Characterization of Carbapenem-Resistant Gram-Negative Bacilli Isolates in Multispecialty Private Hospitals in Lagos, Nigeria

Abstract

Background/Objectives: Carbapenem-resistant Gram-negative bacilli (CR-GNB) pose a growing challenge to public health worldwide due to limited treatment options. This cross-sectional study investigated the characteristics of CR-GNB isolated from clinical specimens in Lagos, Nigeria. Methods: Gram-negative bacilli (GNB) and clinical data were obtained from three multi-specialist private hospitals between March and June 2023. The GNB were identified using the Analytical Profile Index (API) and investigated for CR-GNB by disk diffusion. Antimicrobial resistance patterns and carbapenemase gene data for presumptive carbapenemase-producing Gram-negative bacilli (CP-GNB) were analyzed using Vitek-2 and polymerase chain reaction (PCR). Results: Of 317 GNB, 29.0% (n = 92) were CR-GNB. Significantly higher numbers of CR-GNB were reported from the intensive care unit and oncology department (p = 0.009). Of all CR-GNB, 17 isolates (18.5%) were classified as presumptive CP-GNB. In this subgroup, resistance rates of ampicillin/sulbactam (100.0%) and trimethoprim/sulfamethoxazole (100.0%) were highest. Ten (10) CP-GNB were confirmed, representing 3.15% of all GNB tested. Seven isolates of New Delhi Metallo-β-lactamase (bla_NDM) were found among P. aeruginosa, K. pneumoniae, E. coli, and A. baumannii. The bla_NDM was identified in strains classified as extensively drug-resistant (XDR) and pandrug-resistant. Conversely, the bla_KPC was detected solely in multidrug-resistant and XDR strains. Conclusions: Emerging CR-GNB, specifically CP-GNB, in Nigeria emphasize the need for specific therapeutic management of infected patients. Antimicrobial stewardship and long-term surveillance efforts must be implemented in healthcare settings, as well as improved, accelerated microorganism identification techniques.

1. Introduction

Antimicrobial resistance (AMR) poses a growing challenge to global health. Recognizing the escalating nature of this threat, the World Health Organization (WHO) designated a specific bacterial priority pathogens list (BPPL) to guide research and facilitate novel drug development [1]. These pathogens include carbapenem-resistant Gram-negative bacilli (CR-GNB), such as Acinetobacter and Pseudomonas, and Enterobacterales. Since these pathogens can produce carbapenemase that render beta-lactam antimicrobials ineffective, they have the potential to spread widely and cause outbreaks with unique variabilities in prevalence and characteristics in diverse settings [1].

CR-GNB develop antimicrobial resistance through various genetic and biochemical adaptations. The widespread emergence of CR-GNB is largely driven by carbapenemase production, a resistance mechanism propagated through the horizontal transfer of encoding genes. Owing to their capacity to hydrolyze various β-lactam antimicrobials, including carbapenems, cephalosporins, penicillin, and aztreonam, Carbapenemase-producing Gram-negative bacilli (CP-GNB) are clinically important. They pose a significant threat by compromising the activity of the last-resort antimicrobials used for treating severe infections. Moreover, epidemiologically, carbapenemase genes can disseminate into various bacteria across almost all geographic regions [2]. CP-GNB are more virulent and associated with higher levels of AMR, worse health outcomes, and more rapid spread.

The variation in CP-GNB’s prevalence is influenced by carbapenemase production. which is categorized into different Ambler classes. Classes A, C, and D carbapenemase include the β-lactamases with serine at their active site, whereas the molecular class B carbapenemase includes metalloenzymes with a zinc active site [2]. Ambler classes A, B, and D carbapenemase are clinically significant among nosocomial pathogens. Class A is notably represented by the Klebsiella pneumoniae carbapenemase (KPC), class B includes MBLs such as New Delhi Metallo-β-lactamases (NDM), Verona integron-encoded (VIM), and imipenemase (IMP) enzymes, and class D comprises carbapenem-hydrolyzing oxacillinases like OXA-48 [2]. Class D includes carbapenem-hydrolyzing oxacillinase (OXA), such as the OXA-48 enzymes [3,4]. Carbapenemase genes were initially found to be chromosomally mediated in specific bacterial species. However, they can now also be mediated by plasmids. This horizontal transfer facilitates the faster spread of resistance between different bacterial species and genera [5]. The most frequently reported plasmid-mediated carbapenemase genes include bla_VIM, bla_IMP, bla_KPC, bla_NDM and bla_OXA-48, which vary across continents [6,7].

In Nigeria, studies in public hospitals have reported a high prevalences of CR-GNB. The most prevalent CR-GNB species are Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa [8,9]. A review of these studies revealed a national prevalence average of 21.3%, with bla_NDM and bla_VIM as the most detected genes [8]. It must be noted that most of the studies focused on Enterobacterales using regional data that varied according to each study. For example, a 7.7% prevalence of carebapenem-resistant Enterobacterales was reported among hospitalized patients in South-west Nigeria, while a 29% prevalence was reported in Southeast Nigeria [10,11]. A more comprehensive study that profiled CR-GNB reported a prevalence of 22% for CR-GNB in a tertiary hospital in South-west Nigeria [9].

Infections caused by carbapenem-resistant microorganisms have worse clinical outcomes than those caused by carbapenem-sensitive microorganisms [12,13]. Some of the notable risk factors for increased mortality are past antimicrobial usage, prolonged hospital stays, surgical interventions, the presence of indwelling devices, intensive care unit (ICU) admission, and extended hospitalization [14,15]. The studies conducted in Nigerian public hospitals over the past ten years indicate that the threat of CR-GNB is escalating [4,6,16,17]. However, despite this and also the vital role of private hospitals in patient care in Nigeria—particularly in Lagos State, which has a population of over 24 million and a private hospital sector that serves more than 60% of its residents—there is sparse data on CR-GNB in these settings [18]. This study thus aimed to investigate the characteristics of CR-GNB, including the occurrence of CP-GNB isolated from clinical specimens in Lagos, Nigeria. This comprehensive data, encompassing both phenotype and genotype, is imperative for the effective treatment and enhanced control of AMR.

2. Materials and Methods

2.1. Study Design and Setting

A cross-sectional study was conducted from March to June 2023 at three multispecialty private hospitals in Lagos, Nigeria: R-Jolad Hospital, Lagoon Hospitals, and Havana Specialist Hospital. These facilities span the three senatorial districts and collectively serve a diverse patient population. R-Jolad Hospital has 215 beds, Lagoon Hospitals has a total of 120 beds across its branches, and Havana Specialist Hospital has 100 beds. While none of these hospitals have a formal antimicrobial stewardship program, they do implement infection prevention and control initiatives.

These hospitals were meticulously chosen for their exceptional ability to manage a diverse range of medical cases and their proficiency in conducting in-house laboratory testing. Furthermore, functional electronic medical records that connect laboratories to other hospital units were important selection criteria. In total, 317 Gram-negative bacilli (GNB) were collected from both inpatients and outpatients, without a specific focus on individuals with severe infections or hospitalizations. These samples were derived from routine laboratory microbiological investigations and were subsequently tested for carbapenem resistance in the next stage of analysis. Among the isolates, based on the presence of clinical signs and symptoms indicative of infection, 296 were identified as responsible for clinical infections, and 21 were categorized as colonizers. Moreover, to identify the associated factors, the patients’ relevant clinical records were obtained after their discharge or death.

2.2. Identification of Gram-Negative Bacilli

Consecutive, non-duplicate isolates of GNB were obtained from the chosen hospitals. The bacterial isolates were recovered from a variety of clinical samples, including urine, high vaginal swab, stool, wound swab, ear swab, throat swab, sputum, catheter tips, semen, tracheal aspirate, and blood. These isolates were cultured on MacConkey agar (Oxoid, Thermo Fisher Scientific, London, UK) to differentiate between lactose fermenters and non-lactose fermenters. From each agar plate, distinct colonies were selected and cultured on Mueller-Hinton agar (Oxoid, Thermo Fisher Scientific, London, UK). Unique identification numbers were assigned to each isolate in mixed cultures to manage potential errors. The oxidase test was then performed using MAST ID™ oxidase strips (MAST Group Limited, Merseyside, UK). All the isolates were then subjected to Gram staining (Sigma-Aldrich, Burlington, MA, USA), and quality control was maintained using P. aeruginosa ATCC 27853 and E. coli ATCC 25922. Gram staining was performed according to the standard protocol established by the American Society for Microbiology [19]. Finally, the pure isolates were identified using API 20 NE for oxidase-positive non-lactose fermenters and API 20E for oxidase-negative lactose fermenters by following the manufacturer’s instructions (bioMérieux, Marcy l’Etoile, France). To minimize misidentification, colony morphology, Gram staining, oxidase testing, and quality control strains (P. aeruginosa ATCC 27853 and E. coli ATCC 25922) were used for cross-validation (Table S1).

2.3. Detection of Carbapenem-Resistant Gram-Negative Bacilli

The identified GNB were tested for CR-GNB by checking their antimicrobial susceptibility to ertapenem, imipenem, and meropenem (MAST Group Ltd., Merseyside, UK). The standard and interpretive criteria recommended by the European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines were followed [20]. In our study, isolates classified as ‘category I’ (now termed ‘susceptible, increased exposure’) were combined with other susceptible isolates under the non-resistant group, in accordance with local laboratory reporting protocols. Subsequently, the isolates demonstrating potential resistance to ertapenem, imipenem, or meropenem were further screened for CP-GNB using a minimum inhibitory concentration (MIC) test via the Vitek 2 system (bioMérieux, Marcy l’Etoile, France). This test was interpreted using the Clinical and Laboratory Standards Institute (CLSI) guidelines [21], and based on the test results, the isolates that showed potential resistance or intermediate susceptibility to imipenem or meropenem were categorized as presumptive CP-GNB [22,23]. In this study, a combined methodology was employed, involving disk diffusion testing followed by EUCAST guidelines [20] and MIC testing, which applied CLSI interpretive criteria [21]. Although this dual approach may appear inconsistent, it reflects the standard workflow in the participating laboratories.

2.4. Antibacterial Susceptibility Test of the Presumptive Carbapenemase-Producing Gram-Negative Bacilli

The CP-GNB were tested for antimicrobial susceptibility to 14 antimicrobial agents using Vitek 2 AST cards (bioMérieux, Marcy l’Etoile, France). The antimicrobials tested were ampicillin/sulbactam, trimethoprim/sulfamethoxazole, piperacillin/tazobactam, cefotaxime, ceftazidime, ceftazidime/avibactam, ceftolozane/tazobactam, cefepime, imipenem, meropenem, gentamicin, ciprofloxacin, amikacin, and colistin. The MIC and clinical breakpoints were determined and interpreted using the CLSI guidelines [21]. Quality control was performed using P. aeruginosa ATCC 27853 and E. coli ATCC 25922. The antimicrobial resistance pattern was divided into three levels. Multidrug-resistant (MDR) organisms were defined as microorganisms that demonstrate non-susceptibility to ≥1 antimicrobial agent in ≥3 antimicrobial categories. Extensively drug-resistant (XDR) organisms were defined as “non-susceptible to at least one agent in all but two or fewer antimicrobial categories”. Pandrug-resistant (PDR) organisms were defined as non-susceptible to all antimicrobial agents in all antimicrobial categories [24].

2.5. Molecular Detection of Carbapenemase Genes

The identified CR-GNB’s DNA templates were prepared using the commercial DNA extraction kit (NIMR, Lagos, Nigeria). PCR techniques were employed to detect five significant carbapenemase genes. The PCR mixture (total 20 µL) comprised 1X FIREPol Master Mix (Solis Biodyne, Tartu, Estonia), 0.2 µM of each primer shown in

Table 1. Oligonucleotide primers and conditions used in this study.

Primer	Sequence (5′–3′)	Amplicon Size (bp)	PCR Condition (35 Cycles)			Reference
			Denature	Annealing	Extension
BlaNDM-F	GGGCAGTCGCTTCCAACGGT	475	95 °C	58 °C	72 °C	[25]
BlaNDM-R	GTAGTGCTCAGTGTCGGCAT		30 s	30 s	1.30 s
BlaOXA-F	TTGGTGGCATCGATTATCGG	438	95 °C	55 °C	72 °C	[26]
BlaOXA-R	GAGCACTTCTTTTGTGATGGC		30 s	30 s	1.30 s
BlaKPC-F	CATTCAAGGGCTTTCTTGCTGC	538	95 °C	55 °C	72 °C	[27]
BlaKPC-R	ACGACGGCATAGTCATTTGC		30 s	30 s	1.30 s
BlaIMP-F	GGAATAGAGTGGCTTAAYTC	232	95 °C	55 °C	72 °C	[28]
BlaIMP-R	TCGGTTTAAYAAAACAACCACC		30 s	30 s	1.30 s
BlaVIM-F	GATGGTGTTTGGTCGCATA	390	95 °C	55 °C	72 °C	[28]
BlaVIM-R	CGAATGCGCAGCACCAG		30 s	30 s	1.30 s

Note: bla_NDM—New Delhi Metallo-β-lactamase; bla_OXA—Oxacillinase-type β-lactamase; bla_KPC—Klebsiella pneumoniae carbapenemase; bla_IMP—Imipenemase Metallo-β-lactamase; bla_VIM—Verona Integron-encoded Metallo-β-lactamase; F—Forward Primer; R—Reverse Primer.

, 2 µL of DNA template (10–100 ng), and deionized water (which was used to make the reaction mixture). Amplification was conducted in a Techne Prime thermal cycler (Cole Palmer, Vernon Hills, IL, USA) with the PCR conditions shown in Table 1. The PCR products were analyzed using electrophoresis on a 1.5% agarose gel with SYBR Safe stain (Thermo Fisher Scientific, Waltham, MA, USA), and their sizes were determined based on a 100 bp DNA ladder (Solis Biodyne, Tartu, Estonia). After electrophoresis, the DNA bands were visualized under a UV transilluminator (Biobase, Jinan, China). The positive controls for bla_KPC, bla_VIM, bla_NDM, bla_IMP, and bla_OXA-48 detection were K. pneumoniae NCTC 13438, K. pneumoniae NCTC 13440, K. pneumoniae NCTC 13443, E. coli NCTC 13476, and K. pneumoniae NCTC 13442, respectively (Table S1).

2.6. Data Analysis

In a previous study conducted in Nigeria, the prevalence of CR-GNB was 22% (39/177 isolates) [9]. Based on this, we estimated a minimum sample size of 326 isolates using a one-proportion formula at a 95% confidence level and a 4.5% margin of error.

AMR was analyzed using the WHONET 2023 software version. Statistical analyses were performed using the SPSS program 18 version for Windows (PASW serial no. 5082357) (SPSS Inc., Chicago, IL, USA). The data were analyzed using descriptive statistics (e.g., frequency distribution, percentage, and interquartile range). Pearson’s chi-square and Fisher’s exact tests were used to analyze the statistical differences in the proportions of non-carbapenem-resistant and carbapenem-resistant isolates based on their expected values. Depending on the data distribution, an independent t-test was conducted to compare the number of non-carbapenem-resistant and carbapenem-resistant isolates within the interquartile range. A confidence interval of 95% was established, with a p-value of less than 0.05 required for statistical significance.

3. Results

3.1. The Identified Gram-Negative Bacilli Isolate

In total, 317 consecutive non-duplicate isolates of GNB were obtained. Of these, 261 (80.3%) GNB isolates were obtained from patients who presented to the general outpatient departments in all the chosen hospitals. The identified bacteria corresponded to Escherichia coli (40.1%) and Klebsiella pneumoniae (40.1%), accounting for about 80.1% of the isolates. In addition, Pseudomonas aeruginosa (7.3%), Proteus mirabilis (4.7%), Enterobacter species (3.2%), and Enterobacter cloacae (1.6%) were also prominent, collectively representing 16.8% of the isolates (Figure 1).

3.2. Demographic Distribution of Carbapenem-Resistant Gram-Negative Bacilli

Table 2. Prevalence of non-carbapenem-resistant Gram-negative bacilli (non-CR-GNB) and carbapenem-resistant Gram-negative bacilli (CR-GNB) by demographics, setting, and specimen type.

	Non-CR-GNB (n = 225)	CR-GNB (n = 92)	p-Value
Age, year (IQR)	36 (29–45)	38.5 (29.8–49.0)	0.050 c
Sex, n (%)			0.758 a
Female	181 (80.4)	72 (78.3)
Male	44 (19.6)	20 (21.7)
Patient Type, n (%)
Outpatient	181 (80.4)	80 (87.0)	0.196 a
Inpatient	44 (19.6)	12 (13.0)
Setting, n (%)
Outpatient Departments
General medicine	154 (85.1)	51 (63.7)	<0.001 b,*
Obstetrics and gynecology	13 (7.2)	14 (17.5)	0.150 a
Medicine	4 (2.2)	3 (3.8)	0.441 b
Pediatrics	3 (1.7)	0 (0.0)	0.555 b
Ear nose and throat	3 (1.7)	4 (5.0)	0.206 b
Surgery	1 (0.6)	3 (3.8)	0.087 b
Nephrology and dialysis	1 (0.6)	0 (0.0)	1.000 b
Neurology	7 (0.6)	0 (0.0)	1.000 b
Urology	1 (0.6)	2 (2.5)	0.223 b
Oncology	0 (0.0)	3 (3.8)	0.028 b,*
Inpatient Departments
General medicine	26 (59.1)	3 (25.0)	0.052 a
Obstetrics and gynecology	5 (11.4)	0 (0.0)	0.574 b
Medicine	3 (6.8)	0 (0.0)	1.000 b
Pediatrics	2 (4.5)	1 (8.3)	0.552 b
Surgery	2 (4.5)	0 (0.0)	1.000 b
Intensive care unit	2 (4.5)	6 (50.0)	0.001 b,*
Nephrology and dialysis	1 (2.3)	0 (0.0)	1.000 b
Neurology	1 (2.3)	0 (0.0)	1.000 b
Oncology	1 (2.3)	2 (16.7)	0.113 b
Cardiology	1 (2.3)	0 (0.0)	1.000 b
Specimen type, n (%)
Urine	178 (79.1)	73 (79.3)	1.000 a
High vagina swab	13 (5.8)	6 (6.5)	1.000 a
Stool	10 (4.4)	1 (1.1)	0.186 b
Wound swab	10 (4.4)	1 (1.1)	0.186 b
Ear swab	6 (2.7)	4 (4.3)	0.484 b
Throat swab	3 (1.3)	1 (1.1)	1.000 b
Sputum	2 (0.9)	1 (1.1)	1.000 b
Catheter	2 (0.9)	2 (2.2)	0.583 b
Semen	1 (0.4)	0 (0.0)	1.000 b
Trachea aspirate	0 (0.0)	2 (2.2)	0.084 b
Blood	0 (0.0)	1 (1.1)	0.290 b

Abbreviation: IQR = Interquartile range. ^a Pearson chi-square test, ^b Fisher-exact test, ^c Independent t-test, * Statistically significant differences (p < 0.05).

presents the patients’ demographic characteristics, the origins of the specimens, and the prevalence of CR-GNB isolates. Among the analyzed GNB, 92 isolates (29.0%) were identified as CR-GNB based on susceptibility testing to ertapenem, imipenem, and meropenem. The median age of patients with CR-GNB isolates (38.5 years, IQR 29.8–49.0) was slightly higher than that of patients with non-CR-GNB isolates (36 years, IQR 29–45), with the difference reaching borderline statistical significance (p = 0.050). Notably, there was no significant difference in the prevalence of CR-GNB isolates between male and female patients. Although most of the isolates were collected from outpatients, there was no statistically significant difference between the outpatient and inpatient settings (p = 0.196). Moreover, a significant distinction was found between the general medicine departments in the outpatient setting, where CR-GNB isolates were identified less frequently compared to non-CR-GNB isolates (p < 0.001). In contrast, CR-GNB isolates were more prevalent (3.8%) in the oncology outpatient departments, whereas none were detected in the non-resistant group (p = 0.028). Within inpatient settings, a significantly higher proportion of CR-GNB isolates were recovered from patients admitted to ICUs (p = 0.001). Among the analyzed specimen types, CR-GNB isolates were most frequently found in urine samples (79.3%), followed by high vaginal swabs (6.5%) and ear swabs (4.3%). However, we did not find any significant differences in the frequency of CR-GNB isolates across the different specimen types.

3.3. Prevalence and Proportion of Carbapenem-Resistant Gram-Negative Bacilli

Out of the 317 Gram-negative bacilli isolates examined, 92 (29.0%) were identified as resistant to carbapenems. The most prevalent carbapenem-resistant organisms were E. coli (46.7%) and K. pneumoniae (28.3%), followed by P. aeruginosa (14.1%). Other species exhibiting carbapenem resistance included P. mirabilis (6.5%), A. baumannii (2.2%), E. asburiae (1.1%), and E. fergusonii (1.1%). Moreover, several species were exclusively present in the non-carbapenem-resistant group and were not detected among the resistant isolates. These species included E. cloacae, A. fergusonii, B. cepacia, R. ornithinolytica, and Salmonella enterica (

Table 3. Prevalence of non-carbapenem-resistant and resistant isolates across bacterial species.

	Non-carbapenem Resistant Gram-Negative Bacilli (n = 225), n (%)	Carbapenem-Resistant Gram-Negative Bacilli (n = 92), n (%)
Klebsiella pneumoniae	101 (44.9)	26 (28.3)
Escherichia coli	84 (37.3)	43 (46.7)
Pseudomonas aeruginosa	10 (4.4)	13 (14.1)
Enterobacter species	10 (4.4)	0 (0.0)
Proteus mirabilis	9 (4.0)	6 (6.5)
Enterobacter cloacae	5 (2.2)	0 (0.0)
Proteus spp.	2 (0.9)	0 (0.0)
Aeromonas fergusonii	1 (0.4)	0 (0.0)
Burkholderia cepacia	1 (0.4)	0 (0.0)
Raoultella ornithinolytica	1 (0.4)	0 (0.0)
Salmonella enterica	1 (0.4)	0 (0.0)
Acinetobacter baumannii	0 (0.0)	2 (2.2)
Enterobacter asburiae	0 (0.0)	1 (1.1)
Escherichia fergusonii	0 (0.0)	1 (1.1)

).

3.4. Antimicrobial Resistance Profiles and Carbapenemase-Encoding Genes of Presumptive Carbapenemase-Producing Gram-Negative Bacilli

The MIC screening of CR-GNB using imipenem and meropenem identified 17 out of 92 isolates (18.5%) as presumptive CP-GNB. Of the 17 CP-GNB, the resistance rates of ampicillin/sulbactam (100.0%) and trimethoprim/sulfamethoxazole (100.0%) were the highest, followed by piperacillin/tazobactam (94.1%), cefotaxime (94.1%), and ceftolozane/tazobactam (88.2%). In contrast, the resistance rate of colistin (11.8%) was the lowest. The resistance rates of imipenem and meropenem were 82.4% and 58.8%, respectively. Most of the presumptive CP-GNB (11 isolates) were XDR, demonstrating non-susceptibility to at least one antimicrobial agent in all but a maximum of two antimicrobial categories (

Table 4. Antimicrobial resistance and carbapenemase gene profiles of presumptive carbapenemase-producing Gram-negative bacilli.

Sample Code	Organisms	Carbapenemase Gene	SAM	TS	TZP	CTX	CTZ	CAZ	CZA	FEP	IMP	MEM	CIP	GEN	AMK	COL	Classify
H1a	P. aeruginosa	blaNDM	-	-	R	R	R	R	R	R	R	S	R	R	S	I	XDR
R123	P. aeruginosa	Nd	-	-	R	R	R	R	R	R	R	S	R	R	S	S	XDR
L65	P. aeruginosa	Nd	-	-	R	R	R	R	R	R	R	R	S	R	R	I	XDR
L73	P. aeruginosa	Nd	-	-	R	R	S	R	R	R	R	S	R	S	S	S	MDR
L23	P. aeruginosa	Nd	-	-	R	R	R	R	R	R	R	S	S	R	S	R	XDR
L74	K. pneumoniae	blaKPC	R	R	R	R	R	R	R	R	R	R	R	S	I	S	XDR
L29	K. pneumoniae	blaKPC	R	R	R	R	R	S	S	S	I	I	S	S	-	I	MDR
L26	K. pneumoniae	blaKPC	R	R	R	R	R	S	S	S	I	I	S	S	-	I	MDR
R102	K. pneumoniae	blaNDM	R	R	R	R	R	R	R	R	R	R	R	R	I	S	XDR
L19	E. coli	blaNDM	R	R	R	R	R	R	R	R	R	R	R	R	R	R	PDR
R135	E. coli	blaNDM	R	R	R	R	R	R	R	R	R	R	R	R	R	S	XDR
H26	E. coli	blaNDM	R	R	R	R	R	R	R	R	R	R	R	R	S	S	XDR
R140	E. coli	Nd	R	R	R	R	R	R	R	R	R	R	R	R	S	S	XDR
R104	A. baumannii	blaNDM	R	R	R	R	R	R	R	R	R	R	R	R	S	S	XDR
R120	A. baumannii	blaNDM	R	R	R	R	R	R	R	R	R	R	R	S	I	S	XDR
R92	E. fergusonii	Nd	R	R	R	R	R	R	S	S	S	R	S	S	S	S	MDR
L61	P. mirabilis	Nd	R	R	S	S	S	S	S	S	R	S	S	I	S	S	MDR

Abbreviations: R, Resistance; I intermediate; S, Susceptible; SAM, Ampicillin/Sulbactam; TS, Trimethoprim/Sulfamethoxazole; TZP, Piperacillin/Tazobactam; CTX, Cefotaxime; CTZ, Ceftolozane/Tazobactam; CAZ, Ceftazidime; CZA, Ceftazidime/Avibactam; FEP, Cefepime; IMP, Imipenem; MEM, Meropenem; CIP, Ciprofloxacin; GEN, Gentamicin; AMK, Amikacin; COL, Colistin; MDR, Multidrug-resistance; XDR, Extensively drug-resistance; PDR, Pandrug-resistance; Nd, not detected; -, no analysis.

).

Out of the presumptive 17 CP-GNB, 10 (58.8%) were carbapenemase-producing strains. Overall, the 10 CP-GNB accounted for 3.15% of the GNB. Specifically, the genes bla_NDM and bla_KPC were detected in seven and three isolates, respectively (Figures S1 and S2). We could not identify the presence of the bla_VIM, bla_IMP, and bla_OXA. The bla_NDM was found in P. aeruginosa, K. pneumoniae, E. coli, and A. baumannii. However, the bla_KPC was exclusively associated with K. pneumoniae. The bla_NDM was identified in the strains classified as XDR (six isolates) and PDR (one isolate). Conversely, the bla_KPC was detected solely in MDR (two isolates) and XDR (one isolate) (Table 4).

3.5. Demographic Data and Clinical Outcome of Presumptive Carbapenemase-Producing Gram-Negative Bacilli

As

Table 5. Clinical and microbiological information on patients with carbapenem-resistant isolates.

Sample Code	Organisms	Hospital Unit/Department	Specimen	Diagnosis	30-Day Mortality Outcome
L23	P. aeruginosa	ICU	Wound Swab	Surgical wound infection	* No
L65	P. aeruginosa	ENT	Ear Swab	Otitis media	* No
L73	P. aeruginosa	ENT	Ear swab	Otitis media	* No
R123	P. aeruginosa	O&G	High Vaginal Swab	Pelvic Inflammatory Disease	* No
H1a	P. aeruginosa	Oncology	Throat Swab	Enlarged adenoid	* No
L19	E. coli	ICU	Blood	Benign neoplasia/Pituitary adenoma	** Died
H26	E. coli	O&G	Urine	UTI/Preterm premature rupture of membrane	* No
R135	E. coli	General Medicine/Surgery Outpatient	Urine	UTI	* No
R140	E. coli	General Medicine/Surgery Outpatient	Urine	UTI	* No
L26	K. pneumoniae	ICU	Catheter Tip	Prostate Enlargement/CVA	* No
L29	K. pneumoniae	ICU	Trachea Aspirate	Adenocarcinoma of the prostate/COPD	* No
L74	K. pneumoniae	General Medicine/Surgery Outpatient	Urine	UTI	* No
R10	K. pneumoniae	Pediatrics	Stool	Neonatal sepsis	* No
R104	A. baumannii	General Medicine/Surgery Outpatient	Urine	UTI	* No
R120	A. baumannii	General Medicine/Surgery Outpatient	Urine	Pyelonephritis/UTI	* No
R92	E. fergusonii	General Medicine/Surgery Outpatient	Urine	UTI	* No
L61	P. mirabilis	General Medicine/Surgery Outpatient	Urine	UTI	* No

Abbreviations: ICU, Intensive care unit; O&G, Obstetrics and gynecology; ENT, Ear nose and throat; UTI, Urinary Tract Infection; CVA, Cerebrovascular Accident; COPD, Chronic Obstructive Pulmonary Disease. * No mortality was recorded within 30 days. ** Patient died within 30 days.

shows, the presumptive CP-GNB was predominantly found in general medicine/surgery outpatient departments, primarily from the urine specimens collected from patients exhibiting signs of presumptive urinary tract infections. Most of the patients were subsequently discharged without any recorded mortality within 30 days. Regrettably, a patient (L19) in the ICU succumbed to a benign neoplasm, specifically a pituitary adenoma, with carbapenem-resistant E. coli identified in the blood sample. This strain also exhibited PDR, indicating resistance to all antimicrobial agents across the tested categories, including colistin, the last-resort antimicrobial.

4. Discussion

AMR remains a major public health challenge, particularly in low-resource regions such as Western Sub-Saharan Africa, including Nigeria. Limited data on AMR from these settings underscore the need to strengthen microbiology laboratory capacity and surveillance systems [29]. This study addressed this gap by assessing the prevalence of CR-GNB in multispecialty private hospitals—a sector where AMR data is scarce. Our study found that 29% of GNB isolates were CR-GNB, higher than the previously reported national mean of 20.5% in Nigeria [8]. This rise may be linked to the nature of multispecialty private hospitals, where antimicrobial prescribing is often less regulated and complicated infections are commonly treated. The significant differences in CR-GNB prevalence across Nigeria highlight how factors such as patient selection, hospital type, and laboratory methods affect resistance estimates, underscoring the need for improved surveillance in private facilities.

Our study highlights notable differences in the distribution of CR-GNB versus non-CR-GNB isolates across clinical settings. Although the overall outpatient proportion was not statistically different (p = 0.196), the CR-GNB isolates were significantly less common in general medicine departments in the outpatient setting, suggesting a lower risk among outpatients with less severe conditions. This finding aligns with previous studies, which found higher rates of multidrug-resistant organisms in inpatient and critical care settings, driven by increased antimicrobial exposure and the use of invasive procedures [30]. Consistently, a significantly higher prevalence of CR-GNB was observed in ICUs and oncology departments (p = 0.009), where patients are more vulnerable due to immunosuppression, prolonged antimicrobial use, and device-associated care [31,32]. The increased use of carbapenems—often as last-resort agents—further contributes to selective pressure in these high-risk units [32]. Notably, 80.3% of the isolates belonged to outpatients, which limited inpatient representation. The small number of inpatients may lead to an underestimation of CR-GNB burden in hospitalized populations. Future studies should thus ensure a more balanced sampling to better reflect AMR in inpatients, particularly in Nigeria, where surveillance of CR-GNB remains limited.

In our data, patients with CR-GNB (38.5 years, IQR 29.8–49.0) were slightly older than those non-CR-GNB (36 years, IQR 29–45), although most were still young adults. This profile is modestly younger than a hospital-based cohort from South-west Nigeria (median 42 years, IQR 29–64), in which the highest CRE proportion occurred among adults aged 40–54 years [9]. By contrast, consolidated national surveillance of CRE bloodstream infections (BSI) in South Africa described a younger case mix overall (median age ~31 years) [33]. Finally, a recent ICU study from Kenya did not report a median age but found that nearly half of ICU patients with GNB infections were between 40 and 60 years old, consistent with an older inpatient case mix typical of critical care [34]. Taken together, our median age lies between Nigerian inpatient cohorts and South African CRE BSI surveillance, and the differences likely reflect setting (outpatient vs. inpatient/ICU), case mix, and study focus (all GNB vs. CRE from bloodstream infections). Careful comparison across studies should therefore account for these design and population differences.

Consistent with reports from Nigeria, the predominant CR-GNB in our study were Escherichia coli and Klebsiella pneumoniae [8]. Our observed resistance rates were higher than those reported in several West African countries but lower than rates from Southern Africa and Egypt [35]. The differences observed may be attributed to variations in study design, patient selection, hospital type, and laboratory methods. Additionally, these variations could be influenced by local healthcare practices, including antimicrobial prescribing behaviors, referral patterns, and the implementation of stewardship programs. Further research is necessary to clarify how these contextual factors shape regional resistance patterns.

Using carbapenemase gene detection, we identified 10 CP-GNB, representing 3.15% of all the GNB—higher than the 2.7% previously reported in Nigeria [36]. These isolates exhibited broad resistance across antimicrobial classes, underlining the role of carbapenemases in AMR [4,16,37]. Moreover, we detected MDR, XDR, and PDR phenotypes in the identified CP-GNB, thereby posing substantial challenges for clinical management and AMR surveillance [4,29]. Notably, CRE, CRAB, and CRPA exhibited resistance to last-line agents, such as piperacillin/tazobactam, ceftolozane/tazobactam, ceftazidime/avibactam, and cefepime. Although these antimicrobials typically demonstrate high efficacy against GNB—with susceptibilities up to 99% among Enterobacterales [38,39]—their effectiveness may be limited in difficult-to-treat infections. For instance, susceptibility among difficult-to-treat P. aeruginosa was reported to be under 20% [40]. Despite not specifically assessing such infections, our observed high resistance rates raise concerns.

Among CP-GNB, the predominant carbapenemase genes were bla_NDM and bla_KPC, differing slightly from previous Nigerian data that highlighted bla_NDM and bla_VIM [8]. The bla_NDM genes, particularly bla_NDM-1, are plasmid-mediated and highly adaptable, indicating widespread horizontal transmission among bacteria. Additionally, the absence of a standardized test for metallo-beta-lactamases leads to many unrecognized asymptomatic carriers [41]. A recent study in Lagos found that plasmids harboring carbapenemase genes shared structural similarities with those from Asia, Australia, and Europe, suggesting active global transmission [4]. While bla_KPC was detected exclusively in K. pneumoniae, it remains uncommon in Nigeria and was first reported in 2015 [42]. The gene’s association with plasmids enhances its epidemic potential and limits its treatment options [43].

Despite A. baumannii’s clinical relevance, Nigerian healthcare settings are limited in their capacity to detect it [15]. A. baumannii often exhibit high resistance to essential antimicrobials [17], with a recent prevalence of Carbapenem-resistant A. baumannii (CRAB) carrying the bla_NDM gene at 27.9% [44]. We identified two A. baumannii from one of the hospitals, albeit smaller in other studies. However, these recovered A. baumannii were bla_NDM-carrying CRAB, indicating a 100% prevalence among these organisms in our sample. Our findings align with the propensity of A. baumannii to develop XDR. These two organisms were susceptible to amikacin and colistin, which is consistent with the findings of a study in public tertiary hospitals in Nigeria [17]. Moreover, our findings support the recommendation of treating CRAB with a combination of amikacin and colistin or minocycline with cefoperazone-sulbactam [31]. A study in China suggests that a combination of amikacin, polymyxin B, and sulbactam can effectively combat MDR A. baumannii [45].

Although A. baumannii was susceptible to amikacin and colistin, our study showed that A. baumannii still retains the potential to disseminate bla_NDM within healthcare settings and environments, leading to potential outbreaks [4,46].

Only one death was reported for the 30-day all-cause mortality outcome among patients with infections caused by CP-GNB in this study. The organism isolated from the patient’s sample was a PDR E. coli, which caused bloodstream infection. The patient had undergone transsphenoidal surgery for benign neoplasia and was admitted to the ICU. Although the patient’s underlying condition and surgical complications may have contributed to the fatal outcome, the presence of a PDR bloodstream infection likely exacerbated the clinical course, consistent with previous reports on the high risk associated with bacteremia [47]. Furthermore, the mortality observed in this study was lower than the rates reported in larger cohorts from Africa, which are approximately 37.2% [15]. This discrepancy is likely due to the small sample size and the predominance of outpatient cases among our isolates. As a result, the actual mortality burden of CR-GNB in hospitalized patients may be underestimated in our findings. Nevertheless, this fatal case illustrates the potential severity of CP-GNB infections and underscores the importance of early detection, infection prevention, and antimicrobial stewardship interventions.

Our study has limitations, particularly in the context of antimicrobial susceptibility testing. Species identification in this study relied on API 20E and API 20NE kits, supplemented by morphology, Gram staining, oxidase testing, and quality control strains. Although widely used in clinical laboratories, these methods have limited accuracy, particularly for distinguishing Enterobacter and Acinetobacter [48,49]. The lack of molecular confirmation methods, such as 16S rRNA sequencing or MALDI-TOF mass spectrometry, is a notable limitation of this study. Due to resource constraints, these methods could not be utilized. MALDI-TOF mass spectrometry and 16S rRNA sequencing provide greater reliability: the former achieved 100% concordance in identifying Acinetobacter with an updated database [50,51], while both methods demonstrate superior speed and precision in routine identification of Enterobacter [52]. We employed the EUCAST method to test carbapenem resistance via agar disk diffusion and the CLSI method to determine the minimum inhibitory concentrations of presumptive CP-GNB. Changes to the standard guidelines may thus impact our AMR assessments. Furthermore, classifying isolates with intermediate inhibition zones as non-resistant may underestimate the actual resistance rates. Another key limitation is that we relied solely on resistance to imipenem or meropenem to identify potential carbapenemase-producing organisms—we might have thus missed cases of low-level resistance. This is due to the variable expression of the carbapenemase gene, porin mutations, or efflux mechanisms [53]. Finally, the short study duration may have underestimated the prevalence of CR-GNB. The achieved sample size (317 vs. 326) was <3% below the target, increasing the margin of error only marginally (4.5% to 4.6%). Such a minor shortfall is unlikely to affect the validity of the findings and is generally acceptable in biomedical research when acknowledged as a limitation [54].

5. Conclusions

Our study offers key insights into the characterization of CR-GNB and CP-GNB in multispecialty private hospitals in Lagos, Nigeria. P. aeruginosa and E. coli were found to be the predominant carbapenem-resistant species, with a high prevalence of extended drug-resistant strains among CP-GNB. Moreover, the frequent detection of bla_NDM along with bla_KPC underscores the need for tailored therapeutic strategies. Accurate detection—beginning with screening for ertapenem, meropenem, or imipenem resistance and confirmed by phenotypic and genotypic methods—is essential for timely treatment and infection control. Given the evolving nature of AMR in healthcare settings, continuous surveillance, antimicrobial stewardship, and robust infection prevention measures are imperative. Our findings support the urgent need for government-led interventions that target the private healthcare sector as part of Nigeria’s broader AMR response.