A total of 103 gram-negative bacteria isolated from various units of Evercare Hospital, previously called Apollo Hospital, Dhaka, were used for this study. These included 1) 13 isolates from outpatient (OPD) 2) 43 isolates from inpatients (IPD) 3) 47 isolates from the Intensive Care Unit (ICU). The ratio of infected males to females, as well as the infection rate in various age groups were recorded accordingly. Bacterial samples were isolated from different clinical specimens including Urine, Tracheal aspirate, Sputum, Blood, ET tube tips, Pus, Suction tips, Wound swabs, Foley catheter tips, Silicon catheters, and CVP tips. All experimental methods were carried out in accordance with guidelines and regulations approved by the “Research and Ethical Practice Committee” of Evercare Hospital, Dhaka, Bangladesh. Oral informed consent was obtained from the study subjects/patients or from their legal guardians and was approved by the “Research and Ethical Practice Committee” at Evercare Hospital, Dhaka, Bangladesh and the data were analyzed anonymously.

All these specimens were inoculated on appropriate culture media (MacConkey agar, Blood agar, Chocolate agar and only for urine samples HiCrome UTI agar was used) and incubated for 24 to 48 hours at 37˚C. After incubation, organisms were identified by standard microbiological procedures (Colonial morphology, Gram stain appearance, oxidase test, triple sugar iron test, motility test, urease test, citrate test and indole test).

Antibiotic susceptibility tests were performed using the Kirby-Bauer disc diffusion techniques according to CLSI guidelines [15] . Inoculates were prepared by suspending the isolates in normal saline equal to the turbidity of 0.5 McFarland turbidity standard and applied on Mueller-Hinton Agar plates. All antibiotic discs were obtained from Oxoid, UK. Antibiotic discs were used depending on the types of microorganisms and specimens (Amoxiclav 30 mcg, Ceftriaxone 30 mcg, Cefepime 30 mcg, Ceftazidime 30 mcg, Imipenem 10 mcg, Meropenem 10 mcg, Polymyxin 300 mcg, Amikacin 30 mcg, Gentamicin 10 mcg, Netilmicin 30 mcg, Tetracycline 30 mcg, Tigecycline 15 mcg, Ciprofloxacin 5 mcg, Levofloxacin 5 mcg, Cotrimoxazole 25 mcg). These discs were incubated along with controls for 18 - 24 hours at 37˚C aerobically.

In the present study, Escherichia coli ATCC 25922 was used for quality control of the antimicrobial tests. Isolated gram-negative bacteria were subjected to an array of antibiotics and the results were interpreted as resistant or sensitive according to criteria set by the Clinical and Laboratory Standards Institute (CLSI). Gram-negative isolates with Imipenem (IPM) resistance were tested for detection of MBL by DPT and DDST method. A total number of 103 gram-negative bacteria showed carbapenem resistance in routine culture.

The bacterial isolates were sub-cultured on MacConkey agar media and incubated overnight at 37˚C. After overnight incubation, 2 - 3 isolated colonies of the organism were picked up from the subculture plate by a sterile inoculating wire loop and suspended in 3 ml sterile peptone water in a screw-capped test tube. The test strain was adjusted to the McFarland 0.5 standard. Within 15 minutes of adjustment of the density of the organism, streaking on Mueller-Hinton agar was done using a sterile cotton swab.

One disc containing Ceftazidime (CAZ) (30 mcg) and two IPM (10 mcg) was placed on the plate. The distance between every CAZ/IPM disc was kept at about 4 cm from center to center [16] . Two blank discs with 10 µl of 0.5 M EDTA added were placed near the CAZ/IPM disc, within a center-to-center distance of 2 cm. A sodium mercaptoacetate (SMA) (Metallo-β-lactamase SMA Eiken; Eiken Chemical Co., Ltd., Tokyo, Japan) disc was placed near the IPM disc respectively disc distance edge to edge 10 mm. The agar plate was incubated at 37˚C overnight [14] [17] .

In the Double Disc Synergy Test (DDST) with 0.5 M EDTA, the enhancement of synergistic growth inhibition zones between the CAZ/IPM disc towards the disc containing 0.5 M EDTA was considered positive for MBLs [16] . The presence of a synergistic inhibition zone of IPM with greater than 5 mm enlargement with the SMA disc side, was interpreted as positive [14] .

The bacterial isolates were sub-cultured on MacConkey agar media and after an overnight incubation at 37˚C, 2 - 3 isolated colonies of the organisms were suspended in 3 ml sterile peptone water. The test strain was adjusted to the McFarland 0.5 standard. Within 15 minutes of adjustment of the density of the organism, streaking on Mueller-Hinton agar was done using a sterile cotton swab.

Disc Potentiation Test (DPT) was done as per Yong et al. using IPM-EDTA [18] [19] . The test depends on comparing the zones given by disc containing IPM with or without EDTA modified the test by using two 10 mcg IPM disc and two 30 mcg Ceftazidime (CAZ) disc [19] [20] . A zone diameter difference between the IPM/CAZ and IPM/CAZ + 0.5 M EDTA disc with greater (>) than 7 mm was interpreted as a positive result.

Two discs containing 30 mcg CAZ and two discs containing 10 mcg IPM were placed on the plates. The distance between every CAZ/IPM disc was kept at about 4 cm from center to center. 10 µl of 0.5 M EDTA was added to one CAZ/IPM disc. The plate was incubated at 37˚C overnight. In this study, DPT with 0.5 M EDTA, the enlargement of the diameter of the growth inhibitory zone around CAZ/IPM + 0.5 M EDTA disc with greater than 7 mm, compared to CAZ/IPM alone was considered positive for MBLs.

For the statistical analyses, data were entered into an Excel spreadsheet and analyzed using the statistical software SPSS version 20.0 (IBM; Chicago, IL, USA) for Windows. Bivariate correlation test was performed. A p-value of <0.05 was considered as statistically significant.

A total of 103 gram-negative strains were isolated from different clinical specimens. All these bacteria isolated from the patients were subjected to an antibiogram test. Antibiotic susceptibility profile (n = 103) was analyzed for the six isolated bacterial spps. All 103 isolates showed 100% resistance to Imipenem, Meropenem, and Amoxiclav. We also investigated the distribution of resistant and susceptible isolates among male and female individuals. Results represented that the number of resistant isolates distribution was significantly high in male patients compared to the females (Table 1).

The percentage of bacterial isolates from male and female patient groups that showed 100% resistance to Imipenem, Meropenem, and Amoxiclav were 70% and 30% respectively (Figure 1(a)).

We also observed that Escherichia coli spp, Klebsiella spp, Pseudomonas spp, Acinetobacter spp, Proteus spp, and Providencia spp. showed a significantly higher

distribution among male patient group compared to female patient group (Figure 1(b)). In addition, our observation of the distribution of resistant bacterial isolates (n = 103) among age group data revealed that the “61 - 75” age group showed the highest number of resistant isolates. The age group also represented the highest resistant isolates from ICU hospital units compared to OPD, and IPD units (Figure 1(c)). We performed a comparative analysis to investigate the level of significance among the five different tests and the statistical levels of significance by Pearson correlation has been shown in supplement Table 1. We also investigated if particular sample types are useful for testing the susceptibility of clinically isolated etiological agents with antibiotics for proper management of the patients.

We analyzed the sample-wise distribution of bacterial isolates and found that Urine, followed by Tracheal aspirate and Wound swab samples represented the highest number of carbapenem-resistant isolates (Figure 2. and Table 2).

While all the bacteria isolated were found to be resistant to antibiotics Imipenem, Meropenem, and Amoxiclav, the highest carbapenem-resistant etiological agents isolated was Acinetobacter spp 40 (38.8%) followed by Pseudomonas spp 27 (26.2%), Klebsiella spp 26 (25.2%), Escherichia coli 8 (7.8%), Proteus spp 1 (1%) and Providencia spp 1 (1%) (Table 3).

Our investigation on the susceptibility of bacterial isolates (n = 103) from all age groups to antibiotics (Amoxiclav, Ceftriaxone, Cefepime, Ceftazidime, Imipenem, Meropenem, Polymyxin, Amikacin, Gentamicin, Netilmicin, Tetracycline, Tigecycline, Ciprofloxacin, Levofloxacin, and Cotrimoxazole) (Table 4).

Among the five different tests performed by DPT and DDST methods compared in this study, of 103 clinical isolates, 61 (59.2%) isolates showed sensitivity by disc potentiation test (DPT) method using imipenem with 0.5 M EDTA and 56 (54.4%) isolates ceftazidime with 0.5 M EDTA. In the DPT test method using imipenem with 0.5 M EDTA, the highest resistance to carbapenem strains carrying MBL enzyme found in IPD patients was 67% (29 out of 43) followed by ICU 53.2% (25 out of 47) and OPD 53.8% (07 out of 13) (Figure 3(a)).

The inflated number of carbapenem-resistant bacterial strains carrying MBL enzyme were found in age groups 0 - 15 years (80%, 4 out of 5) followed by 76 - > 90 years (76%, 13 out of 17) (Figure 3(b)). The proportion of males infected with MBL enzyme-producing resistant bacteria was higher than the proportion of females (44 out of 72) and (17 out of 31) respectively (Figure 3(c)). The highest percentage of MBL-producing strain was Escherichia coli (100%, 8 out of 8) (Figure 3(d)). Using ceftazidime with 0.5 M EDTA, the excessively high resistant strains to carbapenem carrying MBL enzyme were found in ICU patients (27 out of 47) followed by IPD (25 out of 43) and OPD (04 out of 13) (Figure 3(a))

and, the highest distribution was found in the age group “61 - 75” years (19 out of 34) (Figure 3(b)). The proportion of males infected with MBL enzyme-producing resistance bacterial strains was higher than the proportion of females (37 out of 72) and (19 out of 31) respectively (Figure 3(c)). The highest number of MBL-producing bacterial strains was Klebsiella spp (15 out of 26) followed by Escherichia coli (4 out of 8) (Figure 3(d)).

In the DDST method using imipenem and 0.5 M EDTA 43 (41.7%) isolates showed sensitive patterns followed by 38 (36.9%) imipenem and SMA and 15 (14.6%) isolates showed susceptibility to ceftazidime and 0.5 M EDTA. In this test method, using imipenem and 0.5 M EDTA, the highest resistant bacterial isolates to carbapenem strains carrying MBL enzyme were found in ICU patients (20 out of 47) followed by IPD (18 out of 43) and OPD (05 out of 13) (Figure 3(a)). The highest carbapenem-resistant bacterial isolates carrying MBL-enzyme were found in the age group 61-75 years (17 out of 34) (Figure 3(b)) and the proportion of males infected with MBL enzyme-producing resistant bacterial strains was higher than the proportion of female (30 out of 72) and (13 out of 31) respectively (Figure 3(c)). The highest percentage of MBL-producing bacterial strains was Escherichia coli (100%, 8 out of 8) (Figure 3(d)). On the other hand,

using imipenem and SMA, the highest number of bacterial isolates showing resistance to carbapenem-carrying MBL-enzyme were found in ICU patients (17 out of 47) followed by IPD (16 out of 43) and OPD (05 out of 13) (Figure 3(a)), and the highest distribution of carbapenem-resistant isolates carrying MBL-enzyme were found in age-group .. years (15 out of 34) (Figure 3(b)) while the proportion of males patients infected with MBL enzyme-producing bacterial isolates was higher than of female (25 out of 72) and (13 out of 31) respectively (Figure 3(c)). The highest number of MBL-producing bacterial strains was Escherichia coli (8 out of 8) (Figure 3(d)). Using ceftazidime and 0.5 M EDTA, the highest number of bacterial isolates showing resistance to carbapenem strains MBL-enzyme was found in ICU patients (07 out of 47) followed by IPD (07 out of 43) and OPD (01 out of 13) (Figure 3(a)). The highest distribution of carbapenem-resistant bacterial isolates carrying MBL-enzyme was observed in the age group 61-75 years (07 out of 34) and the proportion of males infected with MBL-enzyme-producing resistant bacterial strains was higher than that of females (12 out of 72) and (03 out of 31) respectively (Figure 3(b) and Figure 3(c) respectively), while the highest number of MBL-producing bacterial strain was Pseudomonas spp (8 out of 27) (Figure 3(d)).