Study design and setting
This cross-sectional study was designed and conducted explicitly to determine the bacteriological profile of aerobes isolated from SSIs. These isolates were identified, and antimicrobial susceptibility testing was performed, in order to analyze the prevalence of multidrug-resistant organisms (MDRO) and particular phenotypes: MRSA, ESBL, and carbapenemase producing organisms (CPO).
Patients who consented to participate in the study were included between January 2019 and January 2020. Patients who had had initial surgery in another hospital or ward (internal medicine or emergency), or had had prior to antibiotic treatment, and those who did not volunteer to participate were excluded from the study. We included the obstretrics (particularly ceasarean sections) and gastrointestinal wards at six public hospitals in Benin: Bethesda, Centre Hospitalier Universitaire de Zone de Suru Lere, Centre National Hospitalier Université Hubert Koutoukou Maga (CNHU-HKM), Centre Hospitalier Universitaire de la Mère et L’enfant (CHUMEL), Centre Hospitalier Universitaire Départemental Ouéme/Plateau (CHUDOP) and Centre Hospitalier Universitaire de Zone d’Abomey Calavi. Our choice of wards lay in the facts that caesarean sections are one of the most frequent surgical procedures worldwide [16], and that gastrointestinal section is available in these six public hospitals. All participating hospitals were located in the south of Benin, thereby allowing daily transport from each hospital to the CNHU-HKM laboratory, where all wound swabs collected were analyzed. All results were confirmed in the laboratory of the Cliniques Universitaires Saint-Luc-UCLouvain (Brussels, Belgium).
Sampling and collection method
The case definitions and clinical criteria of SSIs (superficial incisional SSI, deep incisional SSI, and organ/space SSI) were taken from the guidelines of the Centers for Disease Control and Prevention on the prevention of SSI [2]. A preliminary step in the project consisted of training nurses in the sampling technique. These nurses then collected all the samples per hospital. On a day when a clinical SSI was detected, one swab was collected aseptically from each infected patient using a sterile cotton swab. Each sample was labeled with the date of sample collection, the collection method, and the patient’s details. The swab was immediately dipped into a sterile tube with transport medium (Amies, Beckton Dickinson) and delivered to the bacteriology laboratory at CNHU-HKM. Coagulase-negative staphylococcus aureus were considered as pathogens only when isolated in two consecutive sampling swabs.
Socio-demographics and clinical data were obtained from the patients’files and by physical examination using a structured questionnaire. The following data were collected: ward to which admitted, age, gender, history of illness, and antibiotics used during and after surgery. Before the actual data was collected, a pretest of the data collection instrument was conducted to ensure the appropriateness of the questionnaire. If necessary, modifications were made. Data collection was supervised daily by the research team.
Processing of samples
Macroscopic and microscopic examination of samples
All the specimens were visually examined for consistency, color, turbidity, and the presence or absence of blood. Gram staining of each specimen was performed [17].
Culture of specimens and isolation of bacteria
Bacterial identification was performed according to the guidelines for microbiological methods of the European Committee of Antimicrobial Susceptibility Testing guidelines [18]. Cultures were incubated for a total of 48h (if there was no growth at 24h) at 37°C in aerobic atmosphere, and then examined for microbial growth. Identification of Gram-positive bacteria was done using Gram staining, and the catalase test, coagulase test, DNase test and Pastorex staphylococci plus test (Pastorex, staph plus Biorad). Gram-negative strains were identified using various biochemical tests: oxidase, and characteristics of the Analytical Profile Index (API 20E, Biomerieux, Lyon) such as the Voges Proskauer (VP) test, indole test and citrate utilization. All identifications were confirmed in Belgium by using Matrix Assisted Laser Desorption Ionization-Time of Flight (MALDI-TOF) mass spectrometry (Brucker Daltonics, Bremen, Germany). Due to budgetary constraints and a lack of laboratory facilities, we were unable to investigate anaerobic bacteria.
Antimicrobial Susceptibility Test (AST)
Antimicrobial susceptibility testing was performed for all isolates according to the modified Kirby Bauer disk diffusion technique as described in the European Committee on Antimicrobial Susceptibility Testing guidelines [18]. Antibiotics were purchased from BioRad (Marnes-la-Coquette, France) and included for S. aureus: ampicillin (10 , cefotaxime (30 g), cefoxitin (30 , gentamicin (10 , amikacin (30 ), ciprofloxacin (5 ), trimethoprim + sulfamethoxazole (25 g), tetracycline (30 g), and chloramphenicol (30 g). The lactose fermenters bacteria were tested for ampicillin (10 , piperacillin (100 ), cefotaxime (30 ), cefoxitin (30 ), ceftriaxone (30 g), gentamicin (10 ), tobramycin (10 ), amikacin (30 g), ciprofloxacin (5 , trimethoprim + sulfamethoxazole (25 g), imipenem (10 , and meropenem (10 . For lactose non-fermenters bacteria, the disks used included ceftriaxone (30 , ceftazidime (30 , gentamicin (10 ), ciprofloxacin and meropenem (10 g). After 24 h of incubation at 37°C, the inhibition zones were measured and the results were analyzed.
Phenotypic test for multidrug-resistant bacteria, extended-spectrum beta-lactamases, inducible clindamycin resistant (ICR) in S. aureus, and carbapenemases production
Multidrug resistance was defined as resistance to three or more antimicrobial classes [19]. The presence of ESBL was detected using the double-disk synergy test (DDST) between clavulanate and third-generation cephalosporins and/or aztreonam [20]. Methicillin-resistant S. aureus isolates were detected using the cefoxitin disk (30 g) method. The diameter of the zone of inhibition for cefoxitin was < 21mm [21]. Similarly, inducible macrolide-lincosamide streptogramin-B (iMLSB) resistance was detected in S. aureus with the D-test disk method using clindamycin (2ug) and erythromycin (15ug) on MHA plates. After overnight incubation, isolates with a flattened zone of inhibition adjacent to the erythromycin disk (referred to as a “D” zone) were considered to exhibit inducible clindamycin resistance [22]. The presence of resistance to at least one carbapenem was checked with the RESIST-3O.K.N.ICT (Coris Bioconcept, Gembloux,Belgium), which detects Oxacillinases (OXA 48), Klebsiella pneumoniae carbapenemases (KPC) and New-Delhi Metallo beta-lactamase (NDM). The final results of the ICT test were read when they became positive, at the latest after 15 min. All ESBL, MRSA and CPO strains were confirmed and characterized by whole-genome sequencing.
Quality control
Standard operating procedures (SOPs) were strictly followed during all bacteriological procedures, starting from sample collection, isolation, identification and antibiotic susceptibility testing. All culture media were prepared according to the manufacturers’ directions, and were checked for their sterility and performance. Two international control bacteria strains, E. coli ATCC 25922 and S. aureus ATCC 25923, were used as reference strains for quality control of the antimicrobial susceptibility and biochemical tests. The same strain of E.coli was considered as a negative control during the screening and phenotypic tests of ESBL-producing lactose fermenters bacteria. In Belgium, the same strain of E.coli was also considered as a control for mass spectrometry (MALDI-TOF). For transportation, we used swabs with transport medium (Amies, Beckton Dickinson).
Data analysis and statistical tests
Data was entered in Epi-data version 3.1, transferred to Statistical Package for Social Sciences (SPSS) software version 25, and Microsoft Excel software for analysis. Quantitative variables were expressed as median with interquartile range (IQR). A P-value less than 0.05 was considered statistically significant.