Clinical and microbiological profile of diabetic foot ulcers and predictors of infection and amputation: a retrospective cohort study from North Africa

¹Department of Internal Medicine, University Hospital of Ben Arous, Ben Arous, Faculty of Medicine of Tunis, Tunis El Manar University, Tunis, Tunisia, ²Centre de Traumatologie et des Grands Brûlés de Ben Arous, Ben Arous, Faculty of Medicine of Tunis, Tunis El Manar University, Tunis, Tunisia

^&Corresponding author
Hiba Jeribi, Department of Internal Medicine, University Hospital of Ben Arous, Ben Arous, Faculty of Medicine of Tunis, Tunis El Manar University, Tunis, Tunisia

Introduction    

Diabetic foot ulcer (DFU) represents a major chronic complication of diabetes, affecting approximately 20-25% of diabetic patients, according to the World Health Organization [1]. Diabetic peripheral neuropathy and peripheral artery disease are the main mechanisms involved in its pathophysiology, often exacerbated by intrinsic and extrinsic factors [1]. Its severity lies in the risk of infection, amputation, and mortality [2], highlighting the importance of appropriate multidisciplinary diagnostic and therapeutic management. The present study aimed to characterize the epidemiological, clinical, and therapeutic profile of patients with DFU and to identify factors independently associated with infection and amputation in a North African cohort.

Methods     

We conducted a single-center retrospective observational study, reported in accordance with the STROBE guidelines. The study was conducted in the Internal Medicine Department of a university hospital in Tunisia. Patients aged over 18 years, diagnosed with diabetes, and hospitalized for a DFU between November 2016 and March 2025 were included. Patients with skin inflammation from non-diabetic causes were excluded. Parameters studied included demographic characteristics, clinical and paraclinical data (both diabetes-related and foot-specific variables), as well as management and outcome data.

Data collection and variables: microangiopathy was defined as the presence of diabetic neuropathy, nephropathy, or retinopathy. Wound infection was classified according to the IWGDF/IDSA criteria (grade ≥2). Data were collected using a pre-established validated data collection form.

Statistical analysis: data entry and analysis were performed using SPSS version 26. Categorical variables were compared using the chi-square or Fisher's exact test. Multivariate logistic regression was used to identify independent predictors of infection and amputation, with results expressed as odds ratios (OR) with 95% confidence intervals (95% CI).

Ethical considerations: this retrospective study was conducted in accordance with the ethical principles of the Declaration of Helsinki. Ethical approval was obtained from the Ethics Committee of the Ben Arous Regional Hospital (approval number 15/25). Given the retrospective design, the requirement for individual informed consent to participate was waived by the same Ethics Committee, and patient data were anonymized prior to analysis.

Results     

The median age of patients was 56 years, with extremes ranging from 23 to 81 years. A clear male predominance was observed, with a sex ratio of 2.23. Ninety-three percent of patients had type 2 diabetes, with a mean duration of diabetes of 12.52 years. Poor glycemic control (mean glycated hemoglobin [HbA1c] 10.8%) was noted in 91% of patients. The main cardiovascular risk factors in this population were hypertension, present in 56% of patients, dyslipidemia in 50%, smoking in 38%, and obesity in 16%. Regarding diabetic complications, diabetic neuropathy was observed in 81% of patients, followed by nephropathy in 60% and retinopathy in 39%. Macrovascular complications included lower limb peripheral arterial disease (PAD) in 35% of cases, stroke in 10%, and coronary artery disease in 9%. A history of prior amputation was reported in more than one-third of the patients, representing 35% of the cohort. The most frequently observed clinical presentations were plantar neuropathic ulcer (32%), infected plantar neuropathic ulcer (24%), and dry gangrene (18%). Forefoot localization predominated, affecting 78% of cases. The majority of ulcers were deep, corresponding to 62% of patients, and extensive, with a surface area ≥1 cm² in 75% of cases. Radiographic evidence of osteomyelitis was observed in 26% of patients who underwent imaging. According to the IWGDF/IDSA classification, 71% of patients presented with an infected ulcer. The SINBAD score was high (3 to 6) in 63% of patients. Neuropathic etiology was predominant, accounting for 68% of ulcers, and the main precipitating factors identified were plantar hyperkeratosis, present in 68% of cases, and interdigital intertrigo, observed in 34% of patients. The most frequently isolated microorganisms were proteus mirabilis/vulgaris in 39% of cases, Pseudomonas aeruginosa in 25%, and Staphylococcus aureus/xylosus in 21%.

Empirical antibiotic therapy was initiated in 66% of patients. The most frequently used regimen was a combination of amoxicillin-clavulanic acid and ciprofloxacin, administered in 22% of cases, with a median duration of 16.5 days. The clinical course was marked by amputation in 23% of cases, with distal amputations predominating. Favorable evolution without amputation was observed in 49% of patients, with a median hospital stay of seven days and a mean of 11 days. Our study identified several factors significantly associated with DFU infection (Table 1), including prior antibiotic therapy (p<0.001), the presence of microangiopathy (p=0.009), neuropathic ulcer type (p<0.001), ulcer depth (p<0.001), and ulcer surface area (p<0.001). Factors associated with amputation (Table 2) included the presence of macroangiopathy (p=0.002), peripheral arterial disease (p=0.005), dry gangrene (p=0.01) or wet gangrene (p=0.05), bone infection (p=0.01), a high SINBAD score (p=0.003), as well as the extent and depth of lesions (p=0.002). On multivariate logistic regression, prior antibiotic therapy was an independent predictor of DFU infection (OR=13.67; 95% CI: 1.03-181.2; p=0.05), whereas dry gangrene was a significant independent predictor of amputation (OR=24.3; 95% CI: 1.39-424.6; p=0.03).

Discussion     

Principal findings: in this retrospective cohort of 68 patients hospitalized for a diabetic foot ulcer (DFU) in Tunisia, the median age was 56 years with a marked male predominance (sex ratio 2.23), and type 2 diabetes accounted for 93% of cases. Wound infection was present in 71% of patients, and 23% underwent amputation. Contrary to most Western series, proteus mirabilis/vulgaris (39%) and pseudomonas aeruginosa (25%) predominated over staphylococcus aureus (21%). In multivariate analysis, prior antibiotic therapy independently predicted infection, and dry gangrene independently predicted amputation.

Comparison with other studies: the mean age of patients in our study was 56.71 years. These results are comparable to those reported by Chaker et al. in Tunisia and Dia et al. in Senegal, who found mean ages of 59 years and 57 years, respectively [3,4]. This age group generally corresponds to a long duration of diabetes, often exceeding 10 years, during which chronic complications become more frequent. The male predominance observed in our cohort, with a sex ratio of 2.23, is consistent with most international literature [5,6]. This predominance may be explained by greater exposure of men to podiatric microtraumas and lower adherence to recommended self-care practices [6]. In our series, type 2 diabetes was largely predominant, accounting for 93% of cases, as reported in other series [3,4]. However, recent data indicate that the risk of diabetic foot ulcers in individuals with type 1 or type 2 diabetes depends mostly on age and disease duration [6]. The mean HbA1c was 10.8%, reflecting significant glycemic imbalance. Comparable results were observed in the study by Dia et al. which reported a mean HbA1c of 9.8% [4]. Chronic hyperglycemia is a major factor in DFU pathophysiology, leading to peripheral neuropathy, lower limb artery disease, and impaired wound healing mechanisms [5].

In our study, diabetic neuropathy was the most frequent microvascular complication, observed in 81% of patients. This result, comparable to the 77.4% reported by Djibril et al. [7], can be explained by the fact that peripheral neuropathy is a primary pathophysiological factor in DFU development due to the loss of protective sensation, which increases the risk of unnoticed trauma. Other common diabetic complications observed included nephropathy in 60% of patients and diabetic retinopathy in 39%. The latter may also be closely associated with DFUs, as visual impairment resulting from retinopathy promotes foot trauma and delays lesion detection, thereby contributing to the occurrence and progression of DFUs. Regarding macrovascular complications, PAD was present in 35% of patients, consistent with the literature, which estimates PAD prevalence between 20% and 50% in patients with diabetic foot ulcers worldwide [8]. The presence of cardiovascular complications, such as stroke in 10% and coronary artery disease in 9% of cases, can be explained by the correlation between DFU occurrence and accelerated progression of existing cardiovascular diseases [9]. Diabetic foot ulcers and cardiovascular diseases are both markers of diabetes severity and duration, and their pathophysiological synergy worsens the overall prognosis [6]. In our study, 51% of patients had a history of DFU other than the current lesion. The International Diabetes Federation (IDF) report noted recurrence prevalence in the Middle East and North Africa ranging from 31% in Egypt to 43% in Turkey [8]. According to the French National Authority for Health (HAS), a prior DFU is considered a grade 3 risk factor for the occurrence of a new lesion [10]. The higher proportion observed in our population may suggest, on one hand, insufficient initial therapeutic management of the first ulcer and/or a low level of patient therapeutic education.

The main lesion type identified in our series was plantar neuropathic ulcer, found in 56% of patients. This predominance has also been reported in the series by Adel et al. (40%) [11]. Conversely, other studies conducted in Tunisia and Senegal report a predominance of gangrene [3,4], which represented 34% of cases in our series, making it the second most frequent lesion type. In the literature, plantar neuropathic ulcer remains the typical DFU and is the primary complication of diabetic neuropathy. In our series, 26% of patients who underwent standard foot radiography presented radiological signs suggestive of osteomyelitis. This result is comparable to those reported by Adel et al. and Djibril et al. [7,11]. The prevalence of osteomyelitis in diabetic patients with DFU varies significantly, ranging from 30% to 80% [12]. Diabetic foot ulcers infection was identified in 71% of patients. Our results are slightly higher than those reported in the literature, which generally places the infection rate around 60% [6]. This difference may be explained by variable infection definitions across studies, as well as differences in lesion stage at the time of consultation. Among the main intrinsic triggering factors observed during foot examination, plantar hyperkeratosis was noted in 65% of patients. These results are comparable to those of Adel et al. (71%) [11]. Indeed, plantar hyperkeratosis is one of the lesions most frequently associated with the development of diabetic foot ulcers [10]. Interdigital intertrigo was the principal extrinsic triggering factor, observed in 34% of patients. These results are higher than those reported by Djibril et al. in Togo (12.9%) [7]. Such differences can be explained by climatic and environmental variations between regions, as a warmer and more humid climate favors the occurrence of interdigital intertrigo. Our results may also be influenced by Muslim ablution practices; when feet are not thoroughly dried after washing, moisture promotes the development of interdigital intertrigo.

The literature indicates that Staphylococcus aureus is the most frequently isolated pathogen in DFUs [1,13]. However, in our series, Proteus spp. predominated (39%), followed by Pseudomonas aeruginosa (25%) and Staphylococcus aureus (21%). This difference may be explained by prior antibiotic therapy received by more than one-third of our patients before admission, favoring the selection of resistant or less frequent pathogens, such as Gram-negative bacilli. It is important to note that, unlike Western countries, developing countries with warm climates, particularly in North Africa and Asia, show a lower prevalence of Staphylococcus aureus and more frequent isolation of Pseudomonas aeruginosa [14]. In our study, 66% of patients received antibiotic therapy. The most frequently used regimen was amoxicillin-clavulanic acid combined with ciprofloxacin, prescribed in 22% of cases. Predominant use of amoxicillin-clavulanic acid, either alone or in combination with another antibiotic, has also been reported in several studies, including those by Chaker et al.and Dia et al. [3,4]. According to the 2023 IWGDF recommendations, antibiotic therapy covering both Gram-positive and Gram-negative bacteria, particularly Pseudomonas aeruginosa, is recommended in tropical or subtropical regions, such as North Africa, or in patients who have recently received antibiotic therapy [14]. Initial monotherapy with amoxicillin-clavulanic acid, a fluoroquinolone, or a cephalosporin is generally recommended [14]. In our cohort, empirical antibiotic therapy was highly variable, with 16 different regimens, highlighting the need to adapt empirical protocols to local microbiological data. Favorable evolution of DFUs without amputation was observed in 49% of cases, an intermediate rate compared to published series [4,11]. The amputation rate was 23%. According to the IDF, the prevalence of lower limb amputations in the Middle East and North Africa is generally below 30% [8]. Higher percentages are often observed in studies with small patient numbers compared to larger samples in the same regions [8]. The mean length of hospital stay was 11 days, shorter than in some African studies, where it may exceed two weeks [4,7]. Diabetic foot ulcers have become one of the main causes of prolonged hospitalization among diabetic patients due to the complexity of management, particularly in cases of infection, associated PAD, or complications requiring surgical intervention [15].

Risk factors for diabetic foot ulcers infection: in our series, prior antibiotic therapy was identified as a risk factor for DFU infection. Although no similar studies were found in the literature, our results can be interpreted according to two hypotheses. The first is bacterial selection and microbiota disruption: non-targeted or inappropriate use of antibiotics without precise microbiological diagnosis could disturb the wound microbiome, promoting the emergence of more virulent or resistant pathogens. The second hypothesis is confounding bias: it is possible that prior antibiotic therapy is not an independent risk factor on its own, but rather a marker of the initial severity of the ulcer. Multivariate analysis in our study demonstrated a significant association between the presence of microangiopathy and the risk of DFU infection (p=0.009). Similar associations have been observed in studies conducted in Australia [9]. This association can be explained by impaired microcirculation leading to tissue hypoxia and a reduced local immune response [16]. Patients with deep ulcers had a 1.2-fold higher risk of infection. These results are comparable to those of Jia et al. who reported a 2.2-fold increased risk of infection in deep wounds [17]. Bone involvement represents an additional risk factor for DFU infection, as deep lesions affecting underlying tissues such as muscles, tendons, or bones are less accessible and more difficult to clean. Neuropathic ulcer origin was also significantly associated with infection in our series, consistent with findings reported by Jia et al. [17]. The loss of protective sensation due to diabetic neuropathy is a major risk factor for the development of foot infections [13].

Risk factors for amputation: univariate analysis in our series showed that the presence of macroangiopathy increased the risk of lower limb amputation by 1.6 times. Moreover, meta-analyses by Sen et al. as well as studies by Chaker et al. and Gong et al. have highlighted a statistically significant association between PAD and the risk of amputation [3,18,19]. Our study supports these findings, confirming that PAD is frequently recognized as a major risk factor for amputation in patients with DFU. A relationship between ulcer surface area and amputation risk was identified (p=0.002). This is consistent with the literature, which frequently reports that larger ulcer surface area constitutes an independent risk factor for amputation [1,2,19]. Likewise, ulcer depth is a major risk factor for amputation [2]. Our results demonstrate a statistically significant association between these two variables (p=0.002), likely because deeper tissue involvement complicates therapeutic management, increases infection risk, and impairs healing, thereby worsening ulcer prognosis. In our series, bone infection was also significantly associated with amputation, with a sevenfold increased risk in affected patients. This strong association is consistent with multiple studies [1,2,18,19] and can be explained by the relative antibiotic resistance of bone infections, which often necessitate surgical removal of necrotic tissue and bacterial biofilms to reduce bacterial inoculum [1]. Several studies worldwide agree that gangrene constitutes a major risk factor for amputation in patients with DFU [2,3,19]. Our results align with this evidence, showing a significant association between both wet gangrene (p=0.05) and dry gangrene (p=0.03) with amputation. Patients with dry gangrene had a 24-fold higher risk of undergoing lower limb amputation. Infection is also among the most frequently cited risk factors for diabetic foot amputation [1,2,10,18]. It is particularly dangerous in diabetic patients due to the increased risk of multi-resistant bacterial infections [20] and frequent association with other complications such as diabetic neuropathy and PAD [16]. Our findings confirm the close association between infection and amputation risk in the context of DFU.

Strengths and weaknesses of this study: the main strengths of this study are its detailed clinical and microbiological characterization of DFU in a North African population - a setting for which such data remain scarce - and its use of multivariate analysis to identify independent predictors of infection and amputation, in accordance with STROBE reporting guidelines. Its main limitations relate to its retrospective, single-center design and its limited sample size, which are detailed in the Limitations section below.

Meaning of the study: taken together, these findings suggest that empirical antibiotic protocols for DFU in North Africa should be adapted to local microbiological patterns, favoring broader gram-negative coverage rather than the Staphylococcus-focused regimens generally recommended in Western guidelines. They also underline the importance of avoiding unnecessary or non-targeted antibiotic exposure prior to hospitalization, and of aggressive, multidisciplinary management of dry gangrene to reduce the risk of amputation.

Unanswered questions and future research: whether the association observed between prior antibiotic therapy and infection reflects a true causal mechanism, such as bacterial selection and microbiota disruption, or confounding by the initial severity of the ulcer, remains to be clarified. Larger, prospective, multicenter studies across North Africa are needed to confirm these microbiological patterns and predictors, and to guide the development of regionally adapted empirical antibiotic protocols for the management of diabetic foot ulcers.

Limitations: this study has several limitations. It is retrospective, with data collected from medical records that were sometimes incomplete. It is also single-center, with a small sample size of 68 patients, which limits the generalizability of the results. Consequently, the wide 95% confidence intervals observed in our analyses reflect this limited sample size and the associated imprecision. Larger multicenter studies are therefore necessary to confirm our findings and provide more precise estimates.

Conclusion     

This study highlights the multifactorial complexity of diabetic foot ulcers, influenced by clinical, vascular, and infectious factors, as well as individual patient history. In long-standing diabetic patients with poor glycemic control, systematic screening for micro- and macroangiopathic complications and rigorous assessment of at-risk feet are essential. Prevention, early detection, patient education, and multidisciplinary management, including antibiotic protocols adapted to local microbiological data, are critical to prevent progression to severe forms or those requiring amputation. Large-scale, multicenter studies are needed to consolidate these findings and better define the risk factors associated with diabetic foot ulcers.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

Hiba Jeribi and Donia Chebbi: conceptualization, data collection, writing - original draft. Hiba Jeribi and Mohamed Othmen Chouchene: formal statistical analysis, writing - review and editing. Meya Abdallah, Fatma Ben Dahmen, Mohamed Abdelkefi and Donia Chebbi: supervision, critical revision for important intellectual content. All the authors have read and agreed to the final manuscript.

Acknowledgments     

The authors acknowledge the patients who participated in this study and the staff of the Internal Medicine Department of the Ben Arous University Hospital. No writing or editorial assistance from a third party was received in the preparation of this manuscript, and no source of funding was received for the study, authorship, or preparation of the manuscript, for any author.

Tables     

Table 1: univariate analysis of clinical and paraclinical variables associated with diabetic foot ulcer infection, aimed at identifying risk factors for infection, among patients hospitalized in the Internal Medicine Department of the University Hospital of Ben Arous (Tunisia) between November 2016 and March 2025 (N=68; no infection, n=20; infection, n=48)

Table 2: univariate analysis of clinical, vascular, and infectious variables associated with lower-limb amputation, aimed at identifying risk factors for amputation, among patients hospitalized for a diabetic foot ulcer in the Internal Medicine Department of the University Hospital of Ben Arous (Tunisia) between November 2016 and March 2025 (N=68; no amputation, n=53; amputation, n=15)

References