Antibiogram Prole of Antibacterial Multidrug Resistance in Democratic Republic of Congo: Situation in Bukavu City Hospitals

Background Bacterial strains carrying multidrug resistance traits are gaining ground worldwide, especially in countries with limited resources. This study aimed to evaluate the spreading of multidrug-resistant bacteria strains in Bukavu city hospitals in the Democratic Republic of Congo. Methods We analyzed 758 antibiogram data recorded in les of patients consulted between January 2016 and December 2017 at three reference hospitals selected as sentinel sites, namely the Panzi General Reference Hospital (HGP), BIO -PHARM hospital (HBP), and Saint Luc Clinic (CSL). Of 758 isolates tested, the laboratories identied 12 bacterial strains in 712 isolates, of which 223(29.42%) presented MDR prole, including Escherichia coli (11.48%), Klebsiella pneumonia (6.07%), Enterobacter (5.8%), Staphylococcus aureus and coagulase-negative staphylococci (1.58%), Proteus mirabilis (1.85%), Salmonella enterica (1.19%), Pseudomonas aeruginosa (0.53%), Streptococcus pneumonia (0.4%)), Citrobacter (0.13%), Neisseria gonorrhea (0.13%), Enterococcus faecalis (0.13%) and Morganella morganii (0.13%). Infected patients were signicantly adults (73.1% vs. 21.5%) compared to children and mainly women (63.7% vs. 30.9%; p = 0.001).

treatment with a rst-line antibiotic, more expensive drugs should be used. Collaterally, prolonging illness and treatment increases health care costs, as well as the nancial burden on families and society [7,11]. Many factors triggering microbial resistance are known, mainly the misuse or irrational use of antibiotics. In 20 to 50% of cases, their use in humans is unnecessary, and in animals, it is questionable in 40 to 80% cases [12].
The increasing proportion of poor-quality generic drugs is a growing concern in the sub-Saharan Africa region [12][13][14].
There are multiple drawbacks to using poor-quality antimicrobial drugs; that lead to microbial resistance, treatment failure, exacerbation of the disease, and increased death rates. The WHO and the US Centers for Disease Control and Prevention (CDCP) have recognized the importance of studying the factors of emergence and risk of resistance and the need to establish control [3,8,15]. The MDR scoreboard itself in the Democratic Republic of Congo (DRC) is not clearly de ned. Epidemiological investigations are essential to monitoring the spread of MDR as in other countries with limited resources [16][17][18]. This study aimed to pro le the spread of MDR bacterial infections in Bukavu, a town in eastern DRC.

Study design and data collection
The study was a retrospective cross-sectional analysis of the antibiogram data recorded in the les of patients consulted between January 2016 and December 2017. Three reference hospitals were selected as sentinel sites, namely the Panzi General Reference Hospital (HGP), BIO -PHARM hospital (HBP), and Saint Luc Clinic (CSL). Laboratories performed antimicrobial sensitivity tests on blood, urine, cerebrospinal uid, and other samples, according to standard recommendations [19][20][21]. The susceptibility outcome was considered sensitive or resistant based on the inhibition diameter and according to data published in the 2017 CA-SFM recommendations.

Ethical considerations
All methods were carried out in accordance with relevant guidelines and regulations. All experimental protocols were approved by the research committee of the Faculty of Medicine (UOB). The clinical directors of the study hospitals allowed the collection of data. As retrospective study, informed consent was waived by institutional ethics review board of the O cial University of Bukavu. We assured for the con dentiality of individual patient information.

Data analysis
The SPSSv20 statistical software and Windows Excel 10 served to run descriptive statistics with a statistically signi cant difference set at p < 0.05; 95%, regarding the demographics of patients with MDR, the frequency of MDR strains in each hospital, and each biological sample.

Results
Identi cation and Prevalence of MDR strains Table 1 shows the total number (N) of each strain identi ed, the prevalence (%) of MDR and NMDR strains by column and row, as well as the ratio of MDR/NMDR. Of 758 isolates examined, laboratory technicians identi ed 12 bacteria present in 712 (93.93%) samples and did not identify the strains in 46 (6.07%) samples. Of those identi ed, 535 (70.58%) were NMDR, and 223 (29.42%) MDR. The column % shows that the most prevalent were EC (49.34%), KP (11.87%), EB (11.61%), and SA(11.61%). Individual MDR frequencies were EC (11.48%), KP(6.07%), EB(5.  Susceptibility of MDR isolates to antibiotics used in hospitals Table 2 shows the number of each strain exposed to a given antibiotic and, in parentheses, the percentage of susceptible strains.   This study aimed to contribute to such a need. The results identi ed 12 bacterial strains were in the samples taken from 758 cultures, mainly the EC strain. The pro les found here are comparable to those reported in other studies in Africa, the USA, and Europe [24], as illustrated below. Some isolates might even be XDR bacteria, but we did not separate them because the data did not come from a controlled study. The GLASS report [22] revealed that the most common MDR bacteria were Escherichia coli, Klebsiella pneumoniae, Staphylococcus aureus, and Streptococcus pneumoniae, followed by Salmonella spp. The 29.4% median rate of MDR strains found in Bukavu ranges between 23% in the USA and 37% in India. In the USA, a study [24] conducted in community hospitals revealed 23% of MDR pathogens, of which the three most common were SA(28%), EC(24%), and coagulase-negative staphylococci (10%); the infecting organism varied according to the place of acquisition. In Rwanda, the prevalence of MDR strains was 28% based on three primary data [26]. Studies in India found 37.1% MDR bacteria, 13.8% XDR and 0% PDR [2,25]. Despite the differences in the prevalence levels reported worldwide, the results support the alert on the increase in MDR bacteria around the world.
Aggregation of results by the hospital showed that more cases were from GHP (53.8%), followed by HBP (30.9%) and CSL (15.2%). The signi cant difference (p = 0.001) is only related to the size of each hospital. The pro le also indicated that most isolates were from sexually transmitted UTIs (61.8%) followed by skin infections (23.8%). The majority of people carrying MDR strains were adults (60%), as expected since children are less prone to UTIs. The high percentage of women (51%) is due to anatomical causes (proximity to the vaginal and anal openings), poor hygiene habits, sexual intercourse, and pregnancy [17,23].
The susceptibility pro les of the strains varied according to the classes of antibiotics. Regarding resistance to the betalactam category, the case of ampicillin and amoxicillin is striking. The test shows that these drugs are less effective against almost all strains. Many bacteria are also resistant to cephalosporins in this study. However, our previous research had shown that they are currently still widely prescribed and also used as self-medication in Bukavu [17,18]. A review article [16] reports that MDR (penicillin + two other classes) is 25% in Africa, 20% in Latin America, 12% in Eastern Europe, 18% in Western Europe, and 26 % in the USA. Data from bacterial resistance surveillance networks show that the distribution of 3rd generation cephalosporin-resistant Enterobacteriaceae species has increased signi cantly [22,27].
According to the authors, this resistance mainly concerns the production of extended-spectrum beta-lactamase (ESBL) and, to a lesser extent, plasma cephalosporinases (AmpC). For instance, the resistance of KP to third-generation cephalosporin is critical on a large scale in all WHO regions of the Americas, the Western Paci c, the Eastern Mediterranean, and the European Region [22]. Community-based infection with resistant E.coli producing extendedspectrum beta-lactamases is ubiquitous in Asia, the Middle East, South America, and parts of Europe [28].
Regarding aminoglycosides (AGs), gentamicin was more effective against EC (60%), resistant to PA, KP, and Salmonella. However, it is used mainly in combination with amoxicillin and azithromycin [18]. In the study by Bala et al. [29], no MDR isolate of gentamicin appeared. These in vitro results suggest that gentamicin may be an effective treatment option for MDR strains. In Bukavu hospitals, gentamicin is used mainly in combination with amoxicillin and azithromycin [18]. Parenteral administration of AGs, which limits their use as self-medication, partly explains their preserved e cacy. By far, the most common mechanism of resistance to AGs is the inactivation of these antibiotics by enzymes modifying their structure [30,31].
This study showed high resistance of many infections to second-generation quinolones. Only 25% of the 76 EC strains were susceptible to cipro oxacin, which backs what some studies reported in Asia and Africa [16,22,[32][33][34]. EC ST131 is a clone of MDR disseminated worldwide that presents resistance to uoroquinolones in addition to the production of ESBL CTX-M. EC ST131 strains tend to induce pyogenic liver abscesses and sometimes metastatic infections, including meningitis. The median resistance of SE Typhi to nalidixic acid is between 15.4-43.2% for pathogens isolated from patients with severe illness [28,[35][36][37].
Finally, the percentage of strains susceptible to meropenem was 41% for EC, 66.7% for SP, and less than 25% for the others, consistent with other studies. However, most clinicians consider carbapenems to be the class of choice for severe infections caused by ESBL-producing Enterobacteriaceae [34,35]. Carbapenems-resistance of PA is the most typical and frequent example of resistance induced by developing cell membrane impermeability [38]. Furthermore, the enzymatic inactivation of carbapenems is the most common resistance mechanism in A. Baumannii [32]. Carbapenem-resistant Enterobacteriaceae (CRE) represents an immediate threat to public health that requires urgent and aggressive action. Community-wide infections are likely to lead to a dramatic increase in the practical use of carbapenems [39,40]. A review article reported that the median prevalence of resistance to chloramphenicol in Enterobacteriaceae, isolated from patients with febrile illness, ranged from 31.0-94.2%.

Conclusion
The ndings con rm the ongoing elevated prevalence of multidrug-resistant bacteria in Bukavu, not withdrawing that the rates found can even be underestimated. In most cases worldwide, the risk factors for antimicrobial resistance are insu cient infection control in hospitals, inadequate public health systems for antimicrobial stewardship, inadequate knowledge of prescribers and users, advertising, and pharmaceutical companies' impact. The observed expansion requires that hospital therapeutic committees set up effective control and clinical management systems and de ne the right combinations of antibiotics. All methods were carried out in accordance with relevant guidelines and regulations. All experimental protocols were approved by the research committee of the Faculty of Medicine (UOB). As retrospective study, informed consent was waived by institutional ethics review board of the O cial University of Bukavu.

Consent for publication
Not applicable Availability of data and material The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Con icts of interest/Competing interests
The authors declare that they have no competing interests