The misuse of antimicrobial agents has been identified as one of the major forces resulting in the rapid spread of resistance, but the nature of this relationship is complex. Resistance to antimicrobials is sometimes brought about by a change in the bacterial genome, which can occur by the transfer of antimicrobial resistance genes which may be found in some of the bacteria found in the environment and transferred to those without the resistance genes and also through other horizontal gene transfer elements [1] . Products that are used in disinfection and sterilization, as well as heavy metals used in industries and households along with antibiotics, creating selective pressure in the environment that lead to mutations in microorganisms [2] .

Indiscriminate waste dumping enhances the breeding of microorganisms that pose a danger to the human population. Urban wastes contain a wide range of things that may have come from different sources e.g. hospitals, animal sheds and may carry antimicrobial resistance bacteria belonging to the human and animal commensal flora, mainly Enterobacteriaceae [3] . Waste degradation is enhanced by the presence of soil microorganisms that create a conducive environment for the resistant bacteria e.g. Salmonella species and E. coli to thrive in thus becoming potential human pathogens and may cause severe health hazards [4] . The presence of rodents in these dumpsites and garbage collection areas enhance the spread of antimicrobial resistant bacteria to other areas. Previous studies have focused on the identification of these disease vectors on dumpsites and have reported cockroach, housefly, black garbage fly, and stable fly to be the most prevalent disease vectors on the dumpsites and garbage collection areas [5] [6] .

Enterobacteriaceae are one of the major causes of infections and deaths around the world. The prevalence of antimicrobial resistance in this family of bacteria e.g. Escherichia coli, Salmonella and Shigella has raised over the years. One of the major reasons for the increase is the spread of Klebsiella pneumoniae carbapenemase (KPC), a class A serine carbapenemase that was originally isolated from Klebsiella pneumoniae in 1996 [7] .

Resistance in pathogenic organisms poses a distinct clinical challenge. However commensal bacteria may enhance the spread of resistant bacteria as they may act as reservoirs of the resistance genes which may have been acquired through various horizontal gene transfer elements. This, therefore, increases the carriage levels of resistant organisms. The resistant microorganisms from clinical samples may gain their entry into the environment through fecal contamination and various wastes from the hospital that has not been sterilized properly. Escherichia coli and Klebsiella species were used in this study as an indicator species as they are commonly associated with humans and animals disease and have also been used in other studies to gauge the spread of acquired resistance [8] .

Dumpsites near schools, residential areas and the municipal general waste dumping sites in Nairobi area were selected for the study and sampled. In order to verify the most accessed area of these dumpsites and garbage collection areas by the street families and other people relying on dumpsites for a living, a qualitative survey of the dumpsites and garbage collection areas was conducted. The purpose of the survey was to determine the most appropriate area to sample for the main study. This was done by visiting the dumpsites and garbage collection areas before the start of the study and surveying the areas and identifying potential barriers to our study.

A total of 17 dumpsites (permanent dumping area) and garbage collection areas (temporary dumping area where garbage is dumped awaiting collection) were randomly sampled in different parts of Nairobi area (Figure 1).

Sample collection was randomly carried out in different days in seventeen different points. A total of 126 samples were collected during the sampling period. At each sampling station, the sub-surface soil, mixed solids, leaking water, stagnant water, swabs and food samples were collected from one squire foot area into sterile sampling bottles and appropriately labeled. Six samples were collected from each site. The samples were then transported to the laboratory for analysis.

One gram (1 g) of each solid samples and 1 mL of the liquid samples and swabs were suspended in 10 mL physiological sterile saline. Serial dilutions of 10 fold, 5 fold, and 1 fold dilutions were prepared from the 10 mL suspension and transferred onto duplicate molten Plate Count Agar (PCA) mixed and allowed to cool at room temperature. This was then incubated at 37˚C for 24 hours. Colonies were determined from duplicate plates and the average counts recorded as mean viable bacteria (colony forming units [CFUs] of the sample. The low and high CFUs were reached by dividing the dumpsites and garbage collection areas into two, those that had CFUs above 5.0 were considered to be high and those below 5.0 considered being low.

A loop full (1 µl) of the mixture incubated in buffered peptone water was then transferred onto MacConkey agar plates and incubated at 37˚C for 24 hours for isolation of E. coli and Klebsiella species. After overnight incubation the plates were then examined for growth and presumptive identification of E. coli and Klebsiella species for lactose-fermenting colonies (pink). Presence of pink non- mucoid colonies for E. coli and pink mucoid for Klebsiella species were further identified by biochemical tests.

Pure discrete colonies of E. coli and Klebsiella species from primary cultures were identified by gram staining and biochemical tests. Colonies that appeared as gram negative rods were tested for indole test, methyl-red test, Voges- Proskauer test and citrate utilization (IMViC). Presence of E. coli was interpreted as a positive reaction for indole and methyl-red tests and a negative reaction for Voges-Proskauer test and lack of citrate utilization. Additionally presence of Klebsiella species was interpreted as a negative reaction for indole and methyl-red tests and a positive reaction for Voges-Proskauer test and citrate utilization [9] .

The antimicrobial susceptibility testing was done on 286 isolates on Mueller- Hinton agar plates (Oxoid). The isolates were tested against the following antibiotics; Ampicillin (10 μg), Cefpodoxime (10 μg), Ceftazidime (30 μg), Cefoxitin (30 μg) Cefepime (30 μg), Amoxicillin-Clavulanic acid (10/100μg ratio), Ciprofloxacin (10 μg), Tetracycline (30 μg), Trimethoprim Sulfamethoxazole (30 μg), Gentamicin (10 μg), Chloramphenicol (30 μg), Streptomycin (25 μg), Nalidixic acid (10 μg), and Meropenem (10 μg). The plates were then incubated at 37˚C for 18 - 24 hours. The inoculums for susceptibility testing were compared against the McFarland 0.5 turbidity standards with E.coli ATCC 25,922 strain being used as the control standard for quality assurance of media and the antimicrobial discs. The interpretation of results was according to Clinical and Laboratory Standards Institute guidelines [10] .

The lowest CFU from any given sampling point was 1.1 × 10¹¹ that was recorded in Dandora dumpsite. The site with the highest CFU value was Muthurwa estate dumpsite that recorded 8.2 × 10¹⁰. Other sites with high CFU counts were Umama garbage collection area (Komarock), Kawangware market dumpsite, Kenyatta staff quarter garbage collection area, Kweria garbage collection area, City market garbage collection area, Central police garbage collection area, Kibera dumpsite, and Kenyatta market dumpsite while Seven of the dumpsites and garbage collection areas (Ayany dumpsite, Dandora dumpsite, Ngara market garbage collection area, Muthurwa market garbage collection area, Masai market dumpsite, Mareba garbage collection area (Kibera), District Commissioner garbage collection area (Kibera) recorded CFUs below 2.0 × 10¹⁰ Table 1.

Resistance pattern to all antimicrobials for E. coli and Klebsiella was above 5% except for ciprofloxacin (3.3% E.coli, and 2.4% Klebsiella), meropenem (1.7% E.coli and 1.8% Klebsiella) and gentamicin (3.3% E.coli and 0% Klebsiella) Table 2.

In general, resistances frequencies were similar for E. coli and Klebsiella obtained from samples with high CFUs to those obtained from samples with low CFUs. In the dumpsites and garbage collection areas that had high CFUs, such as Muthurwa estate dumpsite, Central police garbage collection area, City market garbage collection area, City park market dumpsite Figure 2(a), there were high resistance prevalence’s of above 25% to streptomycin, ampicillin, tetracycline and trimethoprim sulfamethoxazole for isolates belonging to both species. In contrast, there was low resistance to meropenem, gentamicin, and ciprofloxacin (≤5%) in both species. The study also found that 42% of E. coli isolates were resistant to ampicillin compared to 59% of Klebsiella isolates found in this study.