Prevalence of Extended-Spectrum β-Lactamase Producing Enterobacteriaceae in Community Patients in Blantyre, Malawi

Onduru G. Onduru (  ogyonduru@yahoo.com ) Department of Pathology, College of Medicine, University of Malawi, Private Bag 360, Blantyre, Malawi. Rajhab S. Mkakosya Department of Pathology, College of Medicine, University of Malawi, Private Bag 360, Blantyre, Malawi. Said Aboud Department of Microbiology and Immunology, Muhimbili University of Health and Allied Sciences, P.O. Box 65001, Dar es Salaam, Tanzania Susan F. Rumisha Directorate of Information Technology and Communication, National Institute for Medical Research, P.O. Box 9653, Dar es Salaam, Tanzania.


Introduction
Extended-spectrum β-lactamase producing Enterobacteriaceae (ESBL-E) were exclusively observed in hospital settings in early years after the introduction of Cephalosporins in 1990s [1,2]. They continue to be a major global health challenge in both clinical and infection prevention control. Currently global problem of antimicrobial resistance (AMR) resulting from the production of ESBL enzymes by Enterobacteriaceae has increased to an alarming magnitude and is among the leading threats to human health [3]; the most affected are the developing countries with limited resources to implement control measures to curb AMR [4]. Increased spread of ESBL-E is not only occurring in hospital settings but also in the communities. The prevalence of community acquired ESBL-E worldwide remained below 10% before 2008 but has rapidly exploded to over 60% in recent years causing a major challenge for antibiotic therapy to both nosocomial and community acquired infections [5,6] In low income countries, lack of enforcement of regulations and guidelines controlling the sale and irrational use of antibiotics in both animal and humans contributed to the emergence and widespread of resistant bacteria [7]. The problem of AMR as a result of failure to treat bacteria that are resistant to locally available antimicrobial agents is implicated in the longer hospitalization and the choice for more expensive drugs which are often not readily available in the resource constrained countries like Malawi [8,9].
Few available studies conducted in Malawi to account for the presence of ESBL producing Enterobacteriaceae in humans were based on observational study and surveillance data from hospital settings [10][11][12]. In 2005, prevalence of ESBL producing Enterobacteriaceae reported after baseline review of ESBL problem and characterization of ESBL enzymes from clinical isolates following the death of patients attributed to multidrug resistant Klebsiella pneumoniae infection was found to be 0.7% [11]. Twelve years later (2017), the study describing longitudinal trends in AMR at a teaching hospital in Blantyre Malawi that focused on prevalence of ESBL resistance among Gram-negative bacteria causing blood stream infections found that the magnitude of ESBL producing Enterobacteriaceae rose in Escherichia coli from 0.7 to 30.3%, 11.8 to 90.5% in Klebsiella spp and 30.4 to 71.9% in other Enterobacteriaceae [12].
However, the current status of ESBL producing Enterobacteriaceae burden in the community settings in Malawi and Blantyre in particular is lacking. As an evidence of the level of ESBL producing Enterobacteriaceae carriage in the communities in Blantyre, ESBL producing E. coli has been recovered from a stool of a patient with neither history of admission nor travel history in six months [13].Therefore, the current study was conducted to provide baseline prevalence and associated factors for carriage of ESBL β-lactamase producing Enterobacteriaceae in community patients in Blantyre, Malawi.

Material And Methods
Study design, duration and location This was a cross-sectional study carried out between March and September 2020 at three randomly selected outpatient health centres representing community settings in Blantyre, Malawi ( Figure.1). Outpatient health centres were chosen as a point of data collection due to conveniences in sample collection and storage and these were Limbe, Zingwangwa and Ndirande health centres. These health centres are the main public health care facilities providing outpatient services to a population of over 800,264 of Blantyre district.
Study population, sample size and sampling Study participants comprised of 300 outpatient adults from three main public health centres representing community settings in Blantyre. Sample size was calculated using a formula described by Lwanga and Lemeshow [14] at 95% con dence interval based on 61.9% prevalence of ESBL producing Enterobacteriaceae in Blantyre as previously reported [12]. All adult patients (≥18 years old) who presented on the day of data collection had equal chances to be included in the study. Patients who were willing and consented to take part in the study were randomly recruited in the study regardless of their reason to seek medical care at the health centres.

Specimen collection
Clinical specimens (urine and rectal swabs) were collected and handled following standard protocol by registered and experienced health personnel working at each health centre in accordance to ethical principles for medical research involving human subjects. Only one type of clinical specimen was collected from each participant. Urine samples were collected exclusively from outpatients who presented with UTI related complains using sterile urine-cup and rectal swabs were collected using Amies ocked swabs (collection and transport system) (COPAN, Brescia-Italia). Collected urine and swabs were transported to Microbiology Laboratory, College of medicine, University of Malawi for culture of ESBL producing Enterobacteriaceae.

Culture of ESBL producing Enterobacteriaceae
Culture of ESBL producing Enterobacteriaceae was performed on CHROMagar TM ESBL medium supplemented with ESBL supplement containing a selective mixture of antibiotics enabling selective growth of ESBL producing Enterobacteriaceae and inhibiting growth of non-ESBL Enterobacteriaceae (CHROMagar, Paris, France). Samples were processed by direct inoculation onto CHROMagar TM ESBL plates using streaking and spreading techniques followed by cover bottom side incubation in aerobic conditions at 37°C for 18-24 hours. Following incubation, a signi cant growth of ESBL producing Enterobacteriaceae and appearance of the colonies were observed.
Phenotypic con rmation of ESBL production by Enterobacteriaceae Phenotypic con rmation of ESBL production by the isolates was done using a combination disk test method (CDT) by comparing inhibition zone diameter around a cephalosporin disks to that of the same cephalosporin plus clavulanate following the recommendations of Clinical and Laboratory Standards Institute (CLSI) 2017. In the current study, we used MAST combination disks (MAST D52C ESBL, Mast Diagnostics, Merseyside, UK) to phenotypically con rm ESBL production in Enterobacteriaceae isolates with no chromosomal de-repressed or inducible AmpC that grew on CHROMagar TM ESBL medium. Both cefotaxime ((CTX-30μg) and ceftazidime (CAZ-30μg) antibiotic disks with and without clavulanic acid (CA-10μg) were used concurrently based on comparing the zones of cefotaxime and cefotaxime plus clavulanate and ceftazidime disk with and without clavulanic acid. The increase in zone diameter of ≥5 mm or a zone expansion of 50% i.e. corresponding to a two-fold dilution between the zone of inhibition of a single disk and in combination with clavulanic acid was indicative of ESBL production as previously described [15].
Quality control ESBL producing Klebsiella pneumonia (ATCC 700603) and non-ESBL producing E. coli (ATCC 25922) were used as positive and negative control, respectively.

Identi cation of Enterobacteriaceae species
Presumptive identi cation of common ESBL producing Enterobacteriaceae isolates was done based on colony color characteristics of bacteria growth on CHROMagar TM plates according to manufacturer's instructions; i.e. ESBL producing Escherichia coli was identi ed by dark pink color, ESBL producing KEC (Klebsiella, Enterobacter, Citrobacter) by metallic blue +/-reddish halo), ESBL producing Proteus by brown halo, ESBL producing Acinetobacter by cream color and ESBL producing Pseudomonas was translucent (+/-natural).
The identity of isolates were subsequently con rmed using the commercially acquired biochemical substrate strips (Microbact™ gram negative identi cation system, Oxoid, GNB 12A). The standardized mirco-subsrate strips (Microbact TM ) were inoculated according to the manufacturer's instructions for identi cation of Enterobacteriaceae. The biochemical tests used were Lysine, Ornithine, H 2 S, Glucose, Mannitol, Xylose, ONPG, Indole, Urease, VP, Citrate and TDA. The interpretation to identify the isolates was done using the Microbact™ computer aided identi cation package (Oxoid) in combination with the Cowan and Steel's Manual for the Identi cation of Medical Bacteria [16] Statistical analysis Data obtained were cleaned and transferred to STATA version 12.0 (Stata Corp LP, College Station, USA) for statistical analysis. Descriptive summary statistics were generated as frequencies and proportions presented in tabular form. Chi square and Fishers Exact tests were used to compare dichotomous variables as appropriate. Univariate association between ESBL-E positivity (outcome) and independent variables was determined by logistic regression analysis. When tting the model, participants who had separated, divorced, widow and single marital status were combined to obtain single variable (unmarried) and was compared with married or cohabiting participants. Results were presented as odds ratio and 95% con dence interval. A p-value ≤0.05 was regarded as statistically signi cant.

Results
Social demographic and clinical characteristics of the study population A total of 300 adult patients attending outpatient clinics in Blantyre district were enrolled in the study. The median age of participants was 29.5 years (IQR=23-38; range, 18-75 years). Participant's male to female sex ratio was 1:1.2. Majority of participants (41.67%) were between the ages of 18-27 years, married or cohabiting (61.33%), had primary education (44.33%) and were unemployed (46%). In the past three months, more than ninety seven percent of participants had no history of admission, 90.33% had no history of surgery and 78% had not used antibiotics (Table 1). Of the 50 ESBL producing Enterobacteriaceae isolates recovered from outpatient adults in Blantyre, 84% were from rectal swabs. The most abundant ESBL-E species isolated were Escherichia coli (66%) followed by Klebsiella spp (8%), Yersinia enterocolitica (6%), Enterobacter lwo i (4%) and others ( Table 2). Community prevalence of ESLB-E was higher in male 56% (28/50) compared to female 44% (22/50); however, this difference was not statistically signi cant X 2 =2.5821, p=0.11. Prevalence of ESBL-E was higher in married or cohabiting 56% (28/50), unemployed 48% (24/50) and those with primary education 46% (23/50). Community patients who lacked history of surgery and history of antibiotic use in the past three months 88% (44/50) and 80% (40/50) respectively had high rate of ESBL producing Enterobacteriaceae carriage compared to those who underwent surgery 12% (6/50) and who used antibiotics in previous three months 20% (10/50). We found that all community patients with ESBL-E carriage had no history of admission in the last three months (Table 1). The summary of unadjusted logistic regression model performed to establish the relationship between patient characteristics and carriage of ESBL-E in community patients is presented in Table 3. Neither demographic characteristics nor clinical characteristics of participants showed any degree of association with the carriage of extended spectrum β-lactamase producing Enterobacteriaceae.

Discussion
Despite the rise in extended spectrum β-lactamase producing Enterobacteriaceae prevalence worldwide, very few reports of community ESBL-E are available from Africa. Less commonly, detection of ESBL Enterobacteriaceae has been carried out from hospital settings and rarely from the community settings in developing countries including Malawi [10,12,[17][18][19]. In this study, we demonstrated moderate prevalence and lack of factors associated with extended-spectrum β-lactamase producing Enterobacteriaceae in community settings in Blantyre, Malawi. Examining the status of ESBL producing Enterobacteriaceae burden and factors associated with contracting strains of ESBL-E in community settings could provide important information that is a requisite for formulation and implementation of strong infection and antimicrobial use control strategies in both community and hospital settings.
Moderate prevalence found in the present study are similar to ndings of low to moderate prevalence of ESBL-E in the community settings that were reported in Central Africa republic [20] and Kenya [21,22]. Contrary to higher prevalence of ESBL-E that was previously reported in hospital setting in Blantyre [12], moderate prevalence observed in this study may be an indication that community acquired ESBL-E is still a minor problem in Blantyre, Malawi. Elsewhere it has been reported that higher rates of antibiotic consumption accelerate both AMR selection and higher rate of nosocomial ESBLs in hospital settings compared to community settings [23,24], this can account for the moderate rate of ESBL-E detected in community patients in the current study.
We found a preponderance of ESBL-E among males compared to females, but there was no statistically signi cant difference between ESBL carriage in males and females. This nding is in conformity to ndings of the study by Shah et al. [25] which reported ESBL positive isolates mostly in males (65.33%) compared to females (34.67%) when relating age and gender to extended-spectrum β-lactamases in Enterobacteriaceae. Another similar results were obtained in Bhopal region of Central India where 52.54% of ESBL isolates were obtained from males while 43.46% from females [26]. However, our ndings differ from the ndings of most studies conducted in hospital settings [12,[27][28][29]. It has been reported that urinary tract infections mostly affecting women and samples of urine are the common sources of bacteria isolates in most microbiology laboratories in different countries [30], whether sample type (rectal vs urine) has something to do with male/female differences in ESBL-E carriage is worth another study.
In the current study, all patients with ESBL-E carriage had no history of hospital admission and there was higher prevalence of ESBL-E in community patients who lacked history of surgery and history of antibiotic use in the past three months, these ndings indicated that there was community acquisition of ESBL-E in Blantyre Malawi but at a moderate level.
Similar to studies that reported E. coli as the most frequent Enterobacteriaceae isolates in Ethiopia [31], Burkina Faso [32], Netherlands [33], Uganda [34] and India [26,35], the current study showed that Escherichia coli were the most abundant Enterobacteriaceae species harboring extended-spectrum βlactamase followed by Klebsiella spp in community patients in Blantyre. Probably, the reason for E. coli to contribute more than 60% of all isolates in the current study could be the community setting and the type of sample taken. Like our study in which rectal samples constituted greater number of clinical samples than urine, E. coli has mostly been isolated from faecal matter compared to urine [36]. Contrary to the current study where participants with ESBL-E carriage were adult community patients who did not report any history of hospital admission in previous three months, other Enterobacteriaceae species speci cally Klebsiella spp have been found to cause a substantial morbidity among paediatric patients accounting for almost more than 50% of all Gram-negative infections in neonates and a signi cant burden of hospital-acquired infections in sub Saharan Africa [30].
It has been highlighted in the past that the risk factors for introduction of ESBL-E into the community includes travel to areas with a higher prevalence of ESBL pathogens, previous hospitalization, antibiotic treatments, old age, comorbidities like diabetes and previous infection by members of Enterobacteriaceae [37][38][39]. However, our study found that there was no relationship between ESBL-E carriage in community patients and their demographic or clinical characteristics similar to the study in The Gambia which found that most of the demographic characteristics had no strong association with the carriage of ESBL producing Enterobacteriaceae [40]. Lack of association between ESBL-E carriage and demographic or clinical characteristics of patients in this study may be due to the fact that all patients with ESBL-E carriage in the current study had no history of admission and few had history of prior antibiotic use in the past three months which are the main predictors of ESBL acquisition.

Limitations
In this study we used both cefotaxime and ceftazidime disks to con rm ESBLs and this was important because ESBLs could be missed if just a single disk was used. However our study could be limited by the accuracy and discriminatory power of the medium used during initial screening of ESBLs between bacteria producing these enzymes and those with other mechanisms of resistance to beta-lactams e.g., inhibitor-resistant β-lactamase and cephalosporinase overproduction. In addition to that, we were not able to perform genotypic characterization to determine the presence of ESBL genes in Enterobacteriaceae recovered from community patients.

Conclusion
Our ndings revealed moderate presence of ESBL-E phenotypes in community patients, participants with ESBL-E carriage lacked history of hospital admission three months before the study. This can suggest community acquisition of ESBL-E in Blantyre, Malawi and should these carrier patients in the community remain untreated, they may serve as a community reservoir of resistant pathogens potential for the transmission and spread of community acquired ESBLs. Low prevalence of ESBL-E in community settings in Blantyre Malawi observed in this study can be maintained if strong infection and antimicrobial use control strategies are to be implemented. Before enrolment into the study, a written informed consent was obtained from participants using a form translated in local language of Chichewa. For the patients without formal education attended, informed consent was obtained from legally authorized representatives.

Consent for publication
Not applicable Availability of data and materials The dataset used and/or analysed in the current study are available from the corresponding author on reasonable request the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.