Chronic diabetic foot ulcers are one of the most common complications among people with diabetes. Nearly a third of people with diabetes suffer from foot ulcers at the course of their lifespan, and over 50% of these ulcers are infected [1]. Diabetic foot infections are associated with high morbidity and mortality and can lead to lower limb amputation if not treated effectively [2]. These lesions are often polymicrobial, with Gram-positive or Gram-negative aerobic bacteria [3][4] as well as anaerobes involved. To prevent antibiotic resistance, reduce costs and minimise adverse effects associated with antibiotic use, the recommendations of the Infectious Diseases Society of America (IDSA) for the diagnosis and treatment of diabetic foot infections advocate the prescription of antibiotics that have proven effective in treating diabetic foot infections, with coverage of Gram-positive cocci and limited coverage of Gram-negative bacteria. Generally, the empirical antibiotics used orally are clindamycin, cotrimoxazole, levofloxacin and amoxicillin + clavulanic acid [5]. However, the rapid emergence of antibiotic resistance to these drug families among the most common bacteria represents the greatest challenge in the management of these infections, especially carbapenem resistance and, in some cases, multidrug resistance [6].

The aim of this study is therefore to determine the profile of the identified bacteria and their sensitivity to the antibiotics used in the treatment of infected diabetic feet.

The study was conducted in the Endocrinology and Metabolism Department of the Hubert Koutoukou Maga National University Hospital Hubert Koutoukou Maga (CNHU HKM) in Cotonou. This was a descriptive cross-sectional study based on retrospective data involving diabetic patients admitted to the diabetic foot infection unit between 1 January 2015 and 31 December 2019. All patients admitted during the study period who had bacteriological testing of pus with an antibiogram were included. Patients admitted for a diabetic foot infection with incomplete medical records were excluded from the study.

The socio-demographic, clinical, laboratory and therapeutic data were collected using a dedicated data collection form. Diabetic foot was defined as an infection, ulceration or destruction of deep foot tissues associated with neuropathy and/or peripheral arterial disease of the lower limbs in a diabetic patient (International Consensus on the Diabetic Foot). Infection was diagnosed on the basis of the presence of fever, suppuration of the foot, leukocytosis on complete blood count, elevated C-reactive protein (CRP), with or without the isolation and identification of bacteria in bacterial culture of the pus. Bacteriological samples have been done by swab. Existing dressings were removed using aseptic precautions. The wounds surface was irrigated with sterile normal saline. Necrotic tissue, slough and superficial debris were removed by sharp debridement. And after cleansing and debridement, a sterile swab was applied using Levine technique. When feasible, deep tissue specimens were obtained using sterile curettes. Specimen were immediately placed into transport media and delivered promptly to the microbiology laboratory. Aerobic and anaerobic cultures were performed according to local laboratory procedures. The results of the susceptibility testing were interpreted in accordance with the recommendations of the French Society of Microbiology’s Susceptibility Testing Committee and the European Committee on Antimicrobial Susceptibility Testing (CASFM-EUCAST) of February 2018 [7].

The different lesions were graded using the Wagner classification [8]. Extended-spectrum beta-lactamase (ESBL)-producing Enterobacteriaceae and methicillin-resistant Staphylococcus aureus (MRSA) were classified as multidrug-resistant bacteria (MDRB). In terms of treatment, an amputation was classified as minor if it was partial and major if it was total (leg or thigh).

Statistical analyses: the data collected were analysed using Epi Info software version 5.4. The results were presented as proportions or means with their standard deviation. The frequency of resistance was calculated by dividing the number of bacteria resistant to the tested antibiotic by the total number of that type of bacteria isolated.

This study identified the main bacteria involved in infected feet, as well as their behaviour in relation to the antibiotics used. It emerged that the most common bacteria were Pseudomonas aeruginosa, Escherichia coli, Klebsiella pneumoniae and Staphylococcus aureus. All these bacteria showed a high level of resistance to the antibiotics tested.

In this study, as in series on diabetic foot, some patients hospitalised for diabetic foot had already experienced a foot ulcer and an amputation in the context of diabetic foot [9][10]. Patients had self-medicated with antibiotics in 47.1% of cases, and amoxicillin + clavulanic acid was the most commonly used molecule, followed by metronidazole and cloxacillin. This almost systematic use may be due to the relative ease of access to these drugs, although this has since been regulated and strictly controlled by the country’s health authorities.

Table 3. Antibiotic resistance profile of bacteria tested in 102 patients hospitalised for diabetic foot, N = 102.

Table 4. Multidrug-resistant bacteria (MDRB) in 102 patients hospitalised for diabetic foot, N = 102.

We found that 28.4% of samples contained two organisms and 5.9% contained three organisms. The polymicrobial nature observed in diabetic foot infections is well-known and described in the published literature [11]-[14] with varying frequencies.

Regarding the bacterial ecology identified in this study, a study conducted by Alassani et al. in 2014 [15] in the same department found a similar bacterial ecology among 26 identified a similar bacterial flora, dominated by Gram-negative bacteria, among the 26 identified pathogens. Other authors have confirmed this bacteriological profile dominated by Gram-negative bacteria in infected diabetic feet [6][10][16][17]. In contrary to the findings of our study and those of Asian and Turkish studies [14][18], European and American studies have found a bacteriological profile dominated by Gram-positive bacteria [9][19][20]. The presence of Pseudomonas among the common bacteria, as seen in our study, is attributed to hospital flora resulting from prolonged hospital stays and the chronic nature of the lesions [17][21].

The antibiotic resistance profile of the most commonly isolated bacteria revealed a high frequency of ESBL-producing Enterobacteriaceae (E. coli 38.5%, Klebsiella pneumoniae 42.3%) and MRSA (68.2%). Multiple studies have reported a high frequency of antibiotic resistance in bacteria from infected diabetic feet [2][17][22]. More specifically, higher resistance to beta-lactam antibiotics is observed, with the exception of imipenem, to which the bacteria remain highly susceptible. This might be partly explained by the high prevalence of ESBL-producing bacteria found: 38.5% of Escherichia coli strains and 42.3% of Klebsiella pneumoniae strains. ESBL-producing strains of Klebsiella pneumoniae and Escherichia coli, as well as multidrug-resistant Pseudomonas, have also been reported by other studies [10][23] in infected diabetic feet. Resistance to the second-generation fluoroquinolones tested remains high for the three Gram-negative bacteria, with the highest frequency observed in Escherichia coli. Mahamane Sani et al. [16] reported 52.7% resistance among Enterobacteriaceae to first- and second-generation fluoroquinolones in their 2017 study conducted in Ivory Coast. This high frequency of bacterial resistance to antibiotics might make it more difficult to choose the initial empirical antibiotic therapy, which is based primarily on beta-lactams ± beta-lactamase inhibitors [5], as resistance is higher and there few matches with antibiotic susceptibility testing results, which accounted for 3.9% in our study. In African studies, the most commonly used combinations are penicillin-imidazoles and fluoroquinolones-imidazoles [16][24][25]. The combination of amoxicillin + clavulanic acid and metronidazole was the most commonly used in our study (32.4%). Moreover, the prolonged length of stay observed in this study could be explained by the high frequencies of deeper lesions (Wagner grades III, IV, V) and osteitis (70.6%). However, this finding remains consistent with those reported by Sidibé et al. in 2006 [24] and Mamane Sani et al. in 2017 [16], which were 54 ± 29 and 59.4 ± 31.6 days respectively. The amputation rate (41.2%) remains similar to that of 41.38% reported by Sidibé et al. [24] but higher than the 22% reported by Nese Saltoglua et al. in 2018 [10].

Our study, however, has certain limitations that should be noted. Its single-center design may not accurately represent the country’s ecology. Furthermore, the study location (the country’s first reference center) and the sometimes lengthy patient journey involving the use of inappropriate and often incompletely documented empirical antibiotics could also contribute to an increased incidence of bacterial resistance.

The bacteriological profile in this study is dominated by Gram-negative bacteria. The bacteria were multi-drug resistant, with a significant frequency of ESBL-producing strains. There is a low degree of consistency between empirical antibiotic therapy and the results of the antibiotic sensitivity test. It may therefore be necessary to review practices regarding the choice of antibiotics in empirical therapy in order to improve patient outcomes.