Changing Bacterial Etiology and Antimicrobial Resistance Profiles as Prognostic Determinants of Diabetic Foot Infections: A Ten-Year Retrospective Cohort Study

Abstract

Background:

In this single-center study, we analyzed a retrospective cohort of patients with diabetic foot infections (DFIs) between 2011 and 2020.

Patients and Methods:

The first and second five-year periods were compared. A poor prognosis was defined as a primary composite end point including re-infection, major amputation, or mortality at six months.

Results:

A total of 484 patients were enrolled. Overall, 269 patients had the primary composite end point. A substantial decrease was detected in the second five-year period in terms of re-infection (n = 132, 66.0% vs. n = 68, 23.9%; p < 0.001) and mortality (n = 22, 11.0% vs. n = 7, 2.5%; p < 0.001). A total of 798 micro-organisms were isolated from 484 patients. A substantial increase was detected in polymicrobial infections (48.5% vs. 65.1%; p = 0.001) as well as Streptococcus spp. (2.5% vs. 9.2%; p = 0.003), Corynebacterium spp. (9.5% vs. 22.9%; p < 0.001), and extended-spectrum β-lactamase (ESBL) producing Escherichia coli (3.0% vs. 12.7%; p < 0.001) in the second five-year period, whereas the prevalence of multi-drug–resistanct (MDR) Pseudomonas aeruginosa (17.0% vs. 10.2%; p = 0.029) and carbapenem-resistant Acinetobacter baumannii (7.5% vs. 2.8%; p = 0.017) decreased. Multivariable regression analysis revealed that MDR Pseudomonas aeruginosa (odds ratio [OR], 1.917; 95% confidence interval [CI], 1.074–3.420; p = 0.028) and carbapenem-resistant Acinetobacter baumannii (OR, 3.069; 95% CI, 1.114–8.453; p = 0.030) were independent predictors for poor prognosis.

Conclusions:

This 10-year cohort study provides reassuring information about the changing epidemiology of DFIs and the prognostic determinants in patients with DFIs.

Diabetes mellitus and related-complications especially foot infections are one of the most significant healthcare burdens worldwide. Diabetic foot infections (DFIs) are associated with prolonged hospitalization and a life time risk of lower extremity limb loss, and substantially contribute to mortality, morbidity, and high financial costs.^1,2

Micro-organisms most frequently isolated from DFIs are gram-positive cocci and gram-negative bacillus. Diabetic foot infections are often polymicrobial and recognizing the complexity of microbiologic interaction will guide physicians to establish best clinical practice.^3,4 Because antimicrobial resistance has increased as a result of broad-spectrum unnecessary antibiotic agents, appropriate narrow-spectrum initiation may be considered as a key element for the management of DFIs.⁵

Guidelines have emphasized the importance of microbiology of DFIs that is vital for monitoring the local antimicrobial resistance.^3,4,6 Surveillance studies not only inform clinicans on the current epidemiologic data but also provide insight into microbial dynamics and local microbiologic trends. In this study, we therefore aimed to explore pathogen micro-organisms, their antimicrobial resistance profiles, and prognostic indicators of poor clinical outcomes in patients with DFIs. In addition, we investigated potential relations between causative agents and demographic characteristics of the patients including age and gender, study period (first and second five-year periods), wound location, laboratory parameters, and history of previous DFIs.

Patients and Methods

In this single-center study, we analyzed a retrospective cohort of patients with DFI between 2011 and 2020 from one interdisciplinary Cerrahpasa DFIs Committe at the Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty. Patients with culture-positive results who were followed up for at least six months were included in the study. Hospitalized patients who were already on oral or intravenous antibiotic agents were not included in the study. However, outpatients who were already on oral antibiotic agents were not excluded from the study.

Demographic characteristics, clinical features, microbiological profiles, laboratory parameters, and outcomes were retrieved from the patients' medical charts. All patients with severe infection and patients with a moderate infection with complicating features were hospitalized. A proper wound sampling was performed based on the national and international guidelines.^2-4,6

Microbiologic examinations were performed before the first administration of antimicrobial agents. Specimens were obtained using wound curettage after debridement, needle aspiration of purulent material, or tissue biopsy and bone biopsy. Deep tissue samples were obtained from the debrided ulcer base. Bone biopsy samples were obtained from suspected osteomyelitis.² Swab samples was not obtained to avoid contamination. For microbiologic identification, standard conventional clinical laboratory methods including staining, culturing, and simple biochemical tests were used for the identification of the micro-organisms. The preference of antibiotic discs and susceptibility testing were determined according to the European Committee on Antimicrobial Susceptibility Testing guidelines. Antibiotic susceptibility testing was performed on Mueller-Hinton agar by the Kirby Bauer disc diffusion method and additionally by the e-test method if required. Multi-drug–resistance (MDR) was defined as acquired nonsusceptibility to three or more different antimicrobial classes.

Mortality was defined as the six-month mortality. Major amputation was defined as any lower extremity resection proximal to the ankle at six months. Minor amputation was defined as any resection at the level of the ankle, foot, or toe at six months.⁷ Re-infection was defined as a recurrent infection on the same foot during the six-month follow-up period after the first month of the DFI treatment. A poor prognosis was defined as primary composite end point including re-infection, major amputation, or mortality. Periods of infection were equally divided into two groups as the first and second years. The International Working Group on the Diabetic Foot (IWGDF) classification was used to evaluate the severity of disease.² In this study, the association between causative micro-organisms and periods of infection, poor prognostic outcomes, age, wound location, and laboratory parameters were demonstrated.

Frequencies (n) and percentages (%) were used to present the descriptive characteristics of the data, wheras numerical variables were represented through mean ± standard deviation (SD). Categorical data were compared with the χ² or Fisher exact tests. The Kolmogorov-Smirnov test was performed to determine data normality. Although the independent sample t-test was used for normally distributed data, the Mann-Whitney U test was used for non-normally distributed data. Multivariable regression analysis was performed for identifying predictors for poor prognosis. Odds ratios (OR) with a 95% confidence interval (CI) were calculated. A p value <0.05 was considered as statistically significant. The analyses were performed using SPSS Statistics, version 22 (IBM Corp, Armonk, NY).

This study was conducted in accordance with the ethical standards of the Declaration of Helsinki. This study was approved by the Ethics Committee of Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty (approval number: E-83045809-604.01.02-2601, date: July 1, 2021). Written informed consent was waived, given the retrospective nature of this study.

Results

A total of 484 patients were enrolled, 267 of whom (55.2%) were male. The mean age was 60.3 (±9.7) years. No difference was observed between the first and second five-year period in terms of age (p = 0.960) and gender (p = 0.478). Two-hundred and fifty-eight (53.3%) patients had amputations, as the minor (n = 166; 34.3%) or major (n = 92; 19.9%). The mortality rate was 6.0% with 29 deaths. Re-infection occurred in 200 (41.7%) patients during the six-month follow-up period. Overall, 269 patients (55.6%) had the primary composite end point. No difference was found between the first and second five-year period in terms of major amputation (n = 35, 17.5% vs. n = 57, 20.1%; p = 0.478). However, a decrease was detected in the second five-year period in terms of re-infection (n = 132, 66.0% vs. n = 68, 23.9%; p < 0.001), mortality (n = 22, 11.0% vs. n = 7, 2.5%; p < 0.001), and the primary composite end point (n = 148, 74.0% vs. n = 121, 42.6%; p < 0.001).

A total of 798 micro-organisms were isolated from 484 patients. Two-hundred eighty-two (58.3%) patients had polymicrobial infection, whereas 202 patients (41.7%) had monomicrobial infection. The most common micro-organisms were Pseudomonas aeruginosa (n = 137; 28.3%) and Escherichia coli (n = 70; 14.5%) among gram-negatives. The most common micro-organisms among gram-positives were Enterococcus spp. (n = 93; 19.2%), coagulase-negative Staphylococcus spp. (CoNS; n = 90; 18.6%) and Staphylococcus aureus (n = 87; 18.0%). Methicillin resistance rates were 27.6% (n = 24/87) and 80.0% (n = 72/90) in Staphylococcus aureus and CoNS respectively. Multi-drug–resistant Pseudomonas aeruginosa was detected in 63 of 137 (46.0%) isolates. Carbapenem-resistant Acinetobacter baumannii was detected in 23 of 28 (82.1%) isolates. Extended-spectrum β-lactamase (ESBL)-producing gram-negative bacteria were detected in 59 isolates of Escherichia coli (n = 42/70; 60.0%) and Klebsiella spp. (n = 17/38; 44.7%).

An increase was detected in polymicrobial infections (n = 97, 48.5% vs. n = 185, 65.1%; p = 0.001) as well as Streptococcus spp. (n = 5, 2.5%, vs. n = 26, 9.2%; p = 0.003), Corynebacterium spp. (n = 19, 9.5% vs. n = 65, 22.9%; p < 0.001), and ESBL-producing Escherichia coli (n = 6, 3.0% vs. n = 36, 12.7%; p < 0.001) in the second five-year period, whereas the prevalence of MDR Pseudomonas aeruginosa (n = 34, 17.0% vs. n = 29, 10.2%; p = 0.029) and carbapenem-resistant Acinetobacter baumannii (n = 15, 7.5% vs. n = 8, 2.8%; p = 0.011) decreased (Table 1). Regarding gender and causative agents, polymicrobial infections were more frequent among male than female patients with DFIs (64.0% vs. 51.2%; OR, 1.701; 95% CI, 1.181–2.451; p = 0.004). In addition, Corynebacterium spp. (21.0% vs. 12.9%; OR, 1.791; 95% CI, 1.093–2.937; p = 0.020) and Citrobacter spp. (4.1% vs. 0.5%; OR, 1.792; 95% CI, 1.093–2.937; p = 0.015) were more frequent among male patients.

Table 1.

Isolated Micro-Organisms from Patients with Diabetic Foot Infections between 2011 and 2020

	2011–2015		2016–2020		In total
	n	%	n	%	n	%	p
Gram-negative	63	31.5	56	19.7	119	24.6	0.001
Gram-positive	40	20.0	43	15.1	83	17.1
Polymicrobial	97	48.5	185	65.1	282	58.3
Staphylococcus aureus	35	17.5	52	18.3	87	18.0	0.819
MRSA	11	5.5	13	4.6	24	5.0	0.645
Coagulase-negative Staphylococcus spp.	38	19.0	52	18.3	90	18.6	0.848
MRCoNS	30	15.0	42	14.8	72	14.9	0.949
Streptococcus spp.	5	2.5	26	9.2	31	6.4	0.003
Enterococcus spp.	33	16.5	60	21.1	93	19.2	0.203
Corynebacterium spp.	19	9.5	65	22.9	84	17.4	< 0.001
Escherichia coli	23	11.5	47	16.5	70	14.5	0.120
ESBL (+) Escherichia coli	6	3.0	36	12.7	42	8.7	< 0.001
Klebsiella spp.	13	6.5	25	8.8	38	7.9	0.354
ESBL (+) Klebsiella spp.	5	2.5	12	4.2	17	3.5	0.310
Proteus spp.	19	9.5	30	10.6	49	10.1	0.703
Pseudomonas aeruginosa	56	28.0	81	28.5	137	28.3	0.900
MDR Pseudomonas aeruginosa	34	17.0	29	10.2	63	13.0	0.029
Acinetobacter baumannii	18	9.0	10	3.5	28	5.8	0.011
CR Acinetobacter baumannii	15	7.5	8	2.8	23	4.8	0.017
Morganella morganii	13	6.5	17	6.0	30	6.2	0.817
Citrobacter spp.	3	1.5	9	3.2	12	2.5	0.375
ESBL (+) gram-negatives	10	5.0	47	16.5	57	11.8	< 0.001
Others	24	12.0	25	8.8	49	10.1	0.251

MRSA = methicillin-resistant Staphylococcus aureus; MRCoNS = methicilline resistant coagulase-negative Staphylococcus spp.; ESBL = extended-spectrum β-lactamase; MDR = multi-drug–resistant.

Staphylococcus aureus (p = 0.026), MDR Pseudomonas aeruginosa (p = 0.015), Acinetobacter baumannii (p = 0.033) and carbapenem-resistant Acinetobacter baumannii (p = 0.025) as well as IWGDF infection grade 4 (p < 0.001) were associated with prognosis (Table 2). Multivariable regression analysis revealed that MDR Pseudomonas aeruginosa (OR, 1.917; 95% CI, 1.074–3.420; p = 0.028) and carbapenem-resistant Acinetobacter baumannii (OR, 3.069; 95% CI, 1.114–8.453; p = 0.030) were the only independent predictors for poor prognosis.

Table 2.

Univariable Associations between Isolated Micro-Organisms and Poor Prognosis during the Six-Month Follow-Up

		Major amputation/re-infection/mortality at 6-mo
		No		Yes
		n	%	n	%	OR (CI)	p
Staphylococcus aureus	Absence	167	77.7	230	85.5
	Presence	48	22.3	39	14.5	0.59 (0.37–0.94)	0.027
MDR Pseudomonas aeruginosa	Absence	196	91.2	225	83.6
	Presence	19	8.8	44	16.4	2.02 (1.14–3.57)	0.015
Acinetobacter baumannii	Absence	208	96.7	248	92.2
	Presence	7	3.3	21	7.8	2.52 (1.05–6.04)	0.033
Carbapenem-resistant Acinetobacter baumannii	Absence	210	97.7	251	93.3
	Presence	5	2.3	18	6.7	3.01 (1.10–8.25)	0.025
IWGDF classification
	Grade-2 (ref.)	92	42.8	84	31.2	1.00
	Grade-3	111	51.6	110	40.9	1.08 (0.73–1.61)	0.685
	Grade-4	12	5.6	75	27.9	6.84 (3.48–13.48)	<0.001

OR = odds ratio; CI = confidence interval; MDR = multi-drug–resistant; IWGDF = International Working Group on the Diabetic Foot; CR = carbapenem-resistant; ESBL = extended-spectrum β-lactamase

Regarding age and causative agents, CoNS (16.8% vs. 31.1%; OR, 0.45; 95% CI, 0.24–0.81; p = 0.008) and methicillin-resistant CoNS (13.2% vs. 26.2%; OR, 0.43; 95% CI, 0.23–0.81; p = 0.008) were less common in patients aged 50 years and older, whereas MDR Pseudomonas aeruginosa (14.7% vs. 1.6%; OR, 10.30; 95% CI, 1.40–75.72; p = 0.005) was more common in patients aged 50 years and older (Table 3).

Table 3.

Univariable Associations between Age and Causative Micro-Organisms

		Age
		<50 y		≥50 y
		n	%	n	%	OR (CI)	p
CoNS	Absence	42	68.9	352	83.2
	Presence	19	31.1	71	16.8	0.45 (0.24–0.81)	0.008
Methicillin-resistant CoNS	Absence	45	73.8	367	86.8
	Presence	16	26.2	56	13.2	0.43 (0.23–0.81)	0.008
MDR Pseudomonas aeruginosa	Absence	60	98.4	361	85.3
	Presence	1	1.6	62	14.7	10.30 (1.40–75.72)	0.005

CoNS = coagulase-negative staphylococci; MDR = multi-drug–resistant.

Staphylococcus aureus was more frequent in the metatarsal region of the forefoot (31.8% vs. 15.4%; OR, 2.56; 95% CI, 1.38–4.76; p = 0.002), whereas it was less frequent in toes except the first toe (9.4% vs. 23.5%; OR, 2.95; 95% CI, 1.43–6.08; p = 0.003). Polymicrobial infection was more frequent both metatarsal region (72.7% vs. 63.6%; OR, 5.46; 95% CI, 1.64–18.26; p = 0.004) and midfoot (79.2% vs. 62.8%; OR, 2.26; 95% CI, 1.11–4.60; p = 0.021). Acinetobacter baumannii (13.2% vs. 1.5%; OR, 9.97; 95% CI, 2.81–35.42; p < 0.001) and carbapenem-resistant Acinetobacter baumannii (9.4% vs. 1.1%; OR, 9.13; 95% CI, 2.11–39.48; p < 0.001) were more frequent in midfoot (Table 4).

Table 4.

Univariable Associations between Site of the Obtained Clinical Samples and Causative Micro-Organisms

Toes except the first	Absence		Presence
	n	%	n	%	OR (CI)	p
Staphylococcus aureus	50	23.5	10	9.4	2.95 (1.43–6.08)	0.003

Metatarsal region	Absence		Presence
	n	%	n	%	OR (CI)	p
Gram-negatives	55	21.7	3	4.5	ref. 1.00
Gram-positives	37	14.6	15	22.7	7.43 (2.01–27.49)	0.003
Polymicrobial	161	63.6	48	72.7	5.46 (1.64–18.26)	0.006
Staphylococcus aureus	39	15.4	21	31.8	2.56 (1.38–4.76)	0.002

Midfoot	Absence		Presence
	n	%	n	%	OR (CI)	p
Monomicrobial	99	37.2	11	20.8	ref. 1.00
Polymicrobial	167	62.8	42	79.2	2.26 (1.11–4.60)	0.021
Acinetobacter baumannii	4	1.5	7	13.2	9.97 (2.81–35.42)	< 0.001
Carbapenem resistant Acinetobacter baumannii	3	1.1	5	9.4	9.13 (2.11–39.48)	< 0.001

OR = odds ratio; CI = confidence interval.

Staphylococcus aureus was less frequent in patients with leukocyte count ≥10,000 mm³ (12.5% vs. 24.1%; OR, 0.45; 95% CI, 0.26–0.75; p = 0.002), while ESBL-producing gram-negative bacteria (14.9% vs. 8.5%; OR, 1.86; 95% CI, 1.03–3.49; p = 0.047) in particularly ESBL-producing Escherichia coli (11.5% vs. 5.0%; OR, 2.45; 95% CI, 1.14–5.27; p = 0.018) were more frequent in patients with leukocyte count ≥10,000 mm³ at diagnosis. Streptococcus spp. (10.1% vs. 4.9%; OR, 2.18; 95% CI, 1.05–4.57; p = 0.034) and Escherichia coli (19.6% vs. 12.4%; OR, 1.71; 95% CI, 1.01–2.91; p = 0.044) were more frequent in patients with C-reactive protein (CRP) ≥130 mg/L. Multi-drug–resistant Pseudomonas aeruginosa (16.3% vs. 8.5%; OR, 2.09; 95% CI, 1.15–3.80; p = 0.015) and Acinetobacter baumannii (7.6% vs. 3.2%; OR, 2.49; 95% CI, 1.14–6.18; p = 0.045) were more frequent in patients with erythrocyte sedimantation rate (ESR) ≥70 mm/h (Table 5).

Table 5.

Univariable Associations between Laboratory Parameters and Causative Micro-Organisms

Leukocyte (mm³)	<10,000 mm³		≥10,000 mm³
	n	%	n	%	OR (CI)	p
Staphylococcus aureus	48	24.1	26	12.5	0.45 (0.26–0.75)	0.002
ESBL+ gram-negative bacteria	17	8.5	31	14.9	1.86 (1.03–3.49)	0.047
ESBL+ Escherichia coli	10	5.0	24	11.5	2.45 (1.14–5.27)	0.018

C-reactive protein (mg/L)	<130 mg/L		≥130 mg/L
	n	%	n	%	OR (CI)	p
Streptococcus spp.	17	4.9	14	10.1	2.18 (1.05–4.57)	0.034
Escherichia coli	43	12.4	27	19.6	1.71 (1.01–2.91)	0.044

ESR (mm/h)	<70 mm/h		≥70 mm/h
	n	%	n	%	OR (CI)	p
MDR Pseudomonas aeruginosa	16	8.5	47	16.3	2.09 (1.15–3.80)	0.015
Acinetobacter baumannii	6	3.2	22	7.6	2.49 (1.14–6.18)	0.045

OR = odds ratio; CI = confidence interval; ESBL = extended-spectrum β-lactamase; ESR = erythrocyte sedimantation rate; MDR = multi-drug–resistant.

In addition, Staphylococcus aureus (22.5% vs. 11.0%; OR, 2.35; 95% CI, 1.39–4.00; p = 0.002), Streptococci (9.9% vs. 1.0%; OR, 10.38; 95% CI, 2.45–44.04; p = 0.002), Pseudomonas aeruginosa (34.0% vs. 24.6%; OR, 1.58; 95% CI, 1.06–2.36; p = 0.025), MDR Pseudomonas aeruginosa (17.3% vs. 10.1%; OR, 1.83; 95% CI, 1.07–3.12; p = 0.026), Acinetobacter baumannii (10.5% vs. 2.7%; OR, 4.17; 95% CI, 1.80–9.67; p < 0.001), and carbapenem-resistant Acinetobacter baumannii (9.4% vs. 1.7%; OR, 5.99; 95% CI, 2.19–16.43; p < 0.001) were more frequent in patients having DFIs within the last three months than those with none.

Discussion

In this 10-year restrospective study, changing epidemiology and antimicrobial resistance in DFIs were evaluated by five-year periods. We presented a detailed retrospective analysis of the microbiologic profiles of 798 micro-organisms in 484 patients with DFIs followed by an interdisciplinary committee at a university hospital. We also investigated several parameters including age, gender, study period, wound location, laboratory parameters, and history of previous DFI in relation to causative agents. In addition, a list of micro-organisms as prognostic indicators of poor clinical outcomes in patients with DFIs were revealed in the present study.

In our study, the patients with DFIs in the second year had improved outcomes in the primary composite end point, re-infection, and mortality, whereas major amputation rates was not different between the first and second five-year periods. This may be due to early surgical intervention, enhanced patient education, and change in patient behavior.

In our study, gram-negative micro-organisms were more frequent compared with gram-positives in both the first and second five-year periods. Gram-negative micro-organisms are often more frequently isolated from DFIs in Asia, Africa, and South America, which have warmer climates,^8–13 whereas gram-positives are more frequently isolated from DFIs in European countries.^14–16 This may be due to the diversity of climatic, social, environmental conditions, personal habits, health systems, or interactions between different micro-organisms.^2,17 In the present study, the prevalence of MDR Pseudomonas aeruginosa and carbapenem resistant Acinetobacter baumannii decreased in the second five-year period. This is probably the effective infection control measures in our unit. However, polymicrobial infections, Streptococcus spp., Corynebacterium spp., and ESBL-producing Escherichia coli. The increase in the prevalence of ESBL-positive infections may be due to the unnecessary use of broad-spectrum antibiotic agents before admission to our unit and the inevitable emergence of resistant strains in circulation.

Macdonald et al.¹⁴ reported a high prevalence of Staphylococcus aureus (n = 65; 32.5%) among all clinical results of patients with DFIs. In their study, which is the first Scottish retrospective report in DFIs, they revealed that Staphylococcus aureus was associated with age but not gender or wound location, and the prevalence of Staphylococcus aureus in DFIs increased with age (p = 0.021). On the other hand, MDR Pseudomonas aeruginosa was associated with advanced age in our study (p = 0.005). It may be due to the high prevalence of MDR colonization, which is thought to be induced with increasing age. In addition, we found an association between gender and polymicrobial infections, Corynebacterium spp. and Citrobacter spp.

In the study by Dörr et al.,¹⁵ with increase in age, prevalence of Staphylococcus aureus and Pseudomonas aeruginosa increased, whereas streptococci became less frequent. Similar to our findings, Corynebacterium spp. was recognized as a growing pathogen among gram-positives. They reported that the forefoot and plantar site infections were mostly caused by gram-positives. However, they did not find any difference in single species between wound location. In our study, Staphylococcus aureus was more frequent in the metatarsal region of the forefoot whereas Acinetobacter baumannii and carbapenem-resistant Acinetobacter baumannii were more frequent in midfoot.

In this study, higher leukocyte counts were determined in Staphylococcus aureus, ESBL-producing gram-negative bacteria, especially in ESBL-producing Escherichia coli. Streptococcus spp. and Escherichia coli were associated with higher CRP levels. Multi-drug–resistant Pseudomonas aeruginosa and Acinetobacter baumannii were associated with higher ESR levels. Dörr et al.¹⁵ reported that inflammatory parameters including leukocyte, CRP, and ESR did not differ between gram-positive and gram-negative micro-organisms in one single species. Another study found poor kidney function and high leukocyte counts as predictors for amputation in patients with DFI.¹⁸ In addition, Aragón-Sánchez et al.¹⁹ found a relation between gram-negatives and higher levels of leukocyte count.

Antimicrobial resistance in DFIs is a fundamental public health concern worldwide. Zubair et al.²⁰ reported a decade ago that 45% of patients were infected with MDR organisms, and the prevalence of ESBL and methicillin resistance was 68.5% and 43.2%, respectively. In the study of Gog et al.²¹ from Malaysia, most patients with DFIs had polymicrobial infections with MDR organisms. In addition, Jouhar et al.²² reported that antibiotimicrobial resistance was significantly related with previous antibiotic use. In our study, MDR Pseudomonas aeruginosa, Acinetobacter baumannii, and carbapenem-resistant Acinetobacter baumannii were associated with poor prognosis in DFIs. In our cohort, patients with DFIs with antibiotic-resistant pathogens had a high rate of major amputation, re-infection, and death. Similarly, Kurup et al.¹² showed that patients with MDR pathogens had more severe infection compared with those with non-MDR organisms (28.1% vs. 16.0%). In a multi-center study from 19 hospitals in Turkey, Klebsiella spp. (OR, 7.7; p = 0.028) was an independent predictor for mortality.²³ Aragón-Sánchez et al.¹⁹ reported that gram-negative micro-organisms in DFIs were associated with the severity of infection. Thus, antimicrobial stewardship is crucial for the management of DFIs.²⁴ In our study, isolation of Staphylococcus aureus was associated with a better prognosis. The fact that our empirical treatment always covers Staphylococcus aureus and includes methicillin-resistant Staphylococcus aureus treatment in risk groups may be related to this result.

In the present study, we demonstrated that Staphylococcus aureus, streptococci, Pseudomonas aeruginosa, MDR Pseudomonas aeruginosa, Acinetobacter baumannii, and carbapenem-resistant Acinetobacter baumannii were more frequent in patients with history of previous DFIs within the last three months. Isolation of MDR pathogens in patients with DFIs could be explained by previous antibiotic exposure or inadequate infection control practices.

This study had several strengths. First, we evaluated and followed up the patients with an interdisciplinary team to accelerate the management of DFIs. Second, we routinely recorded the patients' medical information. Therefore, an accurate information could easily be obtained from the patients' medical charts. Third, we followed up the patients with DFIs for a six-month period. However, our study had some limitations. First, the study was conducted retrospectively. Second, we did not include other relevant risk factors for poor prognosis in DFIs as we focused on the microbiologic evaluation. However, we included IWGDF classification in the multivariable analysis, since IWGDF classification was found as an associated risk factor for poor prognosis in the univariable analysis. Third, a power analysis could not be performed due to the exploratory nature of the study. Last, our results thus may not be generalized. Unfortunately, patients with complicated DFIs who failed previous treatments or who were not appropriately treated may be pre-colonized with resistant bacteria when they visited to our unit.

Conclusions

In conclusion, this 10-year cohort study provides reassuring information about the changing epidemiology of DFIs in a university hospital multidisciplinary follow-up of patients with DFIs and suggests that age, gender, wound location, laboratory parameters, and previous history of DFIs are somehow associated with causative bacterial agents. Therefore, each unit should implement continuous surveillance for DFIs and develop antibiotic use strategies. In addition, clinicians should consider MDR Pseudomonas aeruginosa and Acinetobacter baumannii isolations as poor prognostic determinants in DFIs.

Footnotes

Acknowledgments

The authors acknowledge all healthcare professionals who contribute to the care of the patients.

Authors' Contributions

Conceptualization: Surme, Saltoglu, Balkan. Methodology: Surme, Saltoglu, Balkan, Can. Software: Kurt, Caglar, Ozdemir, Vatan. Validation: Surme, Can. Formal analysis: Surme, Can. Investigation: Surme, Saltoglu, Karaali, Balkan. Resources: Surme, Vatan. Data curation: Surme, Togluk-Yigitoglu, Budak. Writing, original draft preparation: Surme, Saltoglu. Writing, review and editing: Balkan, Arapi, Seker, Can, Gonen, Cetinkale. All authors have reviewed and approved the final version of the article.

The Cerrahpasa Medical Faculty Diabetic Foot Infections Study Group (DAİÇG)

Ahmet Furkan Kurt, Department of Infectious Diseases and Clinical Microbiology, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Ali Seker, Department of Orthopaedics and Traumatology, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Aslı Vatan, Department of Infectious Diseases and Clinical Microbiology, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Berk Arapi, Department of Cardiovascular Surgery, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Beyhan Budak, Department of Infectious Diseases and Clinical Microbiology, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Bilge Caglar, Department of Infectious Diseases and Clinical Microbiology, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Eylem Togluk-Yigitoglu, Department of General Surgery, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Gunay Can, Department of Public Health, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Ilker Inanc Balkan, Department of Infectious Diseases and Clinical Microbiology, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Meryem Ozdemir, Department of Infectious Diseases and Clinical Microbiology, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Mustafa Sait Gonen, Department of Internal Medicine, Division of Endocrinology and Metabolism, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Nese Saltoglu, Department of Infectious Diseases and Clinical Microbiology, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Oguz Cetinkale, Department of Plastic and Reconstructive Surgery, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Ridvan Karaali, Department of Infectious Diseases and Clinical Microbiology, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Semih Baghaki, Department of Plastic and Reconstructive Surgery, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey. Serkan Surme, Department of Infectious Diseases and Clinical Microbiology, Istanbul University-Cerrahpasa, Cerrahpasa Medical Faculty, Istanbul, Turkey.

Funding Information

The authors received no financial support for the research, authorship, and/or publication of this article.

Author Disclosure Statement

The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Uçkay

, Aragón-Sánchez

, Lew

, et al. Diabetic foot infections: What have we learned in the last 30 years?. Int J Infect Dis, 2015; 40:81–91; doi: 10.1016/j.ijid.2015.09.023.

Lipsky

, Senneville É, Abbas

, et al. Guidelines on the diagnosis and treatment of foot infection in persons with diabetes (IWGDF 2019 update). Diabetes Metab Res Rev 2020;36(Suppl 1):e3280; doi: 10.1002/dmrr.3280.

Diabetic foot problems: prevention and management. National Institute for Health and Clinical Excellence (NICE) guideline. August 26, 2015. Available from: www.nice.org.uk/guidance/ng19 [Last accessed: December 15, 2021].

Saltoğlu

, Kılıçoğlu Ö, Baktıroğlu

, et al. Diagnosis, treatment and prevention of diabetic foot wounds and infections: Turkish consensus report. Klimik J, 2015; 28(1):2–34.e

Saltoglu

, Surme

, Ezirmik

, et al. The Effects of antimicrobial resistance and the compatibility of ınitial antibiotic treatment on clinical outcomes in patients with diabetic foot ınfection. Int J Low Extrem Wounds 2021;Apr 15:15347346211004141; doi: 10.1177/15347346211004141.

International Diabetes Federation. Clinical Practice Recommendation on the Diabetic Foot: A guide for health care professionals. 2017. Available from: https://www.idf.org/e-library/guidelines/119-idf-clinical-practice-recommendations-on-diabetic-foot-2017.html [Last accessed July 18, 2022].

van Netten

, Bus

, Apelqvist

, et al. Definitions and criteria for diabetic foot disease. Diabetes Metab Res Rev 2020;36(Suppl 1):e3268; doi: 10.1002/dmrr.3268.

Vatan

, Saltoglu

, Yemisen

, et al. Association between biofilm and multi/extensive drug resistance in diabetic foot infection. Int J Clin Pract 2018;72(3):e13060; doi: 10.1111/ijcp.13060.

Goh

, Bajuri MY

C N

, adarajah

, et al. Clinical and bacteriological profile of diabetic foot infections in a tertiary care. J Foot Ankle Res. 2020;13(1):36; doi: 10.1186/s13047-020-00406-y.

10.

Kateel

, Augustine

, Prabhu

, et al. Clinical and microbiological profile of diabetic foot ulcer patients in a tertiary care hospital. Diabetes Metab Syndr, 2018; 12(1):27–30; doi: 10.1016/j.dsx.2017.08.008.

11.

Ayed

, Ababneh

, Alwin Robert

, et al. Common pathogens and antibiotic sensitivity profiles of ınfected diabetic foot ulcers in Saudi Arabia. Int J Low Extrem Wounds, 2018; 17(3):161–168; doi: 10.1177/1534734618793557.

12.

Kurup

, Ansari

. A study to identify bacteriological profile and other risk factors among diabetic and non-diabetic foot ulcer patients in a Guyanese hospital setting. Diabetes Metab Syndr, 2019; 13(3):1871–1876; doi: 10.1016/j.dsx.2019.04.024.

13.

Goh

, Bajuri MY

C N

, adarajah

, et al. Clinical and bacteriological profile of diabetic foot infections in a tertiary care. J Foot Ankle Res 2020;13(1):36; doi: 10.1186/s13047-020-00406-y.

14.

Macdonald

, Jordan

, Crichton

, et al. A retrospective analysis of the microbiology of diabetic foot infections at a Scottish tertiary hospital. BMC Infect Dis 2020;20(1):218; doi: 10.1186/s12879-020-4923-1.

15.

Dörr

, Freier

, Schlecht

, et al. Bacterial diversity and inflammatory response at first-time visit in younger and older individuals with diabetic foot infection (DFI). Acta Diabetol, 2021; 58(2):181–189; doi: 10.1007/s00592-020-01587-5.

16.

Neves

, Duarte

, Pinto

, et al. Diabetic foot infection: Causative pathogens and empiric antibiotherapy considerations. The experience of a tertiary center. Int J Low Extrem Wounds, 2019; 18(2):122–128; doi: 10.1177/1534734619839815.

17.

Tomic-Canic

, Burgess

, O'Neill

, et al. Skin microbiota and its ınterplay with wound healing. Am J Clin Dermatol. 2020 Sep;21(Suppl 1):36–43; doi: 10.1007/s40257-020-00536-w.

18.

Dörr

, Schlecht

, Chatzitomaris

, et al. Predictive effect of ınflammatory response and foot ulcer localization on outcome in younger and older ındividuals with ınfected diabetic foot syndrome. Exp Clin Endocrinol Diabetes, 2021; 129(12):878–886; doi: 10.1055/a-1149-8989.

19.

Aragón-Sánchez

, Lipsky

, Lázaro-Martínez

. Gram-negative diabetic foot osteomyelitis: Risk factors and clinical presentation. Int J Low Extrem Wounds, 2013; 12(1):63–8; doi: 10.1177/1534734613477423.

20.

Zubair

, Malik

, Ahmad

. Clinico-microbiological study and antimicrobial drug resistance profile of diabetic foot infections in North India. Foot (Edinb), 2011; 21(1):6–14; doi: 10.1016/j.foot.2010.10.003.

21.

Jouhar

, Jaafar

, Nasreddine

, et al. Microbiological profile and antimicrobial resistance among diabetic foot infections in Lebanon. Int Wound J, 2020; 17(6):1764–1773; doi: 10.1111/iwj.13465.

22.

Goh

, Bajuri MY

C N

, adarajah

, et al. Clinical and bacteriological profile of diabetic foot infections in a tertiary care. J Foot Ankle Res 2020;13(1):36; doi: 10.1186/s13047-020-00406-y.

23.

Saltoglu

, Ergonul

, Tulek

, et al. Influence of multidrug resistant organisms on the outcome of diabetic foot infection. Int J Infect Dis, 2018; 70:10–14; doi: 10.1016/j.ijid.2018.02.013..

24.

Uçkay

, Berli

, Sendi

, et al. Principles and practice of antibiotic stewardship in the management of diabetic foot infections. Curr Opin Infect Dis, 2019; 32(2):95–101; doi: 10.1097/QCO.0000000000000530.