Does Metabolic Control Have Any Influence on the Clinical Presentation and Short-Term Outcomes of Diabetic Foot Infections?

Abstract

Objective:

We hypothesized that patients with poor glycemic control undergoing treatment for diabetic foot infections (DFIs) would have a poorer prognosis than those with better metabolic control assessed by glycated hemoglobin (HbA1c).

Approach:

We analyzed a retrospective cohort of 245 patients with moderate and severe DFIs. HbA1c values were dichotomized (<7% or ≥7% and ≤75th percentile (P75) and >P75) to analyze patient outcomes regarding metabolic control. The present study adhered to the STROBE guidelines for cohort studies.

Results:

One hundred sixty-nine patients (69%) were men. Their mean age was 60.7 years (10.8). HbA1c ≥7% was detected in 203 patients (82.9%). P75 HbA1c was 10.9%. After performing univariate analysis, we found an association of HbA1c <7% with major amputations and mortality. However, after applying the logistic regression model, we did not find HbA1c <7% to be a predictive factor of major amputation. The risk factors for mortality following application of Cox's proportional hazards model were osteomyelitis (HR: 0.2, 95% CI: 0.07–0.62, p < 0.01), eGFR <60 mL/min/1.73 m² (HR: 2.7, 95% CI: 1.0–7.5, p = 0.04), and HbA1c <7% (HR: 4.9, 95% CI: 1.8–13.2, p < 0.01).

Innovation:

The group with optimal glycemic control (HbA1c <7%) had a shorter survival time than those with worse metabolic control.

Conclusions:

We did not find a longer duration of hospitalization, a higher rate of amputations, or longer healing times in the groups with worse metabolic control. HbA1c <7% was a risk factor for mid-term mortality.

Javier Aragón-Sánchez, MD, PhD

INTRODUCTION

Diabetic foot ulcer (DFU) is a complication of diabetes mellitus (DM) that is increasing worldwide and may lead to infection, placing the lower extremity at risk. Diabetic foot infections (DFIs) are associated with the need for hospitalization, a major economic burden, and amputations. Risk factors for developing DFUs in populations with DM are mainly distal polyneuropathy, peripheral arterial disease (PAD), foot deformities, and minor trauma.^1,2 The association of hyperglycemia with macro- and microvascular complications in DM has been established, as has the usefulness of glycated hemoglobin (HbA1c) as a value in the assessment of glycemic control. The goal of achieving a value of HbA1c of <7% without significant hypoglycemia has been recommended by the American Diabetes Association (ADA).³ It decreases the risk of micro- and macroangiopathic complications,³ such as distal polyneuropathy, as well as PAD, which are both related to DFUs.⁴ Finally, poor glycemic control has been reported to have a deleterious effect on polymorphonuclear cell function and may increase the risk of infection.⁵ It is, therefore, well accepted that chronic hyperglycemia leads to complications related to the development of DFUs and infectious complications.^6,7 From a theoretical point of view, patients with DFIs and poorer glycemic control may be associated with worse outcomes. However, some authors have reported that HbA1c is not associated with worse outcomes when treating diabetic foot osteomyelitis.⁸ HbA1c was not an independent risk factor for amputation in patients with DFUs in other studies.^9
–11 Therefore, this issue is controversial, and additional research is needed.

CLINICAL PROBLEM ADDRESSED

We hypothesized that patients undergoing treatment for DFIs with poorer glycemic control would be characterized by a different clinical presentation, a longer hospitalization duration, a longer period of antibiotic treatment, a higher rate of major amputations and reinfections, a longer time to healing, and increased mortality. Therefore, clinicians could select the patients in whom to implement intensive and tight control of glucose level during the foot infection episode. It could improve outcomes regarding limb loss and mortality.

MATERIALS AND METHODS

We analyzed a retrospective cohort of patients with moderate and severe DFIs consecutively collected in our diabetic foot research database in the Diabetic Foot Unit at San Juan de Dios Hospital in San José de Costa Rica, Costa Rica from January 1, 2017 to December 31, 2018. Infection was clinically diagnosed based on the presence of local or systemic signs or symptoms of inflammation, as recommended by the International Working Group on the Diabetic Foot (IWGDF).¹² The database contains data on medical histories, laboratory tests, microbiological culture results, duration of antibiotic treatment, number and type of surgical procedures to arrest the infection, and patient follow-up. Mortality was always assessed on the hospital records of every patient and reassessed at the Central Civil Registry of Deaths of Costa Rica. Laboratory test results, including HbA1c, were collected from the database at the initial evaluation of the case. Electronic laboratory notebooks were not used. HbA1c was not routinely tested during follow-up. The severity of the infection was classified according to the scheme developed by the IWGDF.¹² Osteomyelitis (DFO) was diagnosed by a combination of clinical evaluation, a probe-to-bone test, and a simple X-ray in two views.¹³ In the case where any patient was initially treated for soft tissue infection and developed osteomyelitis during follow-up, the case was included as osteomyelitis in the database. PAD was diagnosed on the following basis: previous vascular surgery, absence of one or both foot pulses, and/or ankle–brachial index (ABI) <0.9 or ≥1.2. The vascular team always evaluated patients when PAD was diagnosed, and additional tests and revascularizations were carried out at the discretion of vascular surgeons and were collected in the database. The results of tests performed by the vascular team were not collected in the database. Patients who underwent revascularization were followed up by our department. We took biopsies during the surgical procedures and sent them to the Departments of Microbiology and Pathology. We took additional bone biopsies from the bone margins in cases of osteomyelitis for the Department of Microbiology. Anaerobic cultures were not routinely performed during the study period. Isolated bacteria were collected in our database and were further classified as follows: Gram +, Gram −, and mixed infections combining Gram + and Gram – isolates. Multidrug-resistant organisms included methicillin-resistant Staphylococcus aureus, vancomycin-resistant Enterococcus, Enterobacteriaceae resistant to third-generation cephalosporins and/or carbapenem, and all antimicrobial susceptibility phenotypes of Pseudomonas aeruginosa and Acinetobacter baumannii. ¹⁴

Infection recurrence was defined as any condition in which the healing course was impaired due to an infection clinically and/or radiologically diagnosed and/or cases that needed additional antibiotic therapy and/or surgical treatment. Time to healing was calculated from the day on which the patient underwent the first surgical procedure until the wound(s) was (were) completely epithelialized. Perihospitalization mortality was defined as death within 30 days of admission. After that date, cases of death were considered as mortality during follow-up. Patients are currently in follow-up, and data were censored on February 28, 2021.

The normal distribution of the variables was tested using the Kolmogorov–Smirnov test for continuous variables. For descriptive purposes, we used the mean and standard deviation (SD) for normally distributed continuous variables and the median [quartile 1 (Q1) and quartile 3 (Q3)] for non-normally distributed continuous variables. We used percentages for discrete variables. Variables, even interesting ones, with missing data were excluded from the statistical analysis. We analyzed independence between discrete variables by using the χ ² test and Fisher's exact test, when indicated. Normally distributed quantitative variables were analyzed using Student's t-test for two independent samples. Non-normally distributed quantitative variables were compared by using the nonparametric Mann–Whitney U test for two independent samples. For pairwise comparisons, HbA1c values were dichotomized into <7% or ≥7% and values at or below the 75th percentile (≤P75) and >P75 to analyze the outcomes in patients with worse metabolic control. After performing univariate analysis, we found an association of HbA1c <7% with major amputations and mortality. An ad hoc binary logistic regression model was built to determine if HbA1c <7% remained an independent predictor of major amputation in a multivariate model. Variables with p < 0.05 in univariate analysis were included in a stepwise backward logistic regression analysis. Cox's proportional hazards model for mortality was performed, and hazard ratios (HRs) were estimated. The Kaplan–Meier method adjusted by sex and estimated glomerular filtration rate (eGFR) was undertaken to analyze survival time, calculated from the date of admission to death or the end of the study. Patient data were censored on February 28, 2021 when the status of the patient, dead or alive, was verified using the Central Registry of Costa Rica. The stratified log-rank test was used to determine differences between patients with HbA1c <7% and ≥7%. The statistical analysis was performed with SPSS version 20 for macOS (SPSS, Inc., Chicago, IL), and a p-value <0.05 (two tailed) was set as the threshold for statistical significance.

The present study was conducted following the Declaration of Helsinki guidelines, 2013 revision. Every patient gave verbal consent to the entry of their clinical information into the database and provided written informed consent regarding the surgical procedures. The database was anonymized and approved for research purposes by the Ethics Committee. This study was approved by the Ethics Committee of San Juan de Dios Hospital (Caja Costarricense de Seguro Social), San José, Costa Rica (HSJD-059-CEC-2020).

The STROBE guidelines for cohorts were adhered to in the present study.

RESULTS

The cohort consisted of 245 patients who underwent treatment for DFIs. One hundred sixty-nine patients (69%) were men; 76 (31%), women. The mean age was 60.7 years (10.8). No significant lifestyle differences were identified regarding tobacco and alcohol use. Seventy-eight patients (31.8%) had a history of previous amputations, 18 (7.3%) of them major. Hypertension was the most frequent comorbidity and was present in 190 patients (77.6%). The median duration of diabetes was 15 years (Q1 = 8, Q3 = 20). Regarding diabetes treatment, 26 patients did not undergo any treatment, 38 (15.5%) took oral hypoglycemic agents, 68 (27.8%) took oral hypoglycemic agents and insulin, and 113 (46.1%) took insulin. Mean HbA1c was 9.30% (SD 2.2). HbA1c ≥7% was detected in 203 patients (82.9%). P75 HbA1c was 10.9%. HbA1c >10.9% was reported in 61 patients (24.9%). Clinical variables related to demographics, clinical presentation, and outcomes of infection are shown in Table 1. Laboratory values are shown in Table 2. Glucose levels were significantly higher in patients with HbA1c ≥7% and >10.9%. Renal function was significantly better in patients with HbA1c ≥7% when comparing eGFR between the two groups (p = 0.03). Patients with HbA1c >10.9% had a significantly higher inflammatory response, determined by the erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). These inflammatory markers were not significantly different between Gram-positive and Gram-negative isolates.

Table 1.

Univariate analysis among patients regarding HbA1c values

	Series (n = 245)	HbA1c<7% (n = 42)	HbA1c≥7% (n = 203)	p-value	HbA1c≤10.9% (n = 184)	HbA1c>10.9% (n = 61)	p-value
Age in years, median (SD)	60.7 (10.8)	65.0 (12.1)	59.8 (10.3)	<0.01	62.1 (10.8)	56.5 (9.5)	<0.01
Men/Women, n (%)	169 (69)/76 (31)	30 (71.4)/12 (28.6)	139 (68.5)/64 (31.5)	0.70	54 (29.3)	22 (36.1)	0.32
Duration of diabetes in years, median (Q1, Q3)	15 (8, 20)	13 (6, 20)	15 (9, 20)	0.21	15 (8, 20)	14 (8, 23)	0.94
Previous amputation, n (%)	78 (31.8)	19 (45.2)	59 (29.1)	0.04	60 (32.6)	18 (29.5)	0.65
Previous major amputation, n (%)	18 (7.3)	5 (11.9)	13 (6.4)	0.21	15 (8.2)	3 (4.9)	0.40
Hypertension, n (%)	190 (77.6)	34 (81)	156 (76.8)	0.56	152 (82.6)	38 (62.3)	<0.01
Previous stroke, n (%)	7 (2.9)	1 (2.4)	6 (3)	0.83	6 (3.3)	1 (1.6)	0.68
Insulin use, n (%)	181 (73.9)	31 (73.8)	150 (73.9)	0.99	143 (77.7)	38 (62.3)	0.01
PAD, n (%)	119 (48.6)	23 (54.8)	96 (47.3)	0.37	94 (51.1)	25 (41)	0.17
No distal pulses, n (%)	78 (31.8)	16 (38.1)	62 (30.5)	0.33	64 (34.8)	14 (23)	0.08
Foul smelling, n (%)	150 (61.2)	24 (57.1)	126 (62.1)	0.55	111 (60.3)	39 (63.9)	0.61
Skin necrosis, n (%)	151 (61.6)	22 (52.4)	129 (63.5)	0.17	114 (62)	37 (60.7)	0.85
Osteomyelitis, n (%)	197 (80.4)	37 (88.1)	160 (78.8)	0.16	150 (81.5)	47 (77)	0.44
Moderate/Severe, n (%)	166 (67.8)/79 (32.2)	30 (71.4)/12 (28.6)	136 (67)/67 (33)	0.57	125 (67.9)/59 (32.1)	41 (67.2)/20 (32.8)	0.91
Penetrating trauma, n (%)	96 (9.2)	14 (33.3)	82 (40.4)	0.39	66 (35.9)	30 (49.2)	0.06
Shoes as trigger of the foot lesions, n (%)	76 (31)	11 (26.2)	65 (32)	0.45	59 (32.1)	17 (27.9)	0.53
Ulcer evolution in days, median (Q1, Q3)	15 (10, 36.5)	30 (15, 90)	15 (8, 30)	0.03	20 (10, 45)	15 (8, 30)	0.30
Antibiotic treatment before evaluation in our department, n (%)	54 (22)	11 (26.2)	43 (21.2)	0.47	36 (19.6)	18 (29.5)	0.10
No growths in cultures at baseline, n (%)	20 (8.2)	4 (9.5)	16 (7.9)	0.72	18 (9.8)	2 (3.3)	0.17
Gram-positive isolates^a, n (%)	70 (31.4)	8 (21.6)	62 (33.3)	0.16	42 (25.6)	28 (47.5)	<0.01
Gram-negative isolates^a, n (%)	79 (35.4)	18 (48.6)	61 (32.8)	0.06	64 (39)	15 (25.4)	0.06
MDRO, n (%)	63 (28)	11 (28.9)	52 (27.8)	0.88	45 (27.1)	18 (30.5)	0.61
Polymicrobial infections, n (%)	138 (63.1)	24 (63.2)	114 (61)	0.80	104 (62.7)	34 (57.6)	0.49
Revascularization, n (%)	71 (29)	12 (28.6)	59 (29.1)	0.94	56 (30.4)	15 (24.6)	0.38
No surgical treatment, n (%)	43 (17.6)	4 (9.5)	39 (19.2)	0.13	29 (15.8)	14 (23)	0.20
Conservative surgery, n (%)	62 (25.3)	14 (33.3)	48 (23.6)	0.18	52 (28.3)	10 (16.4)	0.06
Minor amputation, n (%)	96 (39.2)	12 (28.6)	84 (41.4)	0.12	69 (37.5)	27 (44.3)	0.34
Major amputation, n (%)	44 (18)	12 (28.6)	32 (15.8)	0.04	34 (18.5)	10 (16.4)	0.71
Any type of amputation, n (%)	140 (57.1)	24 (57.1)	116 (57.1)	1.0	103 (56)	37 (60.7)	0.52
Hospital stay, median (Q1, Q3)	14 (9, 22)	12 (8, 24)	15 (9, 21)	0.56	14 (8.5, 22)	15 (10, 21)	0.67
Duration of antibiotics in days, median (Q1, Q3)	13 (7, 23)	11.5 (7, 31)	13 (7, 22)	0.95	12 (7, 24)	14 (7, 22)	0.51
Recurrence of the infection, n (%)	34 (13.9)	8 (19)	37 (18.2)	0.9	22 (12)	12 (19.7)	0.13
Time to healing in weeks, median (SD)	18.5 (11.5, 27.1)	16.7 (10.9)	21.1 (14.8)	0.07	18.4 (11.2, 27.1)	18.5 (13.7, 27.1)	0.63
Death during follow-up, n (%)	18 (7.3)	8 (19)	10 (4.9)	<0.01	17 (9.2)	1 (1.6)	0.04

Mono or polymicrobial cultures.

Table 2.

Laboratory values among patients regarding HbA1c values

	Series (n = 245)	HbA1c ≤7% (n = 42)	HbA1c >7% (n = 203)	p-value	HbA1c ≤10.9%	HbA1c >10.9%	p-value
Glucose mg/dL at admission, median (Q1, Q3)	179 (138, 271)	143 (122, 187)	194 (143.5, 285)	<0.01	174 (132, 256)	207 (153.5, 306)	0.01
Blood urea nitrogen, mg/dL median (Q1, Q3)	18 (13, 25)	21.5 (15, 37)	17 (13, 24)	0.01	19 (14, 26.5)	18 (12, 24)	0.13
Serum creatinine mg/dL, median (Q1, Q3)	1 (0.8, 1.4)	1.1 (0.8, 1.7)	1.0 (0.8, 1.3)	0.04	1.0 (0.8, 1.4)	1.0 (0.8, 1.2)	0.44
Sodium mM, median (Q1, Q3)	134 (132, 137)	134 (131.5, 136.5)	134 (132, 137)	0.77	134 (132, 137)	133 (130, 135)	<0.01
eGFR mL/min/1.73 m,² mean (SD)	69.3 (32.4)	59.4 (32.9)	71.3 (32.0)	0.03	68.3 (32.8)	72.1 (31.3)	0.43
Albumin g/dL, median (Q1, Q3)	2.9 (2.6, 3.2)	2.9 (2.8, 3.4)	2.9 (2.6, 3.2)	0.24	2.9 (2.7, 3.3)	2.9 (2.5, 3.1)	0.11
Hemoglobin level g/dL, median (SD)	11.4 (2.1)	10.9 (2.0)	11.5 (2.1)	0.07	11.4 (2.1)	11.3 (2.0)	0.78
WBC, cells/mm³ count, mean (SD)	14531.3 (5219.0)	13330.9 (5149.8)	14779.7 (5211.3)	0.1	14314.2 (5219.5)	15186.3 (5205.4)	0.25
Erythrocyte sedimentation rate, mm/h median (Q1, Q3)	90 (67, 107)	86 (66, 102)	91 (69, 107)	0.37	88 (65, 106)	96 (82, 111)	0.04
C-reactive protein, mg/dL, median (Q1, Q3)	24.8 (9.1, 36.5)	22.7 (7.1, 29.8)	25.9 (9.1, 36.5)	0.14	23.4 (6.2, 35.3)	30.5 (19.6, 41.1)	<0.01

Patients with worse metabolic control (Hb1c ≥7% and HbA1c >10.9%) were younger that those included in the groups with better metabolic control (p < 0.01). Patients with HbA1c <7% had undergone more previous amputations (45.2%) than those with HbA1c ≥7% (29.1%), p = 0.04. The median duration of the ulcer was 30 days in patients with HbA1c <7% and 15 days in those with HbA1c ≥7%, p = 0.03. Finally, 28.6% of patients with HbA1c <7% had undergone major amputations versus 15.8% in those with HbA1c ≥7%, p = 0.04 (Fig. 1). Patients with HbA1c ≤10.9% had a significantly greater frequency of history of hypertension, used insulin at a higher rate, and presented Gram-positive isolates at lower rate than those with HbA1c >10.9%.

Figure 1.

Histogram showing univariate analysis regarding major amputations and mortality.

After the logistic regression model was applied, HbA1c <7% was not a risk factor for undergoing a major amputation in the present cohort of patients.

The mean follow-up time was 36.7 months: 34.2 months in those with HbA1c <7% and 37.2 months in those with HbA1c ≥7%. Three patients were lost to follow-up. All patients, except these three and three others who died before the follow-up period, achieved healing after a median of 18.5 weeks (Q1 = 11.5, Q3 = 27.1). We found significant differences regarding overall mortality when comparing patients with HbA1c values ≤10.9% and >10.9% (p = 0.04) and patients with HbA1c values <7% and ≥7% (p < 0.01), (Fig. 1).

No perihospitalization mortality was found, but 18 patients died during follow-up. Risk factors of mortality following application of the Cox's proportional hazards model were osteomyelitis (HR: 0.2, 95% CI: 0.07–0.62, p < 0.01), eGFR <60 mL/min/1.73 m² (HR: 2.7, 95% CI: 1.0–7.5, p = 0.04), and HbA1c <7% (HR: 4.9, 95% CI: 1.8–13.2, p < 0.01).

Global survival time was 47.6 months (95% CI: 456.5–48.7). Mean survival time in patients with HbA1c ≤10.9% was 47.0 months (95% CI: 45.6–48.4) and 49.3 months (95% CI: 48.0–50.6) for patients with HbA1c >10.9%. The log-rank (Mantel–Cox) test between the two groups resulted in χ ² = 3.7 (p = 0.05). The mean survival time in patients with HbA1c <7% was 42.4 months (95% CI: 38.7–46.0) and 48.4 months (95% CI: 47.4–49.4) for patients with HbA1c ≥7%. The log-rank (Mantel–Cox) test between the two groups was χ ² = 12.5 (p < 0.01). The Kaplan–Meier curve is shown in Fig. 2.

Figure 2.

Kaplan–Meier survival curve regarding metabolic control.

DISCUSSION

The results of our study did not support the hypothesis that patients with poorer metabolic control would have worse outcomes of DFI treatment. By contrast, patients with better metabolic control in our experience had a shorter survival time. Regarding clinical presentation of infection, Gram-positive bacteria were isolated at a higher rate in patients with HbA1c >10.9%, and patients with HbA1c ≥7% had a shorter ulcer duration. Patients with HbA1c >10.9% had a significantly higher inflammatory response and a lower mean sodium level. No other differences in clinical presentation were found. In the present study, we used two cutoffs for research purposes: 7%, which is the level recommended by the ADA, and P75 to define patients with worse metabolic control in the present cohort. Most patients in our series underwent surgical treatment to arrest the infection. There is some evidence associating HbA1c levels with postoperative complications and surgical-site infections following elective foot and ankle surgery.^15,16 However, the impact of metabolic control as determined by HbA1c on the outcomes of treatment of DFIs is not sufficiently understood.

We found that patients with HbA1c >P75 were younger. It has recently been reported that older patients, despite having more unfavorable factors for adequate glycemic control, had a lower risk of having high values of HbA1c.¹⁷ This finding is consistent with our results. Furthermore, although the duration of diabetes was the same in both groups, our patients with poorer metabolic control were treated with insulin at a lower rate. This is probably associated with suboptimal treatment of diabetes in that group and could be related to poor adherence to insulin treatment, as demonstrated in other series.¹⁸ It is also associated with younger patients and a lower socioeconomic level, consistent with our population.¹⁸

Ulcer duration was longer, that is, a median of 30 days, in patients with good metabolic control at a cutoff of 7%. We were not able to find similar results in the medical literature. Evolution of ulcers longer than 30 days has been a predictor of lower extremity amputation (LEA) in the presence of HbA1c values greater than those found in our analysis.⁹ The association of infection and amputation with a longer ulcer duration has been widely established.^11,19,20 Patients with better metabolic control also had a higher rate of history of amputations. We could theorize that patients fear a new amputation and put off medical consultation until the infection is evident.

We found no difference in the severity of infection according to the IWGDF severity scheme. This finding is consistent with one study that did not find differences in HbA1c values among patients with moderate and severe infections.²¹ However, other authors reported that HbA1c values were statistically significantly higher in patients with severe infections when compared with those with moderate ones.²² HbA1c was not a prognosis factor in that experience. No differences in HbA1c values were found between patients who underwent amputation and those who did not, nor between those who underwent any major amputation and those who died.²²

According to laboratory findings, patients with HbA1c >P75 had lower values of sodium and higher ESR and CRP levels. These findings could be related to a more severe inflammatory response in patients with higher values of HbA1c. Although higher values of CRP and the ESR are likely related to infection in the present series, they could also be related to a more severe state of subclinical inflammation that has been reported in patients with worse metabolic control.²³ Furthermore, high levels of CRP are associated with advanced stages of atherosclerosis in type 2 diabetic patients, especially those with high HbA1c levels.²⁴ This finding regarding CRP and ESR was not apparent when the cutoff was fixed at 7%.

We hypothesized that poor metabolic control would have a deleterious effect on wound healing, observed as a delayed healing time. However, we did not find any difference in healing time between groups. The impact of metabolic control on wound healing has been debated in many series, but no definitive conclusion has been reached.²⁵ One study reported that higher HbA1c values and high variability over time were associated with longer healing times in DFUs.²⁶ Other authors demonstrated a positive association between wound healing with an increase in the baseline level of HbA1c during the healing period.²⁷ Still other authors suggested that a baseline HbA1c between 7% and 8% during treatment would benefit ulcer healing more than an HbA1c <7%.²⁸ We believe that healing time was not delayed in patients with worse metabolic control due to the characteristics of our series. However, this study dealt with complicated DFUs. All our patients had acute infections requiring surgery in most cases. After surgical debridement and/or any type of amputation, chronic DFUs, characterized by chronic inflammation and a deleterious wound microenvironment, turn into acute wounds.²⁹ A previous study compared conservative versus surgical management of neuropathic foot ulcers.³⁰ No differences regarding HbA1c were found between the two groups, but healing time was shorter in surgical patients.³⁰ Another study dealing with DFIs, specifically with surgical osteomyelitis, did not find significant differences in healing times between patients according to HbA1c values.⁸ Both experiences are consistent with the present study.

In our analysis, we found no differences when comparing groups with poorer metabolic control regarding the amputation rate. However, we found an apparent controversial association: the group with HbA1c <7% had a higher rate of history of previous amputations and underwent a higher rate of major amputations during the index episode in the univariate analysis. However, this association did not remain after we applied a logistic regression model. HbA1c has been documented as a predictor of LEA.^31,32 Another study found that levels of HbA1c >7.5% were associated with a decreased risk of amputation.³³ When the authors studied minor and major amputations separately, HbA1c >7.5% remained as a factor that decreased the risk of minor but not major amputations.³³ We found no between-group differences in the amputation rate. A study reporting a mean HbA1c of 8.2% found that amputation was not associated with lower values of HbA1c.³⁴ Other authors also reported higher values of HbA1c a successfully treated group when compared with a group that underwent amputations.³⁵ However, the difference was not statistically significant.³⁵ In our experience, glucose at admission was significantly different when considering both cutoff points (HbA1c ≥7% and >10.9%). Higher values of glucose at admission may be a consequence of previous poor metabolic control but could also be associated with infection severity. In one report, glucose levels at admission were significantly higher in severe infections when compared with moderate ones.³⁶ Other authors reported that worse metabolic control during admission, determined by glucose monitoring three times a day, was a predictive factor for amputation.⁸

Readmission of 47.4% of the patients with HbA1c >10% due to recurrence or progression of the infection was reported in one series.³⁷ However, we did not find differences regarding the readmission rate in the group with worse metabolic control.

Even though we had a short follow-up period, the results of our analysis showed that HbA1c <7% was a risk factor for mortality. Patients with HbA1c ≥7% had a longer survival time after adjusting for sex, age, and eGFR. Some studies are consistent with our findings.^34,38 The most accepted explanation for this apparently controversial issue is that episodes of hypoglycemia occur in populations with high cardiovascular risk while they are trying to achieve tight glucose control.^39,40 Other authors have suggested chronic kidney disease as another risk factor for hypoglycemia associated with mortality.⁴¹ This is consistent with our research, since the group with HbA1c <7% had lower values of eGFR, which could play an additional role in the occurrence of hypoglycemia in that group. Indeed, we have found that eGFR <60 mL/min/1.73 m² was a risk factor for mortality. Our group reported the same findings in a previous study in which patients were followed up for 8 years.⁴² We did not find differences in survival times when the cutoff of HbA1c was established at 10.9%, which is consistent with the aforementioned theories. However, it is difficult to attribute these findings to foot infections treated in our department because only three patients died before healing. It is most probable that these risk factors of mortality are related to the clinical course of diabetes over the years. We could have provided more information if the cause of the death had been collected in the database. We found an unexpected infection-related “protective” factor against mortality: history of treatment for osteomyelitis. It is not easy to understand the meaning of this finding. We did not find similar information in the medical literature to the best of our knowledge. We previously reported that osteomyelitis was not associated with worse prognosis when compared with soft tissue infections, although mortality was not addressed in that study.⁴³ Other authors have associated the presence of osteomyelitis with worse prognosis.^44,45 Additional research will be necessary to interpret this controversial association.

Our study has some limitations. First, it was retrospective in nature. Second, our results are based on the baseline HbA1c on admission to our hospital; no serial determinations were carried out. Third, patients with noninfected DFUs and those with mild infections were excluded from the present study. Fourth, the small size of the sample of patients with good metabolic control may limit the value of the results. Fifth, the cause of death was not collected in our database. Finally, most of our patients underwent surgery, which could produce a bias in the results.

This study has some strengths: all surgical procedures and follow-up were performed by the same team, meaning that there was no bias in the indication of amputation. Only a few patients were lost to follow-up and mortality was always assessed using the Civil Death National Registry.

INNOVATION

We designed the present study because it was not clear whether high levels of HbA1c are associated with poorer prognosis when patients undergo treatment for DFIs. Few studies have specifically dealt with the impact of HbA1c on the outcomes of DFI treatment. We found that patients with poor metabolic control did not have a longer duration of hospitalization, a higher amputation rate, or longer healing times. Patients with HbA1c <7% had a shorter mid-term survival time.

KEY FINDINGS

Poor metabolic control was not associated with worse outcomes in cases of DFIs in our series, in which most patients underwent surgical treatment.

Patients with worse metabolic control (HbA1c >10.9%) were younger, presented a higher rate of Gram-positive isolates, and had higher values of inflammatory biomarkers and lower sodium values. Nevertheless, they did not have a longer duration of hospitalization, a higher rate of amputations, or longer healing times.

The group with optimal glycemic control (HbA1c <7%) had a shorter survival time.

Footnotes

AUTHOR DISCLOSURE AND GHOSTWRITING

The authors declare that there is no conflict of interest. No ghostwriter was used for the preparation of this article.

AUTHORSHIP CONFIRMATION STATEMENT

J.A.-S. designed the clinical database used in this study, analyzed the data, performed the statistical analysis, and wrote the article. G.V.-M. collected data from the beginning of the project, created the database for research purpose, and edited and contributed to the writing of the article. M.E.L.-V. contributed to the writing of the article, especially, the discussion and reviewing the English language. J.M.R.-B. researched the data reviewed/edited the article, and contributed to the discussion. C.M.-V. researched the data, reviewed/edited the article, and contributed to the discussion.

ABOUT THE AUTHORS

Javier Aragón-Sánchez, MD, PhD, is a General Surgeon. He is the Medical Director and the Chief of the Department of Surgery and Diabetic Foot Unit of La Paloma Hospital in Las Palmas de Gran Canaria (Spain). He is the author of three textbooks, more than 22 chapters, and more than 95 peer-reviewed articles and several monographs, especially dealing with diabetic foot infections and osteomyelitis. He is a member of the Board of the IWGDF infections. Gerardo Víquez-Molina, MD, is the Chief of the Diabetic Foot Unit at San Juan de Dios Hospital in San José de Costa Rica, Costa Rica. His main research fields are diabetic foot infections and osteomyelitis. María Eugenia López-Valverde, MD, is an Endocrinologist interested in the epidemiology, prevention, and management of diabetic foot ulcers and amputations. She is developing a study dealing with nutrition and outcomes of treatment of foot ulcers. José María Rojas-Bonilla, MD and Christian Murillo-Vargas, MD, are staff members of the Diabetic Foot Unit at San Juan de Dios Hospital in San José de Costa Rica, Costa Rica. They are interested in research of complicated foot problems managed in a multidisciplinary way.

FUNDING INFORMATION

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Abbreviations and Acronyms

References

Boyko

, Ahroni

, Stensel

, Forsberg

, Davignon

, Smith

. A prospective study of risk factors for diabetic foot ulcer. The Seattle Diabetic Foot Study. Diabetes Care, 1999; 22:1036–1042.

Armstrong

, Boulton

AJM

, Bus

. Diabetic foot ulcers and their recurrence. N Engl J Med, 2017; 376:2367–2375.

American Diabetes

6. Glycemic Targets: Standards of Medical Care in Diabetes-2021. Diabetes Care, 2021; 44:S73–S84.

Tesfaye

, Chaturvedi

, Eaton

, et al. Vascular risk factors and diabetic neuropathy. N Engl J Med, 2005; 352:341–350.

Delamaire

, Maugendre

, Moreno

, Le Goff

, Allannic

, Genetet

. Impaired leucocyte functions in diabetic patients. Diabet Med, 1997; 14:29–34.

Arya

, Binney

, Khakharia

, et al. High hemoglobin A1c associated with increased adverse limb events in peripheral arterial disease patients undergoing revascularization. J Vasc Surg, 2018; 67:217–228 e211.

Cha

, Kim

, Ko

, et al. Influence of preprocedural glycemic control on clinical outcomes of endovascular therapy in diabetic patients with lower extremity artery disease: an analysis from a Korean multicenter retrospective registry cohort. Cardiovasc Diabetol, 2020; 19:97.

Aragon-Sanchez

, Lazaro-Martinez

. Impact of perioperative glycaemia and glycated haemoglobin on the outcomes of the surgical treatment of diabetic foot osteomyelitis. Diabetes Res Clin Pract, 2011; 94:e83–e85.

Ugwu

, Adeleye

, Gezawa

, Okpe

, Enamino

, Ezeani

. Predictors of lower extremity amputation in patients with diabetic foot ulcer: findings from MEDFUN, a multi-center observational study. J Foot Ankle Res, 2019; 12:34.

10.

Pickwell

, Siersma

, Kars

, et al. Predictors of lower-extremity amputation in patients with an infected diabetic foot ulcer. Diabetes Care, 2015; 38:852–857.

11.

Uysal

, Arda

, Tasbakan

, et al. Risk factors for amputation in patients with diabetic foot infection: a prospective study. Int Wound J, 2017; 14:1219–1224.

12.

Lipsky

, Aragon-Sanchez

, Diggle

, et al. IWGDF guidance on the diagnosis and management of foot infections in persons with diabetes. Diabetes Metab Res Rev, 2016; 32 Suppl 1:45–74.

13.

Aragon-Sanchez

, Lipsky

, Lazaro-Martinez

. Diagnosing diabetic foot osteomyelitis: is the combination of probe-to-bone test and plain radiography sufficient for high-risk inpatients?. Diabet Med, 2011; 28:191–194.

14.

Henig

, Pogue

, Martin

, et al. The impact of multidrug-resistant organisms on outcomes in patients with diabetic foot infections. Open Forum Infect Dis, 2020; 7:ofaa161.

15.

Domek

, Dux

, Pinzur

, Weaver

, Rogers

. Association between hemoglobin A1c and surgical morbidity in elective foot and ankle surgery. J Foot Ankle Surg, 2016; 55:939–943.

16.

Wukich

, Crim

, Frykberg

, Rosario

. Neuropathy and poorly controlled diabetes increase the rate of surgical site infection after foot and ankle surgery. J Bone Joint Surg Am, 2014; 96:832–839.

17.

Sia

, Kor

, Tu

, Liao

, Chang

. Predictors of treatment failure during the first year in newly diagnosed type 2 diabetes patients: a retrospective, observational study. PeerJ, 2021; 9:e11005.

18.

Osborn

, Gonzalez

. Measuring insulin adherence among adults with type 2 diabetes. J Behav Med, 2016; 39:633–641.

19.

Lavery

, Armstrong

, Wunderlich

, Mohler

, Wendel

, Lipsky

. Risk factors for foot infections in individuals with diabetes. Diabetes Care, 2006; 29:1288–1293.

20.

Margolis

, Allen-Taylor

, Hoffstad

, Berlin

. Diabetic neuropathic foot ulcers: the association of wound size, wound duration, and wound grade on healing. Diabetes Care, 2002; 25:1835–1839.

21.

Wukich

, Hobizal

, Raspovic

, Rosario

. SIRS is valid in discriminating between severe and moderate diabetic foot infections. Diabetes Care, 2013; 36:3706–3711.

22.

Lin

, Hung

, Huang

, Yeh

, Huang

. Diabetic foot infection presenting systemic inflammatory response syndrome: a unique disorder of systemic reaction from infection of the most distal body. J Clin Med, 2019; 8:1538.

23.

Elimam

, Abdulla

, Taha

. Inflammatory markers and control of type 2 diabetes mellitus. Diabetes Metab Syndr, 2019; 13:800–804.

24.

Bosevski

, Stojanovska

, Apostolopoulos

. Inflammatory biomarkers: impact for diabetes and diabetic vascular disease. Acta Biochim Biophys Sin (Shanghai), 2015; 47:1029–1031.

25.

Rastogi

, Mukhopadhyay

, Sahoo

, et al. Intensive Glycemic Control for Diabetic Foot Ulcer Healing: A Multicentric, Randomized, Parallel Arm, Single-Blind, Controlled Study Protocol (INGLOBE Study). Int J Low Extrem Wounds, 2020 [Epub ahead of print]; DOI: 10.1177/1534734620952245.

26.

Dhatariya

, Li Ping Wah-Pun Sin

, Cheng

JOS

, et al. The impact of glycaemic variability on wound healing in the diabetic foot - A retrospective study of new ulcers presenting to a specialist multidisciplinary foot clinic. Diabetes Res Clin Pract, 2018; 135:23–29.

27.

Fesseha

, Abularrage

, Hines

, et al. Association of hemoglobin A1c and wound healing in diabetic foot ulcers. Diabetes Care, 2018; 41:1478–1485.

28.

Xiang

, Wang

, He

, Xu

, Zhang

, Tang

. Reasonable glycemic control would help wound healing during the treatment of diabetic foot ulcers. Diabetes Ther, 2019; 10:95–105.

29.

Ferguson

, Herrick

, Spencer

, Shaw

, Boulton

, Sloan

. The histology of diabetic foot ulcers. Diabet Med, 1996; 13 Suppl 1:S30–S33.

30.

Piaggesi

, Schipani

, Campi

, et al. Conservative surgical approach versus non-surgical management for diabetic neuropathic foot ulcers: a randomized trial. Diabet Med, 1998; 15:412–417.

31.

Goldman

, Clark

, Craven

, et al. Effect of intensive glycemic control on risk of lower extremity amputation. J Am Coll Surg, 2018; 227:596–604.

32.

Pemayun

, Naibaho

, Novitasari

, Amin

, Minuljo

. Risk factors for lower extremity amputation in patients with diabetic foot ulcers: a hospital-based case-control study. Diabet Foot Ankle, 2015; 6:29629.

33.

Won

, Chung

, Park

, et al. Risk factors associated with amputation-free survival in patient with diabetic foot ulcers. Yonsei Med J, 2014; 55:1373–1378.

34.

Winkley

, Stahl

, Chalder

, Edmonds

, Ismail

. Risk factors associated with adverse outcomes in a population-based prospective cohort study of people with their first diabetic foot ulcer. J Diabetes Complications, 2007; 21:341–349.

35.

Lee

, Kim

, Lee

, Choi

. Risk factors of treatment failure in diabetic foot ulcer patients. Arch Plast Surg, 2013; 40:123–128.

36.

Ryan

, Crisologo

, Oz

, La Fontaine

, Wukich

, Lavery

. Do SIRS Criteria Predict Clinical Outcomes in Diabetic Skin and Soft Tissue Infections?. J Foot Ankle Surg, 2019; 58:1055–1057.

37.

Ang

, Lim

. Recurrent admissions for diabetic foot complications. Malays Orthop J, 2013; 7:21–26.

38.

Rubio

, Jimenez

, Lazaro-Martinez

. Mortality in Patients with Diabetic Foot Ulcers: Causes, Risk Factors, and Their Association with Evolution and Severity of Ulcer. J Clin Med, 2020; 9:3009.

39.

Riddle

, Ambrosius

, Brillon

, et al. Epidemiologic relationships between A1C and all-cause mortality during a median 3.4-year follow-up of glycemic treatment in the ACCORD trial. Diabetes Care, 2010; 33:983–990.

40.

McAlister

, Zheng

, Westerhout

, et al. Association between glycated haemoglobin levels and cardiovascular outcomes in patients with type 2 diabetes and cardiovascular disease: a secondary analysis of the TECOS randomized clinical trial. Eur J Heart Fail, 2020; 22:2026–2034.

41.

Moen

, Zhan

, Hsu

, et al. Frequency of hypoglycemia and its significance in chronic kidney disease. Clin J Am Soc Nephrol, 2009; 4:1121–1127.

42.

Aragon-Sanchez

, Viquez-Molina

, Lopez-Valverde

, Aragon-Hernandez

, Rojas-Bonilla

, Murillo-Vargas

Long-term mortality of a cohort of patients undergoing surgical treatment for diabetic foot infections. An 8-year Follow-up Study. Int J Low Extrem Wounds 2021. [Epub ahead of print]; DOI:10.1177/15347346211009425.

43.

Aragon-Sanchez

, Viquez-Molina

, Lopez-Valverde

, Rojas-Bonilla

, Murillo-Vargas

Surgical diabetic foot infections: is osteomyelitis associated with a worse prognosis? Int J Low Extrem Wounds 2021. [Epub ahead of print]; DOI: 10.1177/1534734620986695.

44.

Lavery

, Ryan

, Ahn

, et al. The Infected Diabetic Foot: Re-evaluating the Infectious Diseases Society of America Diabetic Foot Infection Classification. Clin Infect Dis, 2020; 70:1573–1579.

45.

Wukich

, Hobizal

, Sambenedetto

, Kirby

, Rosario

. Outcomes of osteomyelitis in patients hospitalized with diabetic foot infections. Foot Ankle Int, 2016; 37:1285–1291.