESBLs) and carbapenemases. Carbapenemase enzymes are primarily the cause of Carbapenem resistance in K. pneumoniae, including class B metallo-β-lactamases (e.g., NDM-1), class A β-lactamases (e.g., KPC), and class D β-lactamases (e.g., OXA-48-like enzymes). Plasmid-mediated dissemination of the blaNDM-1 gene is of major concern because of its role in XDR strains. Joint consensus document of the Centers for Disease Control and Prevention (CDC) and European Centre for Disease Prevention and Control (ECDC) defines XDR isolates as non-susceptible to at least one agent in all but two or fewer antimicrobial categories. Resistance to available antimicrobial agents is exhibited by these strains, further complicating therapeutic management [11]. The presence of several resistance determinants along with the presence of β-lactamases mostly results in multidrug resistance (MDR) or even pan-drug resistance (PDR), posing a significant threat in pediatric and neonatal intensive care units. Effective clinical management is hindered by these resistance mechanisms, but it can also be further escalated by prolonged hospitalizations, healthcare expenditures, and resource-constrained settings can worsen patient outcomes [12, 13]. K. pneumoniae infections can be deciphered by molecular characterization of resistance genes for their evolutionary and epidemiological dynamics. These analyses provide vital information about virulent attributes, genetic diversity, and the design of targeted antimicrobial stewardship initiatives. K. pneumoniae has an extensive burden globally, but there is a data scarcity of clinical isolates genetic profiling from Khyber Pakhtunkhwa (KP), Pakistan [14, 15]. In neonates and pediatric patients, there is a gap particularly related to MDR and XDR infections. The present study aims to fill the existing gap by investigating K. pneumoniae isolates from clinical specimens collected in tertiary care hospitals across Khyber Pakhtunkhwa, Pakistan. To our knowledge, this is the first comprehensive analysis in the region that integrates molecular characterization, antimicrobial resistance profiling, and assessment of sociodemographic risk factors among neonatal and pediatric patients of Khyber Pakhtunkhwa, Pakistan. This study will generate crucial data on the epidemiological and genetic features of K. pneumoniae, supporting the development of targeted clinical management strategies and public health interventions. The findings are expected to have significant implications for reducing the antimicrobial resistance burden and improving health outcomes among neonates and children in Khyber Pakhtunkhwa. Materials and methods Ethical considerations and study settings The present study was granted ethical approval by Kohat University of Science and Technology (KUST) Ethical Committee under reference No. Ref. No,/KUST/Ethical Committee/395 dated 06 July 2022. Parents or legal guardians provided the written informed consent, and confidentiality of the data was maintained according to ethical standards. Sample collection was done from January to December 2023 from selected hospitals (tertiary care centers and district-level facilities) of Khyber Pakhtunkhwa (KP), Pakistan (Fig. 1). In Peshawar, samples were collected from Lady Reading Hospital (LRH), Hayatabad Medical Complex (HMC), and Khyber Teaching Hospital (KTH), equipped with PICUs, NICUs, and general pediatric wards. District Headquarters Hospitals (DHQs) in Kohat, Mardan, and Swat were selected for additional sampling.Fig. 1Study districts of Khyber Pakhtunkwa (Peshawar, Swat, Mardan, Kohat) Study districts of Khyber Pakhtunkwa (Peshawar, Swat, Mardan, Kohat) Study population The study included neonates, infants, and children admitted to the selected secondary and tertiary care hospitals. Clinical samples were collected from patients diagnosed with K. pneumoniae infections, including those with urinary tract infections (UTIs), respiratory tract infections (RTIs), sepsis, and pneumonia. Samples were collected from both symptomatic and asymptomatic carriers for surveillance purposes. Inclusion criteria Neonates (0–28 days), infants (1 month to 1 year), and children (1–16 years) [16–18].Patients with confirmed or suspected K. pneumoniae infections (from blood, urine, sputum, pus, or other body fluids) Neonates (0–28 days), infants (1 month to 1 year), and children (1–16 years) [16–18]. Neonates (0–28 days), infants (1 month to 1 year), and children (1–16 years) [16–18]. Patients with confirmed or suspected K. pneumoniae infections (from blood, urine, sputum, pus, or other body fluids) Patients with confirmed or suspected K. pneumoniae infections (from blood, urine, sputum, pus, or other body fluids) Exclusion criteria Patients with infections caused by pathogens other than K. pneumoniaeImmunocompromised children with conditions like cancer or transplant historyPatients receiving antibiotics before admission or having a history of prior K. pneumoniae infections within the last six monthsNeonates with congenital abnormalities or severe comorbid conditions that would affect the study’s outcome. Patients with infections caused by pathogens other than K. pneumoniae Patients with infections caused by pathogens other than K. pneumoniae Immunocompromised children with conditions like cancer or transplant history Immunocompromised children with conditions like cancer or transplant history Patients receiving antibiotics before admission or having a history of prior K. pneumoniae infections within the last six months Patients receiving antibiotics before admission or having a history of prior K. pneumoniae infections within the last six months Neonates with congenital abnormalities or severe comorbid conditions that would affect the study’s outcome. Neonates with congenital abnormalities or severe comorbid conditions that would affect the study’s outcome. Sociodemographic data collection Patient sociodemographic data was obtained through medical record reviews and structured interviews with parents or legal guardians. Using a predesigned questionnaire, we collected information on the child’s age group (neonate, infant, or child), gender, ethnicity, and residence status (local or Afghan refugee). Additional variables included prior antibiotic use, history of hospitalization, length of hospital stay, underlying medical conditions (e.g., prematurity, low birth weight, congenital anomalies), and exposure to invasive procedures (e.g., intubation, catheterization, surgery). Nutritional indicators, such as breastfeeding status and birth weight, were also recorded. These factors were selected due to their established or potential association with K. pneumoniae infections in neonates and children. To maintain confidentiality, all data were anonymized and coded before analysis. The collected variables were assessed for potential associations with infection rates and antimicrobial resistance, providing critical insights into the epidemiological and clinical characteristics of K. pneumoniae infections in this vulnerable population. Sampling procedures Clinical samples (Urine, Blood, Sputum, Pus, and Tracheal aspirates) based on the suspected infection site were collected. Standard aseptic techniques were used for sample collection to avoid contamination. Suprapubic aspiration or sterile bags were used for urine collection, and peripheral venipuncture was done for blood collection. Sputum and tracheal aspirates were collected from intubated or ventilated patients, while older children provided expectorated samples using sterile suction. Nasopharyngeal aspirates were also obtained in respiratory cases. For pediatric patients (1–16 years), blood was collected via venipuncture, and urine via the clean-catch method or catheterization. Pus was obtained aseptically from surgical sites or wounds, and sputum via expectoration or endotracheal tubes when required. All specimens were processed at the Medical Microbiology Laboratory, Department of Microbiology, KUST, Kohat, and the Mubarak Diagnostic Laboratory & Research Center, Peshawar. Culture and identification of K. pneumoniae Specimens were inoculated onto MacConkey agar (Merck, Germany) and incubated at 37 °C for 18–24 hours. Colonies with characteristic morphology were selected and identified using standard microbiological and biochemical tests. Confirmation of K. pneumoniae was performed using the API 20E identification kit (bioMérieux), following the manufacturer’s instructions. All confirmed isolates were stored in 25% glycerol at − 70 °C for subsequent analyses. Only confirmed K. pneumoniae isolates were included in the final analysis. Analysis of antibiotic susceptibility patterns Antibiotic susceptibility was assessed using the Kirby-Bauer disk diffusion method on Mueller–Hinton agar (MHA) (Merck, Germany), following CLSI guidelines (2023) [19]. The panel included 12 antibiotics commonly used against K. pneumoniae, Amoxicillin-clavulanic acid (AMC, 30 µg), Cefepime (CEF, 30 µg), Ampicillin (AMP, 10 µg), Cefixime (CFM, 10 µg), Ciprofloxacin (CIP, 5 µg), Levofloxacin (LEV, 5 µg), Norfloxacin (NOR, 10 µg), Trimethoprim-sulfamethoxazole (SXT, 1.25/23.75 µg), Amikacin (AMK, 30 µg), Erythromycin (E, 15 µg), Nalidixic acid (NA, 30 µg), Imipenem (IPM, 10 µg), and Meropenem (MER, 10 µg). Multidrug-resistant (MDR) isolates were defined according to European Centre for Disease Prevention and Control (ECDC) criteria. Control strains K. pneumoniae (ATCC 700603) and E. coli (ATCC BAA-2326), from the Department of Microbiology, KUST, were used to validate susceptibility testing. Minimum Inhibitory Concentration (MIC) testing Minimum inhibitory concentration (MIC) testing was performed for K. pneumoniae isolates using the broth microdilution method, under the Clinical and Laboratory Standards Institute (CLSI) guidelines [19]. Standardized 0.5 McFarland bacterial suspensions were inoculated into Mueller-Hinton broth containing serial two-fold dilutions of imipenem, meropenem, and ertapenem. Plates were incubated at 35 ± 2 °C for 18–24 hours, and MIC values were recorded as the lowest antibiotic concentration that inhibited visible bacterial growth. Results were interpreted using CLSI breakpoints. Phenotypic detection of β-lactamases Klebsiella pneumoniae isolates were initially screened for ESBL production based on resistance to ceftazidime. Screen-positive strains were further tested using the double-disc synergy test (DDST) and the CLSI-recommended confirmatory method. For DDST, an amoxicillin-clavulanic acid (AMC) disc was placed at the center of a 90 mm Mueller–Hinton agar plate, with ceftazidime (CAZ, 30 μg) and cefotaxime (CTX, 30 μg) discs positioned 20 mm apart from the AMC disc. A keyhole-shaped enhancement of inhibition between the β-lactam and AMC discs indicated ESBL production, confirming synergy with the β-lactamase inhibitor. Phenotypic detection of carbapenemase production Isolates were initially screened for carbapenem resistance using the Kirby-Bauer disk diffusion method. Carbapenem-resistant isolates were further evaluated for carbapenemase production using the EDTA-Disc Synergy Test (EDTA-DST) and the Modified Carbapenem Inactivation Method (mCIM), following CLSI guidelines. Molecular analysis of antibiotic resistance genes Genes encoding β-lactamases (blaTEM, blaSHV, blaCTX-M, blaCTX-M-15, blaOXA, and carbapenemases (blaKPC, blaOXA-48, blaNDM, and blaIMP) were amplified by PCR using primers described by Bhaskar et al. (2019) [20]. PCR amplification was performed in a 25 μL reaction using standard conditions as described previously by Bhaskar et al. 2019, with specific primers targeting resistance genes. Reactions were run in a ProFlex thermal cycler (Bio-Rad, USA), and amplicons were visualized on 1.5% agarose gels under UV illumination. Statistical analysis Descriptive statistics were used to calculate the frequency and distribution of K. pneumoniae infections, resistance profiles, and sociodemographic characteristics. Chi-square tests were applied to assess the association between sociodemographic factors and the prevalence of MDR and XDR K. pneumoniae isolates. Bivariate and multivariable logistic regression analyses were performed to calculate the crude odds ratio (COR) and adjusted odds ratio (AOR). Data were analyzed using GraphPad Prism version 8.0.2 (GraphPad Software, San Diego, CA, USA). A p-value of less than 0.05 was considered statistically significant. Results Clinical distribution of Klebsiella pneumoniae isolates A total of 2,100 clinical specimens, including blood, urine, sputum, and pus, were analyzed, of which 371 (17.6%) were positive for K. pneumoniae. Among these, 68% were non-multidrug-resistant (non-MDR), 24% were multidrug-resistant (MDR), and 7% were extensively drug-resistant (XDR). Blood was the most common source of isolates, followed by urine. The highest proportion of MDR strains (13.7%) was found in blood, while pus contributed the lowest (2.96%). XDR isolates were detected only in blood and urine; no XDR strains were identified in sputum or pus samples from neonates, although MDR strains were present in sputum (4.3%) and pus (2.96%). In infants, blood and urine accounted for 15.9% and 6.4% of total isolates, respectively. Among blood isolates, non-MDR strains comprised 9.9%, while MDR and XDR strains accounted for 4.3% and 1.6%, respectively. Urine isolates in this group showed similar trends, with MDR at 1.34% and XDR at 0.53%. Among pediatric patients, blood and urine were the predominant sources, together comprising 48% of isolates. MDR prevalence reached 18%, indicating a substantial resistance burden, with XDR strains at 6.6%. Blood isolates showed a high proportion of non-MDR strains (54.2%), while no XDR isolates were detected in sputum or pus in this group. The distribution of isolates by age group and specimen type is illustrated in Fig. 2.Fig. 2Percentage distribution of Klebsiella pneumoniae isolates by age group, resistance pattern, and specimen type. The x-axis represents specimen types, while the y-axis shows the percentage of positive isolates. Bars are color-coded by resistance category and age group: non-MDR, MDR, and XDR across neonates, infants, and pediatric patients Percentage distribution of Klebsiella pneumoniae isolates by age group, resistance pattern, and specimen type. The x-axis represents specimen types, while the y-axis shows the percentage of positive isolates. Bars are color-coded by resistance category and age group: non-MDR, MDR, and XDR across neonates, infants, and pediatric patients Prevalence was calculated by specimen type and age group, and statistical associations were assessed using the Chi-square test. In neonates, the highest detection rate was observed in urine samples (29.17%, 28/96), though this was not statistically significant (p = 0.15). A significant association was found in blood samples (24.50%, 98/400; p = 0.02). Lower detection rates occurred in sputum (17.00%, 17/100; p = 0.47) and pus (11.82%, 13/110; p = 0.60), both non-significant. Among infants, urine again showed the highest prevalence (26.67%, 24/90), followed by blood (14.75%, 59/400; p = 0.01), sputum (14.55%, 16/110), and pus (12.22%, 11/90). Only the blood samples showed a statistically significant association (p < 0.05). In pediatric patients, the highest rate was recorded in urine (19.15%, 18/94; p = 0.03), followed by blood (14.96%, 57/381; p = 0.05), sputum (17.05%, 15/88), and pus (10.64%, 15/141). The associations in sputum and pus samples were not statistically significant (p > 0.05). Overall, K. pneumoniae prevalence varied across age groups and specimen types. Blood and urine consistently yielded the highest number of positive isolates. Significant associations were primarily observed in neonatal and infant blood samples, as well as in pediatric urine samples, highlighting their diagnostic relevance for clinical surveillance and targeted infection control. Antibiotic susceptibility profiles of Klebsiella pneumoniae isolates Among the 371 K. pneumoniae isolates, non-MDR strains accounted for 42% (n = 156), MDR for 51.7% (n = 192), and XDR for 6.1% (n = 23). β-lactam resistance was widespread. High resistance to ampicillin (AMP) was observed across all groups: 77% in non-MDR, 93% in MDR, and 100% in XDR isolates. A similar pattern was noted for amoxicillin-clavulanic acid (AMC), with resistance rates of 79%, 94%, and 100%, respectively. Cefixime (CFM) and cefepime (FEP) resistance exceeded 85% and 90% in MDR and XDR isolates, while even non-MDR strains showed considerable resistance to CFM (73%) and FEP (68%). Fluoroquinolone resistance was also prominent. Ciprofloxacin (CIP) and levofloxacin (LEV) resistance exceeded 60% across all categories, reaching the highest levels in MDR (88% and 87%) and XDR (82% each) isolates. Trimethoprim-sulfamethoxazole (SXT) resistance followed a similar trend, 40% in non-MDR, rising sharply to 85% in MDR and 95% in XDR strains. Among aminoglycosides, amikacin (AMK) remained the most effective, with 91% susceptibility in non-MDR isolates. However, this decreased to 42% in MDR and only 26% in XDR isolates. Carbapenems showed the highest overall activity among β-lactams. Imipenem (IMP) and meropenem (MER) retained susceptibility rates of 74% and 76%, respectively, in non-MDR isolates. In contrast, resistance increased markedly in XDR isolates—73% (IMP) and 69% (MER), indicating the emergence of carbapenem resistance. Erythromycin (ERY) resistance was also high across all categories, particularly in MDR (84%) and XDR (95%) isolates, reducing its potential clinical utility. Antibiotic susceptibility patterns by resistance class are summarized in Fig. 3.Fig. 3Percent sensitivity of non-MDR, MDR, and XDR Klebsiella pneumoniae isolates to different antibiotic classes. The bar chart presents the proportion of isolates sensitive to representative agents from major drug classes: β-lactams (amoxicillin–clavulanic acid, cefepime, cefixime, ampicillin), fluoroquinolones (ciprofloxacin, levofloxacin, norfloxacin), aminoglycosides (amikacin), sulfonamides (trimethoprim–sulfamethoxazole), macrolides (erythromycin), quinolones (nalidixic acid), and carbapenems (imipenem, meropenem). Bars indicate non-MDR, MDR, and XDR isolate groups Percent sensitivity of non-MDR, MDR, and XDR Klebsiella pneumoniae isolates to different antibiotic classes. The bar chart presents the proportion of isolates sensitive to representative agents from major drug classes: β-lactams (amoxicillin–clavulanic acid, cefepime, cefixime, ampicillin), fluoroquinolones (ciprofloxacin, levofloxacin, norfloxacin), aminoglycosides (amikacin), sulfonamides (trimethoprim–sulfamethoxazole), macrolides (erythromycin), quinolones (nalidixic acid), and carbapenems (imipenem, meropenem). Bars indicate non-MDR, MDR, and XDR isolate groups Minimum Inhibitory Concentration (MIC) and susceptibility profiles The Minimum Inhibitory Concentration (MIC₅₀) of the isolates was evaluated and interpreted according to CLSI breakpoints (Table 1).Table 1Minimum inhibitory concentration (MIC) ranges, MIC₅₀, CLSI breakpoints, and percentage of susceptible, intermediate, and resistant clinical isolates for six tested antibioticsAntibioticMIC Range (µg/mL)MIC₅₀ (µg/mL)CLSI Breakpoints (S/I/R µg/mL)% Susceptible% Intermediate% ResistantImipenem (IPM)0.25–321≤1/2/≥456.80%5.30%37.70%Meropenem (MER)0.25–320.5≤1/2/≥461.70%8.00%30.10%Amikacin (AMK)1–642≤16/32/≥6444.70%3.20%52.00%Cefepime (FEP)0.5–644≤2/4/≥852.20%4.30%43.30%Ciprofloxacin (CIP)0.06–81≤0.25/0.5/≥143.90%6.10%49.80%Trimethoprim/Sulfamethoxazole (SXT)0.5/9.5– > 4/762≤2/38/–/≥4/7638.50%7.80%53.60% Table 1Minimum inhibitory concentration (MIC) ranges, MIC₅₀, CLSI breakpoints, and percentage of susceptible, intermediate, and resistant clinical isolates for six tested antibioticsAntibioticMIC Range (µg/mL)MIC₅₀ (µg/mL)CLSI Breakpoints (S/I/R µg/mL)% Susceptible% Intermediate% ResistantImipenem (IPM)0.25–321≤1/2/≥456.80%5.30%37.70%Meropenem (MER)0.25–320.5≤1/2/≥461.70%8.00%30.10%Amikacin (AMK)1–642≤16/32/≥6444.70%3.20%52.00%Cefepime (FEP)0.5–644≤2/4/≥852.20%4.30%43.30%Ciprofloxacin (CIP)0.06–81≤0.25/0.5/≥143.90%6.10%49.80%Trimethoprim/Sulfamethoxazole (SXT)0.5/9.5– > 4/762≤2/38/–/≥4/7638.50%7.80%53.60% Minimum inhibitory concentration (MIC) ranges, MIC₅₀, CLSI breakpoints, and percentage of susceptible, intermediate, and resistant clinical isolates for six tested antibiotics Among the carbapenems, imipenem exhibited a MIC₅₀ of 1 µg/mL, with 56.8% of isolates categorized as susceptible, 5.3% as intermediate (MIC = 2 µg/mL), and 37.7% as resistant (MIC ≥ 4 µg/mL). Meropenem showed a MIC₅₀ of 0.5 µg/mL, with a higher susceptibility rate of 61.7%. In contrast, 8.0% of isolates were intermediate (MIC = 2 µg/mL) and 30.1% were resistant (MIC ≥ 4 µg/mL). These findings suggest that meropenem demonstrated greater efficacy than imipenem against the tested isolates. For the aminoglycoside amikacin, the MIC₅₀ was 2 µg/mL. Susceptibility was observed in 44.7% of isolates, while 3.2% were intermediate (MIC = 32 µg/mL) and 52.0% were resistant (MIC ≥ 64 µg/mL), indicating reduced effectiveness. Among β-lactams, cefepime showed a MIC₅₀ of 4 µg/mL, with 52.2% of isolates susceptible, 4.3% intermediate (MIC = 4 µg/mL), and 43.3% resistant (MIC ≥ 8 µg/mL). For fluoroquinolones, ciprofloxacin demonstrated a MIC₅₀ of 1 µg/mL. A total of 43.9% of isolates were susceptible, 6.1% were in the intermediate range (MIC = 0.5 µg/mL), and 49.8% were resistant (MIC ≥ 1 µg/mL), reflecting a significant resistance burden. Trimethoprim-sulfamethoxazole (SXT) showed the highest resistance rate among the tested antibiotics. Only 38.5% of isolates were susceptible (MIC ≤ 2/38 µg/mL), while 53.6% were resistant (MIC ≥ 4/76 µg/mL). Overall, ciprofloxacin and SXT exhibited the highest resistance levels, whereas meropenem maintained the most favorable susceptibility profile. The distribution of resistant and intermediate isolates highlights the limited therapeutic options available for managing multidrug-resistant K. pneumoniae infections. Phenotypic carbapenemase and ESBL production Across 371 clinical isolates, MDR strains exhibited the highest rates of Extended-Spectrum Beta-Lactamase (ESBL) (29.1%) and carbapenemase (22.6%) production. Non-MDR isolates showed 42% ESBL production but no carbapenemase activity, while XDR isolates had the lowest rates (3.5% ESBL and 2.6% carbapenemase). The distribution of K. pneumoniae isolates categorized as non-MDR, MDR, and XDR according to ESBL and carbapenemase production across specimen types is summarized in Table 2, along with corresponding statistical significance.Table 2Prevalence of ESBL- and carbapenemase-producing isolates stratified by antimicrobial profile across clinical specimen sourcesClinical SpecimenTotal positiveNon-MDR ESBLsn (%)Non-MDR Carbapenemases n (%)MDR ESBLs n (%)MDR Carbapenemasesn (%)XDR ESBLs n (%)XDR Carbapenemases n (%)P valueBlood214108 (29.0%)0 (0%)51 (13.7%)44 (11.8%)6 (1.6%)5 (1.3%)0.002Urine7015 (4.0%)0 (0%)30 (8.0%)17 (4.5%)5 (1.3%)3 (0.8%)0.003Pus4817 (4.5%)0 (0%)16 (4.3%)13 (3.5%)1 (0.2%)1 (0.2%)0.426Sputum3916 (4.4%)0 (0%)11 (2.9%)10 (2.6%)1 (0.2%)1 (0.2%)0.812Total371156 (42.0%)0 (0%)108 (29.1%)84 (22.6%)13 (3.5%)10 (2.6%)– Table 2Prevalence of ESBL- and carbapenemase-producing isolates stratified by antimicrobial profile across clinical specimen sourcesClinical SpecimenTotal positiveNon-MDR ESBLsn (%)Non-MDR Carbapenemases n (%)MDR ESBLs n (%)MDR Carbapenemasesn (%)XDR ESBLs n (%)XDR Carbapenemases n (%)P valueBlood214108 (29.0%)0 (0%)51 (13.7%)44 (11.8%)6 (1.6%)5 (1.3%)0.002Urine7015 (4.0%)0 (0%)30 (8.0%)17 (4.5%)5 (1.3%)3 (0.8%)0.003Pus4817 (4.5%)0 (0%)16 (4.3%)13 (3.5%)1 (0.2%)1 (0.2%)0.426Sputum3916 (4.4%)0 (0%)11 (2.9%)10 (2.6%)1 (0.2%)1 (0.2%)0.812Total371156 (42.0%)0 (0%)108 (29.1%)84 (22.6%)13 (3.5%)10 (2.6%)– Prevalence of ESBL- and carbapenemase-producing isolates stratified by antimicrobial profile across clinical specimen sources In urine samples (n = 70 positive isolates), ESBL production was more frequent among MDR isolates (8%) compared to non-MDR (4%) and XDR (1.3%) isolates. Carbapenemase production was also higher in MDR isolates (4.5%), followed by XDR isolates (0.8%). These differences were statistically significant (p = 0.002). In blood samples (n = 214), ESBL production predominated in non-MDR isolates (29%) and MDR isolates (13%), whereas XDR isolates showed a lower rate (1.6%). The observed variations were statistically significant (p = 0.003). Among pus isolates (n = 48), ESBL production was nearly similar in non-MDR (4.5%) and MDR (4.3%) isolates, with XDR isolates showing only 0.2%. Carbapenemase production in pus was limited to MDR (3.5%) and XDR (0.2%) isolates. However, these differences were not statistically significant (p = 0.426). In sputum samples (n = 39), ESBL production occurred in 4.4% of non-MDR, 2.9% of MDR, and 0.2% of XDR isolates. Carbapenemase production was slightly higher in MDR (2.6%) compared to XDR (0.2%) isolates. These differences were also not statistically significant (p = 0.812). Overall, MDR isolates demonstrated the highest frequency of both ESBL and carbapenemase production, particularly in urine and blood samples, which showed statistically significant associations. These findings emphasize the predominance of MDR K. pneumoniae as a major reservoir of β-lactam resistance mechanisms in pediatric clinical infections. Molecular detection of ESBL and Carbapenemase genes Among the Extended-Spectrum Beta-Lactamase (ESBL) genes, blaSHV was the most prevalent, detected in 15.6% of the 371 isolates. Its occurrence was highest in non-MDR isolates (12.1%), followed by MDR (2.9%) and XDR (0.5%) isolates. Among the other ESBL genes analyzed (blaTEM, blaCTX-M-14, blaCTX-M-15, and blaOXA), blaCTX-M-14 (11%) was the most common in non-MDR isolates, followed by blaTEM (9.9%), blaCTX-M-15 (3.7%), and no detection of blaOXA. In MDR isolates, blaCTX-M-15 (4.3%) was the most frequent ESBL gene, while blaOXA (1.3%) was the least common. Among XDR isolates, blaCTX-M-15 (1.6%) remained the predominant gene, followed by blaOXA (0.2%). For carbapenemase genes, blaOXA-48 (4.3%) was the most frequently detected in MDR isolates, followed by blaNDM-1 (3.7%), blaKPC (1%), and blaIMP (0.8%). Among XDR isolates, blaOXA-48 (0.8%) was again the most prevalent, followed by blaNDM-1 (0.5%), blaKPC (0.2%), and blaIMP (0.2%). The amplicon sizes of different genes are shown in figure S supplementary file Figure 4 illustrates the distribution of K. pneumoniae isolates based on the presence of key ESBL and carbapenemase genes across non-MDR, MDR, and XDR categories.Fig. 4Distribution of ESBL and carbapenemase genes among non-MDR, MDR, and XDR isolates. (5A) percentage of isolates carrying ESBL genes, including blaSHV, blaTEM, blaCTX-M-14, blaCTX-M-15, and blaOXA. (5B) percentage of isolates carrying carbapenemase genes, including blaOXA-48, blaKPC, blaNDM-1, and blaIMP. The isolates are categorized based on resistance profiles: Non-MDR (checkered bars), MDR (gray bars), and XDR (red bars). Error bars represent the standard error of the mean (SEM) Distribution of ESBL and carbapenemase genes among non-MDR, MDR, and XDR isolates. (5A) percentage of isolates carrying ESBL genes, including blaSHV, blaTEM, blaCTX-M-14, blaCTX-M-15, and blaOXA. (5B) percentage of isolates carrying carbapenemase genes, including blaOXA-48, blaKPC, blaNDM-1, and blaIMP. The isolates are categorized based on resistance profiles: Non-MDR (checkered bars), MDR (gray bars), and XDR (red bars). Error bars represent the standard error of the mean (SEM) Sociodemographic factors associated with Klebsiella pneumoniae infections in neonates, infants, and pediatric patients Neonates and infants were found to be significantly more susceptible to Klebsiella pneumoniae infections. Neonates exhibited the highest odds of infection (adjusted OR: 2.10; 95% CI: 1.05–4.23; p = 0.033), followed by infants (adjusted OR: 1.75; 95% CI: 1.02–3.02; p = 0.041). Although children showed a moderately elevated odds ratio (adjusted OR: 1.50; 95% CI: 0.98–2.31), the association did not reach statistical significance (p = 0.062). Male gender was significantly associated with a higher infection risk (adjusted OR: 1.63; 95% CI: 1.10–2.41; p = 0.016), possibly reflecting differences in immune response or exposure risk, whereas female gender showed no significant association (adjusted OR: 1.10; 95% CI: 0.77–1.55; p = 0.621). Ethnicity also influenced infection susceptibility. Afghan refugees had a markedly higher infection risk (adjusted OR: 2.10; 95% CI: 1.33–3.28; p = 0.001), likely reflecting disparities in healthcare access, sanitation, and living conditions. In contrast, local ethnicity showed a modest but non-significant increase in odds (adjusted OR: 1.40; 95% CI: 0.98–2.00; p = 0.068). Prior antibiotic use emerged as a major risk factor, strongly associated with infection (adjusted OR: 2.95; 95% CI: 1.87–4.68; p = 0.0002), underscoring the impact of antibiotic misuse in resistance development. Conversely, those without previous antibiotic exposure demonstrated a significantly reduced risk (adjusted OR: 0.50; 95% CI: 0.32–0.79; p = 0.004). A history of hospital admission also increased infection likelihood (adjusted OR: 1.95; 95% CI: 1.30–2.92; p = 0.002), emphasizing the role of nosocomial transmission. Furthermore, a prolonged hospital stay (≥7 days) was associated with an elevated infection risk (adjusted OR: 2.00; 95% CI: 1.35–3.03; p = 0.002), likely reflecting cumulative exposure to hospital-acquired pathogens. Among underlying medical factors, prematurity and low birth weight significantly contributed to infection susceptibility. Premature infants showed higher odds of infection (adjusted OR: 1.75; 95% CI: 1.08–2.83; p = 0.022), while low birth weight further increased the risk (adjusted OR: 1.80; 95% CI: 1.16–2.77; p = 0.010). In contrast, breastfeeding had a protective effect, significantly reducing infection risk (adjusted OR: 0.50; 95% CI: 0.28–0.89; p = 0.022). Overall, these findings highlight age group, prior antibiotic exposure, and hospitalization as the key determinants of K. pneumoniae infection risk among pediatric patients. Table 3 summarizes the sociodemographic factors associated with infection across different age categories.Table 3Sociodemographic factors associated with K. pneumoniae infections in different age groups using univariate and multivariate regression analysis (p < 0.05)Sociodemographic FactorCategoryPositive n (%)Negative n (%)Univariate OR (95% CI)Univariate p-valueMultivariate OR (95% CI)Multivariate p-valueAge GroupNeonates43 (11.6)2057 (88.4)2.30 (1.14–4.63)0.0182.10 (1.05–4.23)0.033Infants58 (15.6)2042 (84.4)1.87 (1.12–3.14)0.0161.75 (1.02–3.02)0.041Children126 (33.9)1944 (66.1)1.65 (1.02–2.68)0.0401.50 (0.98–2.31)0.062GenderMale145 (39.1)226 (60.9)1.75 (1.22–2.48)0.0031.63 (1.10–2.41)0.016Female126 (33.9)245 (66.1)1.20 (0.87–1.66)0.2601.10 (0.77–1.55)0.621EthnicityLocal185 (49.9)86 (50.1)1.45 (1.02–2.06)0.0341.40 (0.98–2.00)0.068Afghan Refugee86 (23.2)285 (76.8)2.25 (1.45–3.51)0.00042.10 (1.33–3.28)0.001History of Antibiotic UseYes162 (43.7)109 (29.3)3.25 (2.12–5.00)0.00012.95 (1.87–4.68)0.0002No109 (29.3)162 (70.7)0.45 (0.29–0.69)0.00030.50 (0.32–0.79)0.004Previous Hospital AdmissionsYes148 (39.9)123 (33.1)2.10 (1.45–3.08)0.00021.95 (1.30–2.92)0.002No123 (33.1)148 (66.9)1.50 (1.03–2.17)0.0301.45 (0.99–2.11)0.061Length of Hospital Stay < 7 days75 (20.2)296 (79.8)1.50 (1.05–2.14)0.0291.40 (0.98–2.02)0.062≥7 days196 (52.8)175 (47.2)2.10 (1.43–3.11)0.0012.00 (1.35–3.03)0.002Underlying Medical ConditionsPrematurity69 (18.6)302 (81.4)1.80 (1.13–2.87)0.0161.75 (1.08–2.83)0.022Low Birth Weight92 (24.8)279 (75.2)1.95 (1.26–3.05)0.0031.80 (1.16–2.77)0.010Nutritional StatusBreastfeeding145 (39.1)226 (60.9)0.63 (0.45–0.87)0.0060.50 (0.28–0.89)0.022n = number, % = percentage, OR = odd ratio, CI = confidence interval Table 3Sociodemographic factors associated with K. pneumoniae infections in different age groups using univariate and multivariate regression analysis (p < 0.05)Sociodemographic FactorCategoryPositive n (%)Negative n (%)Univariate OR (95% CI)Univariate p-valueMultivariate OR (95% CI)Multivariate p-valueAge GroupNeonates43 (11.6)2057 (88.4)2.30 (1.14–4.63)0.0182.10 (1.05–4.23)0.033Infants58 (15.6)2042 (84.4)1.87 (1.12–3.14)0.0161.75 (1.02–3.02)0.041Children126 (33.9)1944 (66.1)1.65 (1.02–2.68)0.0401.50 (0.98–2.31)0.062GenderMale145 (39.1)226 (60.9)1.75 (1.22–2.48)0.0031.63 (1.10–2.41)0.016Female126 (33.9)245 (66.1)1.20 (0.87–1.66)0.2601.10 (0.77–1.55)0.621EthnicityLocal185 (49.9)86 (50.1)1.45 (1.02–2.06)0.0341.40 (0.98–2.00)0.068Afghan Refugee86 (23.2)285 (76.8)2.25 (1.45–3.51)0.00042.10 (1.33–3.28)0.001History of Antibiotic UseYes162 (43.7)109 (29.3)3.25 (2.12–5.00)0.00012.95 (1.87–4.68)0.0002No109 (29.3)162 (70.7)0.45 (0.29–0.69)0.00030.50 (0.32–0.79)0.004Previous Hospital AdmissionsYes148 (39.9)123 (33.1)2.10 (1.45–3.08)0.00021.95 (1.30–2.92)0.002No123 (33.1)148 (66.9)1.50 (1.03–2.17)0.0301.45 (0.99–2.11)0.061Length of Hospital Stay < 7 days75 (20.2)296 (79.8)1.50 (1.05–2.14)0.0291.40 (0.98–2.02)0.062≥7 days196 (52.8)175 (47.2)2.10 (1.43–3.11)0.0012.00 (1.35–3.03)0.002Underlying Medical ConditionsPrematurity69 (18.6)302 (81.4)1.80 (1.13–2.87)0.0161.75 (1.08–2.83)0.022Low Birth Weight92 (24.8)279 (75.2)1.95 (1.26–3.05)0.0031.80 (1.16–2.77)0.010Nutritional StatusBreastfeeding145 (39.1)226 (60.9)0.63 (0.45–0.87)0.0060.50 (0.28–0.89)0.022n = number, % = percentage, OR = odd ratio, CI = confidence interval Sociodemographic factors associated with K. pneumoniae infections in different age groups using univariate and multivariate regression analysis (p < 0.05) n = number, % = percentage, OR = odd ratio, CI = confidence interval Discussion This study provides a comprehensive analysis of antimicrobial resistance (AMR) profiles, extended-spectrum beta-lactamase (ESBL) production, carbapenemase production, and molecular resistance determinants in Klebsiella pneumoniae isolates from neonates, infants, and pediatric patients in a tertiary care setting. The overall prevalence of K. pneumoniae in this study was 10.7% and 19.6%, aligning with a hospital-based study from Punjab and Balochistan, Pakistan [21, 22]. Neonates showed higher prevalence (22%) compared to infants (15.9%) and pediatric patients (14.9%), reflecting their heightened vulnerability, which is supported by regional and international reports [23–25]. Comparable prevalence trends have been reported from India and Bangladesh, particularly among neonatal intensive care units (NICUs) [26, 27]. Studies from Saudi Arabia and Iran also reported similar rates of neonatal K. pneumoniae infections, reinforcing the regional burden across South Asia and the Middle East [28, 29]. In blood and urine samples of neonates and pediatric patients, multidrug-resistant (MDR) and extensive drug-resistant (XDR) isolates were identified. Similar findings were reported in China, where bloodstream infections in neonates were predominantly caused by MDR isolates, likely due to exposure to invasive procedures and prolonged hospital stays [30, 31]. In pediatric urine samples, the high prevalence of MDR is possibly related to increased urinary tract infections, consistent with other studies [32]. Resistance to beta-lactams was high across all age groups, particularly ampicillin (AMP) and amoxicillin-clavulanic acid (AMC), driven by AmpC enzymes and ESBL production. These findings align with studies reporting high beta-lactam resistance rates in Pakistan and globally [33]. MDR isolates also exhibited higher resistance to cefixime (CEF) and cefepime (FEP), indicating dissemination of cephalosporin resistance mediated by ESBLs and carbapenemase genes [34]. Resistance to fluoroquinolones (levofloxacin and ciprofloxacin) was observed across all age groups, especially in MDR and XDR isolates, primarily linked to gyrA and parC mutations and efflux pump overexpression [35–38]. A moderate effect was noted with carbapenems; however, XDR isolates showed markedly reduced susceptibility, reflecting the emergence of carbapenemase-producing strains. Resistance to sulfamethoxazole-trimethoprim (SXT) exceeded 85% in MDR and 95% in XDR isolates, while amikacin (AMK) demonstrated the highest efficacy, consistent with previous studies [39–41]. High erythromycin resistance further limits clinical options [42]. Regional studies from Pakistan support the high prevalence of resistant strains observed here [43, 44], while global reports confirm high resistance in neonates, infants, and pediatric populations [32, 45, 46]. Studies from other low- and middle-income countries further emphasize similar resistance dynamics [47]. Statistical analysis revealed that MDR strains were most prevalent in neonates. XDR isolates were predominantly found in neonates, highlighting the vulnerability of neonatal ICUs. These findings align with studies from India and Cambodia reporting high MDR and XDR prevalence in pediatric cohorts [48, 49]. Among the positive isolates, ESBL production was highest in MDR isolates (29%) and carbapenemase production in 22.6%. XDR strains in Pakistan are predominantly associated with blaOXA-48, while blaNDM-1 has also been reported [50]. In the present study, XDR isolates contained ESBL genes (blaSHV, blaTEM, blaCTX-M-14, blaCTX-M-15, blaOXA) and carbapenemase genes (blaOXA-48, blaKPC, blaNDM-1, blaIMP), consistent with reports from China, India, and Egypt highlighting ESBL and carbapenemase prevalence among MDR and XDR isolates [41, 45, 51]. Molecular analysis showed that blaCTX-M-14 was the most prevalent ESBL gene, followed by blaSHV, while blaOXA-48 was the most common carbapenemase gene [46]. Importantly, the emergence ofblaOXA-48 and blaIMP in pediatric isolates is of particular clinical concern. blaOXA-48 is recognized as one of the most globally disseminated carbapenemase genes, conferring low-level carbapenem resistance that may evade early detection [52]. The co-existence of blaOXA-48 with blaNDM-1 has been increasingly reported in South Asia, suggesting plasmid-mediated horizontal gene transfer [50]. Similarly, blaIMP encodes a metallo-β-lactamase that hydrolyzes nearly all β-lactams, and its detection in Pakistan represents an emerging threat [53]. Both genes are associated with nosocomial outbreaks and high mortality rates in neonatal units [31, 51]. Sociodemographic and clinical factors further highlighted high-risk groups. Neonates and infants were the most susceptible to MDR and XDR infections, likely due to frequent hospitalizations and underdeveloped immunity [54]. Prolonged hospitalization and prior antibiotic use were strongly associated with resistant infections, consistent with reports linking hospital exposure and antibiotic misuse to MDR and XDR emergence [55–57]. Males exhibited higher susceptibility, while Afghan refugees had increased risk, possibly due to limited healthcare access and overcrowding. Prematurity and low birth weight were significant risk factors, while breastfeeding was associated with a reduced risk, underscoring its protective role. Overall, this study reveals a high prevalence of MDR and XDR Klebsiella pneumoniae among neonates, infants, and pediatric patients in KP, Pakistan. The alarming resistance patterns and molecular profiles underscore the urgent need for stringent infection prevention measures, strengthened antimicrobial stewardship, and implementation of neonatal care bundles. Future genomic studies should broaden resistance profiling through sequencing to uncover hidden genetic determinants and inform more precise, evidence-based management strategies. Limitations The present study has several limitations to be considered. Firstly, sample collection from selected hospitals hinders the generalizability of the findings for the geographic region. PCR-based detection of ESBL and carbapenemase genes in K. pneumoniae was performed, but financial constraints did not allow the identification of the full spectrum of resistance genes and resistance mechanisms involved. Although sociodemographic and clinical data were carefully collected, certain self-reported variables, such as breastfeeding status, ethnicity, and refugee background, may be subject to recall or reporting bias. These factors could influence data accuracy and should be interpreted with caution. Mutational analysis of the selected genes would have been crucial for comparison to the sequences previously reported and their impact on resistance. Detailed treatment regimens were absent, and potential confounding effects of co-infections are among other limitations that could influence the interpretation of resistance patterns. Despite these limitations, the study provides valuable baseline data on antimicrobial resistance and risk factors, forming a foundation for future genomic surveillance and targeted interventions Conclusion This study demonstrates a high prevalence of MDR and XDR K. pneumoniae isolates among neonates, infants, and pediatric patients in KP province, with neonatal ICUs identified as especially vulnerable. The findings underscore the critical role of prolonged hospital stays, prior antibiotic exposure, and sociodemographic factors in driving the emergence and transmission of resistant infections. Molecular analysis revealed widespread ESBL and carbapenemase genes, particularly blaOXA-48 and blaNDM-1, which are key contributors to resistance in this context. These results underscore the urgent need for strengthened infection control measures, targeted antimicrobial stewardship, and ongoing molecular surveillance. Future research should focus on evaluating intervention strategies and informing policy to prevent the further spread of resistant K. pneumoniae isolates in pediatric settings. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2