Between January and December 2023, strains of Klebsiella spp. from patient clinical specimens were isolated and identified using the API 20 E biochemical gallery (BioMérieux SA, Lyon, France) at the YUTH’s Bacteriology Laboratory. The sampling method used was non-probabilistic convenience. Only one isolate was included per patient. The strains we collected were those isolated on a specific medium for enterobacteriaceae (EMB or MacConkey agar) and identified as Klebsiella. The presumed Klebsiella spp. strains isolated from the bacteriology laboratories of the YGH and the YGOPH were transported to the bacteriology laboratory of the YUTH for analysis, in compliance with bacterial strain transport conditions. These strains were innoculated into brain heart infusion broth supplemented with 10% glycerol, stored in an isothermal cooler. The strains were thus transported from the laboratories of the YGH and the YGOPH to the bacteriology laboratory of the YUTH. The antibiotics used were those recommended by the Antibiogram Committee of the French Society of Microbiology (EUCAST-CASFM, 2023). The study of the sensitivity of Klebsiella strains to antibiotics was determined by the disc diffusion method in agar medium (Mueller Hinton). The antibiotics tested were chosen according to the recommended panel for enterobacteriaceae. Susceptibility test results were interpreted using EUCAST-CASFM criteria, and interpretive reading of the antibiograms enabled the determination of the various phenotypes expressed by Klebsiella spp. The antibiotics studied (Oxoid brand) for all the strains isolated were the following: amoxicillin (20 μg), amoxicillin + clavulanic acid (20/10μg), ticarcillin (75 μg), ticarcillin + clavulanic acid (75/10μg), piperacillin (30 μg), piperacillin-tazobactam (30/6μg), temocillin (30 μg), cefalexin (30 μg), cefoxitin (30 μg), cefixime (5 μg), ceftazidime (10 μg), ceftazidime + avibactam (10/4μg), cefotaxime (5 μg), ceftriaxone (30 μg), cefepime (30 μg), aztreonam (30 μg), imipenem (10 μg), meropenem (10 μg), and ertapenem (10 μg). The detection of extended-spectrum β-lactamases was performed during the antibiotic susceptibility tests by highlighting an image of synergy between the inhibitors and C3G and/or C4G and/or aztreonam. Regarding the screening of carbapenemase-producing strains, the EUCAST-CASFM flowchart (2023) was used to detect strains suspected of producing carbapenemase, despite a categorization that may be “sensitive” to carbapenems.

As concerns quality control, the EMB/MacConkey culture media were inoculated with reference strains of Staphylococcusaureus ATCC 25923 and Escherichiacoli ATCC 25922, then incubated for 24 hours at 37˚C to verify their specificity and fertility. The absence of bacterial growth for Staphylococcusaureus and the presence of growth for Escherichiacoli confirmed that the culture media were selective. The antibiotic disks were tested with reference strains of Escherichiacoli ATCC 25922, and the agreement between the measured diameters and the expected diameters confirmed the good quality of the antibiotic disks. For ESBL detection, Klebsiellapneumoniae ATCC 700603 was used as a positive control, and E.coli ATCC 25922 as a negative control, in accordance with EUCAST-CASFM recommendations.

The data obtained were digitized using CSPro 7.1 software. Statistical analysis was performed using R software (version 4.3) and Microsoft Excel. Descriptive statistics were used to summarize socio-demographic and microbiological data as frequencies and percentages. Associations between categorical variables (carbapenemase production status and age group, sex, hospital ward, and clinical specimen type) were assessed using Pearson’s chi-squared test or Fisher’s exact test when expected cell counts were below 5. The significance threshold was set at p < 0.05 for all analyses. Effect sizes were estimated using Cramér’s V. For the age-group analysis, the eight original age categories were collapsed into four broader groups (0 - 20, 21 - 40, 41 - 60, and >60 years) to ensure adequate expected cell frequencies, yielding df = 3 for the chi-squared test. Post hoc pairwise comparisons between hospital wards were performed using odds ratios with 95% confidence intervals, with p-values derived from Fisher’s exact test. R and Excel were used to extract information from the database and to construct the various graphs and tables.

2.2.1. Socio-Demographic Characteristics

Table 1 below shows the isolated strain distribution according to the sex of the patients. Of the 264 isolates, 135 (51.1%) were from female patients and 129 (48.9%) were from male patients with sex ratio of 1:1.05. The mean age of the patients was 46.39 years (±21,349). The majority of strains were obtained from patients aged [40 - 50] years (20.5%), followed by those in the [50 - 60] and [60 - 70] age groups (19.3% of strains each) as seen in Table 1 below. This table equally shows the distribution of patients per department. Most isolates were from the emergency department and medicine unit, accounting for 22.7% and 22.0%, respectively. Only ten strains of Klebsiella spp. came from the neonatal department (3.8%).

Table 1. Socio-demographic characteristics of the patients.

2.2.2. Overall Antibiotic Resistance Profile

Figure 1 summarizes the resistance rates of the 264 Klebsiella spp. isolates against 19 antibiotics spanning seven pharmacological classes. Universal resistance (100%) was observed for three unprotected penicillins: amoxicillin, ticarcillin, and piperacillin. Temocillin, a penicillin with enhanced stability against certain β-lactamases, displayed a considerably lower resistance rate (31.4%).

Color scale: green (low resistance, ≤25%) → yellow (moderate) → red (high resistance, ≥75%).

Figure 1. Antibiotic resistance rates.

Among β-lactam/β-lactamase inhibitor (BL/BLI) combinations, resistance ranged from 37.1% for piperacillin/tazobactam to 66.7% for ticarcillin/clavulanic acid. Cephalosporin resistance was substantial: cefalexin (1st generation) at 75.4%, cefoxitin (2nd generation) at 34.0%, and third-generation cephalosporins ranging from 68.5% (ceftazidime) to 74.7% (cefixime). Cefepime (4th generation) showed 69.7% resistance, and aztreonam (monobactam) exhibited 68.2% resistance.

The carbapenems retained the lowest resistance rates among all antibiotic classes tested: imipenem (10.6%), meropenem (14.3%), and ertapenem (30.4%), yielding a mean carbapenem resistance rate of 18.4%. The novel combination ceftazidime/avibactam demonstrated a resistance rate of 21.7%. The radar plot (Figure 2) illustrates the mean resistance rate aggregated per antibiotic class.

Figure 2. Mean antibiotic resistance rates aggregated per pharmacological class.

2.2.3. Phenotypic Classification of the Strains Studied

The phenotypic classification of the strains based on their susceptibility profiles made it possible to identify the following phenotypes: Extended-spectrum β-lactamase (112 strains or 42.4%), low-level/wild-type penicillinase (53 strains or 20.1%), derepressed penicillinase (8 strains or 3.0%), carbapenemases (89 strains or 33.7%) and high-level cephalosporinase (13 strains or 4.9%). The most observed phenotypes were ESBL (42.4%), followed by the carbapenemase phenotype (33.7%). The derepressed penicillinase phenotype was the least identified (3.0%). Collectively, ESBL production and suspected carbapenemase production accounted for 76.1% of all resistance mechanisms identified, as shown in Figure 3 below. However, we would like to point out that in some bacterial strains, we found the expression of two phenotypes, such as the ESBL and carbapenemase phenotypes within the same strain.

ESBL = extended-spectrum β-lactamase.

Figure 3. The prevalence of β-lactam resistance mechanisms.

2.2.4. Proportion of Carbapenemase-Producing Strains

A total of 264 Klebsiella spp. isolates were subjected to antimicrobial susceptibility testing and subsequently screened for carbapenemase production using the established phenotypic algorithm. Of these, 89 isolates (33.7%) were classified as suspected carbapenemase producers, whereas 175 isolates (66.3%) were classified as non-carbapenemase producers. This finding indicates that approximately one in three clinical Klebsiella spp. isolates in the study setting demonstrated phenotypic characteristics consistent with carbapenemase production.

2.2.5. Association between Carbapenemase Production and Socio-Demographic Factors

1) Distribution per age group

The distribution of carbapenemase production status across eight age categories is presented in Figure 4. The highest proportion of suspected carbapenemase producers was observed in the 11 - 20 years age group (55.6%; 5 of 9 patients), while the lowest rate was recorded in the 21 - 30 years age group (27.8%; 10 of 36). The remaining age categories exhibited relatively uniform carbapenemase production rates ranging from 29.0% to 39.2%.

Chi-squared analysis revealed no statistically significant association between carbapenemase production and patient age group (χ² = 3.620; df = 3; p = 0.292). Cramér’s V was 0.126, indicating a negligible effect size. The elevated rate in the 11 - 20-year-old age group should be interpreted cautiously, given the small sample size (n = 9) in that category.

Figure 4. The distribution of carbapenemase production status per age group.

2)Distributionperhospitalward

The distribution of carbapenemase production across nine hospital wards/services is presented in Figure 5. The Medical unit exhibited the highest proportion of suspected carbapenemase producers (48.3%; 28 of 58 isolates), followed by the Intensive Care Unit (ICU; 41.7%; 10 of 24), Pediatrics (38.5%; 5 of 13), and Hemodialysis (33.3%; 4 of 12). The lowest rate was observed among outpatient isolates (17.9%; 7 of 39).

Figure 5. Suspected carbapenemase production rates per hospital ward.

Despite these observed variations, chi-squared analysis did not demonstrate a statistically significant overall association between carbapenemase production and hospital ward (χ² = 11.829; df = 8; p = 0.159; Cramér’s V = 0.220). However, post hoc odds ratio analysis with Fisher’s exact test revealed that isolates from the medical ward had significantly elevated odds of carbapenemase production compared to all other wards combined (OR = 2.25; 95% CI: 1.23 - 4.11; p = 0.011). Conversely, outpatient isolates demonstrated significantly reduced odds (OR = 0.37; 95% CI: 0.16 - 0.89; p = 0.026), suggesting a protective effect of non-hospitalization status.

2.2.6. Comparative Resistance Profile: Carbapenemase Producers vs. Non-Producers

A comparative analysis of antibiotic resistance rates between the two groups is detailed in Table 2. Both groups demonstrated identical universal resistance (100%) to amoxicillin, ticarcillin, and piperacillin. However, suspected carbapenemase producers exhibited significantly elevated resistance rates across the majority of the remaining 16 antibiotics tested.

Table 2. Comparative antibiotic resistance rates (%) between Klebsiella spp. non-carbapenemase-producers (n = 175) and suspected carbapenemase-producers (n = 89).

CP = carbapenemase producer; Δ = absolute difference between groups; NS = not significant. Significant p-values are highlighted in bold red.

The most pronounced differences (Δ ≥ 30%) were observed for piperacillin/tazobactam (60.7% vs. 25.1%; Δ = 35.6%), cefoxitin (57.3% vs. 22.3%; Δ = 35.0%), and ceftazidime/avibactam (44.0% vs. 10.3%; Δ = 33.7%). All three carbapenems showed significantly higher resistance in the carbapenemase-producing group: imipenem (21.3% vs. 5.1%), meropenem (30.3% vs. 6.3%), and ertapenem (45.0% vs. 22.9%), all with p < 0.001. Statistically significant differences were observed for 14 of 19 antibiotics. The five antibiotics without significant differences were amoxicillin, ticarcillin, and piperacillin (both groups at 100%), as well as imipenem and ertapenem.

Our study took place in the bacteriology laboratories of three public reference hospitals in the city of Yaounde (YUTH, YGH and GOPHY). The bacteriological analyses took place at the bacteriology laboratory of the YUTH. We included 264 strains of Klebsiella spp. isolated from the clinical specimens of 264 patients who had previously undergone a medical consultation. The majority of strains were obtained from patients aged [40 - 50] years (20.5%), followed by those in the [50 - 60] and [60 - 70] age groups (19.3% of strains each). In this study, there was a predominance of Klebsiella spp. in the clinical specimens of women. This predominance is also found in Kenkouo’s study [7]. This is thought to be due to the high frequency of urinary tract infections in female patients. Of the isolates obtained, Klebsiellapneumoniaepneumoniae was the most common subspecies (n = 164 or 62.1%), followed by K.pneumoniaeozaenae and K.pneumoniaerhinoscleromatis, representing 18.6% (n = 49) and 13.3% (n = 35) respectively. Urine was the most representative clinical specimen, followed by suppurations and blood, accounting for 39.8%, 31.1% and 28.8%, respectively. Mbamyah et al. [8] had also found that the lead clinical specimen was urine (68.7%), followed by suppurations (13.5%). This predominance can be explained by the fact that Klebsiella spp. is a bacterium of the intestinal commensal flora and can easily be found in the urinary tract due to poor hygiene, especially in women, given their genital anatomy.

Most patients were inpatients (85.2%), while 14.8% were outpatients. According to the study by Ebongue and collaborators in 2015 [9] on the evolution of antibiotic resistance in enterobacteriaceae isolated from clinical specimens at the Douala General Hospital, it appears that the majority of strains came from inpatients, i.e., 58.39%. Our rate is much higher than that found in their study, and this is explained by the fact that we were particularly interested in the Klebsiellas and not in all enterobacteriaceae. In addition, this bacterium is most often indexed as a nosocomial germ, which would explain the tendency to encounter it most often in inpatients.

Antibiotic susceptibility results show that the Klebsiellapneumoniae studied shows a very high level of resistance to the antibiotics tested. The universal resistance (100%) observed against unprotected penicillins (amoxicillin, ticarcillin, piperacillin) is concordant with the intrinsic chromosomal β-lactamase (SHV-1) production characteristic of Klebsiella spp. and the high prevalence of acquired resistance determinants in this genus [10][11]. These findings corroborate with published literature and definitively confirm the clinical obsolescence of unprotected penicillins for the empirical treatment of Klebsiella spp. Infections [12].

The isolated species were tested against 19 antibiotics belonging to the beta-lactam family. The high resistance rates observed for third-generation cephalosporins (cefotaxime: 72.1%; ceftriaxone: 73.0%; cefixime: 74.7%) underscore the extensive dissemination of ESBL-producing strains within the study population. These rates exceed the global median for ESBL-producing enterobacteriaceae reported by Mbamyah et al. [13] and the World Health Organization (WHO) Global Antimicrobial Resistance and Use Surveillance System (GLASS), which documents median rates of approximately 40% - 60% in low- and middle-income countries (LMICs) [14]. The wild resistance phenotype has been confirmed for all of our strains (CASFM, 2023). High resistance rates were observed with cefalexin, cefixime, ceftriaxone and cefotaxime (75.4%; 74.7%; 73.0% and 72.1% respectively). High resistance to cefalexin, cefixime, ceftriaxone and cefotaxime is consistent with the worldwide expansion of CTX-M-type ESBL enzymes [15][16].

Notably, the carbapenem class maintained the lowest overall resistance rates among all antibiotic classes evaluated, with imipenem demonstrating the highest retained susceptibility (resistance: 10.6%), followed by meropenem (14.3%) and ertapenem (30.4%). The 18.4% mean carbapenem resistance represents a clinically significant departure from the near-universal carbapenem susceptibility reported in earlier surveillance studies [14][17]. The disproportionately elevated ertapenem resistance relative to imipenem and meropenem is pharmacologically predictable, as ertapenem is more susceptible to hydrolysis by carbapenem-hydrolyzing enzymes, particularly OXA-48-like class D β-lactamases [18][19]. This differential susceptibility pattern supports the established practice of using ertapenem as a sentinel screening marker for carbapenemase production.

Overall, we observe a slight increase in the level of antibiotic resistance compared to a 2015 study (Betbeui et al., 2015) [20] and a 2024 study (Chafa et al., 2024) [21] with resistance rates to C1G (cefalotin), C2G (cefoxitin), C3G (ceftazidime), C4G (cefepime), monobactam (aztreonam) and imipenem. This observed increase in the level of resistance to β-lactam is probably the consequence of the inappropriate prescription and uncontrolled use of antibiotics in this family [22].

The resistance rate to ceftazidime/avibactam (21.7%) merits particular attention. This combination agent was specifically engineered to overcome resistance mediated by class A serine β-lactamases (including KPC-type carbapenemases) and class D β-lactamases (e.g., OXA-48). The observed resistance may therefore indicate the presence of class B metallo-β-lactamases (MBLs), such as NDM, VIM, or IMP, which are intrinsically refractory to avibactam inhibition [19][23]. This hypothesis warrants verification through molecular characterization studies, including targeted PCR or whole-genome sequencing.

Based on a precise arrangement of antibiotic discs with or without inhibitors, and the rigorous measurement of inhibition diameters, we were able to perform a phenotypic characterization of 264 strains of Klebsiella. It appears from this classification that the distribution of phenotypic frequencies is as follows: extended-spectrum β-lactamase (42.4%), low-level/wild-type penicillinase (20.1%), derepressed penicillinase (3.0%), carbapenemases (33.7%) and high-level cephalosporinase (4.9%). The most observed phenotypes are ESBL (42.4%), followed by the carbapenemase phenotype (33.7%). The high prevalence of “resistant” phenotypes, acquired mainly from the strains (79.9%) isolated from inpatient clinical specimens, is justified by the existence of increased antibiotic selection pressure [24]. The proportion of strains of Klebsiella spp. ESBL producers is 42.4%. According to these results, we can observe an increase in the frequency of ESBL-producing strains in Cameroon because a study carried out at the Yaounde Central Hospital in 2005 had revealed a prevalence of 18.8% of Klebsiella spp. ESBL producers [25] and the one carried out at the YUTH in 2015 showed a frequency of 30.3% of Klebsiella spp. ESBL producers [20]. However, the one conducted in 2023 by Mbamyah et al. revealed a percentage of 71.6% [8]. Moreover, in France, the RAISIN states that 17% of K.pneumoniae responsible for bacteremia are ESBL producers. In addition, Klebsiella spp. is one of the bacteria known to produce ESBL [26]. Our results are quite high as compared to those of other countries. This could be explained by the regular consumption of β-lactams, which have become increasingly familiar to bacteria, or by the prolonged hospital stays of patients.

The present study revealed that 33.7% of clinical Klebsiella spp. isolates were suspected of producing carbapenemases, representing a substantial epidemiological burden. This prevalence is notably elevated compared to several recently published studies from sub-Saharan Africa, where reported carbapenem-resistant enterobacteriaceae (CRE) rates generally range between 5% and 25%, and a study carried out by Betbeui et al. in Cameroon (2015), which revealed an 11.1% rate of carbapenemase-producing strains [20] and the one conducted in 2024 by Mbamyah et al., revealed a percentage of 10.7% [8]. The elevated prevalence observed in our setting likely reflects the confluence of multiple contributing factors, including the escalating selective pressure arising from empirical carbapenem prescribing, suboptimal infection prevention and control (IPC) practices, and the facilitation of horizontal gene transfer of resistance determinants within the healthcare environment. This high rate observed in our study could be explained by the fact that E.coli and Klebsiella spp. are the enterobacteriaceae most reported in literature as being involved in resistance linked to carbapenemase production [24]. The concurrent high prevalence of ESBL (42.4%) and suspected carbapenemase production (33.7%) in this study population is particularly concerning, as it suggests a progressive evolutionary trajectory of resistance mechanisms. In this paradigm, ESBL production may represent a transitional phenotype preceding the acquisition of carbapenem-hydrolyzing enzymes, consistent with the stepwise resistance model described in literature.

The predominance of carbapenemase-producing strains in inpatients of the medical, intensive care and pediatric wards (48.3%, 41.7% and 38.5% respectively) can be explained by the fact that ESBL and carbapenemases are enzymes produced mainly by nosocomial strains responsible for hospital epidemics [27] and in long-term care facilities [28].

The absence of a statistically significant association between carbapenemase production and patient age group (χ² = 3.620; p = 0.292; Cramér’s V = 0.126) suggests that carbapenemase-producing Klebsiella spp. infections affect patients across all age categories without a clear age-related predilection. This finding implies that the primary risk factors for acquisition of carbapenemase-producing organisms are more likely related to healthcare-associated exposures (e.g., duration of hospitalization, antibiotic exposure, invasive device utilization) rather than intrinsic demographic characteristics. The elevated rate in the 11 - 20 years age group (55.6%) should be interpreted with considerable caution, given the small sample size (n = 9), which substantially limits the statistical reliability of this estimate.

The overall chi-squared test did not demonstrate a statistically significant association between carbapenemase production and hospital ward (χ² = 11.829; p = 0.159). However, the post hoc odds ratio analysis yielded clinically significant findings: the medical unit demonstrated 2.25-fold higher odds of harboring carbapenemase-producing isolates relative to all other wards combined (OR = 2.25; 95% CI: 1.23 - 4.11; p = 0.011). This association likely reflects the characteristic clinical profile of patients admitted to medical wards, including prolonged lengths of stay, frequent empirical broad-spectrum antibiotic administration, the presence of multiple comorbidities, and increased rates of invasive device utilization, all recognized risk factors for multidrug-resistant organism (MDRO) acquisition.

The relatively high carbapenemase production rates observed in the ICU (41.7%) and Pediatrics (38.5%), although not individually statistically significant, are consistent with the well-established literature identifying these settings as high-risk environments for MDRO colonization and transmission [29][30]. The significantly lower odds observed among outpatient isolates (OR = 0.37; 95% CI: 0.16 - 0.89; p = 0.026) provide complementary evidence that healthcare-associated exposure plays a pivotal role in the acquisition of carbapenemase-producing strains, consistent with the nosocomial transmission paradigm described for CRE globally [31][32].

The significantly elevated resistance rates observed in suspected carbapenemase producers across 14 of 19 antibiotics tested (p < 0.001 for the majority) demonstrate the extensive co-resistance phenotype typically associated with carbapenemase production [19][31]. The most clinically impactful differences were observed for piperacillin/tazobactam (Δ = 35.6%), cefoxitin (Δ = 35.0%), and ceftazidime/avibactam (Δ = 33.7%). The substantially higher cefoxitin resistance (57.3% vs. 22.3%) in the carbapenemase-producing group is particularly noteworthy, as cefoxitin resistance is characteristically associated with AmpC cephalosporinase production or porin loss mechanisms that frequently co-exist with carbapenemase production on multi-resistance plasmids [29].

These observations support the paradigm of co-resistance and co-selection, whereby mobile genetic elements encoding carbapenemases (particularly large conjugative plasmids of the IncF, IncL/M, and IncN families) frequently carry additional resistance determinants conferring resistance to aminoglycosides, fluoroquinolones, trimethoprim-sulfamethoxazole, and other β-lactams. The resulting pan-resistant or extensively drug-resistant (XDR) phenotype critically narrows the therapeutic armamentarium, often limiting effective treatment options to polymyxins, tigecycline, and potentially ceftazidime/avibactam, underscoring the urgent need for comprehensive antimicrobial stewardship and robust infection prevention programs.

The findings of this study carry substantial implications for empirical antibiotic prescribing, infection control policy, and antimicrobial resistance (AMR) surveillance. With approximately one-third of Klebsiella spp. isolates exhibiting suspected carbapenemase production, clinicians should exercise heightened vigilance in selecting empirical antibiotic regimens, particularly for patients admitted to high-risk wards such as the medical and the ICU. Routine phenotypic carbapenemase screening should be integrated into clinical microbiology laboratory workflows, complemented where feasible by molecular confirmation methods (e.g., multiplex PCR for blaKPC, blaNDM, blaOXA-48, blaVIM, blaIMP) to guide targeted therapy.

Institutional antimicrobial stewardship programs should prioritize the following strategic interventions: 1) restriction of empirical carbapenem use through prospective audit and feedback mechanisms; 2) promotion of carbapenem-sparing combination regimens for documented CRE infections, guided by antimicrobial susceptibility testing; 3) adoption of rapid diagnostic technologies for carbapenemase detection to minimize time-to-appropriate therapy; and 4) implementation of de-escalation protocols following receipt of definitive culture and susceptibility results.

Enhanced infection prevention and control measures, including rigorous contact precautions for CRE-colonized or infected patients, active surveillance cultures on admission to high-risk units, environmental decontamination protocols, and antimicrobial stewardship integration, are essential to curtail the nosocomial dissemination of carbapenemase-producing Klebsiella spp.