Abstract
Background
Patients with diabetes mellitus are susceptible to bacterial infections and may be exposed to repeated or prolonged antimicrobial therapy, which can increase the risk of multidrug-resistant organism (MDRO)-positive cultures. However, the distribution of MDROs across different specimen types and the relationship between glycemic control and MDRO isolation remain insufficiently characterized in regional clinical microbiology settings. This study investigated specimen-specific MDRO profiles and evaluated the association between glycated hemoglobin (HbA1c) levels and MDRO-positive cultures in adults with diabetes.
Methods
This retrospective clinical microbiology study was conducted at a regional teaching hospital in southern Taiwan. Culture-positive isolates obtained from adult patients between January and December 2023 were reviewed. After predefined exclusions, 1989 isolates from abscess, urine, sputum, and blood specimens were included. The analysis focused on six predefined clinically monitored MDRO phenotypes routinely reported by the hospital clinical microbiology laboratory and infection-control program. Diabetes status was determined from documented physician diagnosis in the electronic medical record and/or laboratory evidence consistent with national diagnostic criteria. Poor glycemic control was defined as HbA1c > 8.0%. Bacterial identification and antimicrobial susceptibility testing were performed using the VITEK-2 Compact automated system, with confirmatory disk diffusion testing when indicated. Associations among diabetes status, specimen type, demographic factors, biochemical parameters, and MDRO-positive cultures were analyzed using chi-square tests, t tests, and logistic regression models.
Results
Among 1989 culture-positive isolates, 902 were obtained from patients with diabetes and 1087 from patients without diabetes. The overall MDRO-positive proportions were 19.5% in the diabetes group and 17.5% in the non-diabetes group. Urine was the most frequent specimen source in patients with diabetes. MDRO distribution differed significantly by specimen type. Methicillin-resistant Staphylococcus aureus predominated in abscess and blood specimens, vancomycin-resistant Enterococcus predominated in urine specimens, and carbapenem-resistant Acinetobacter baumannii and carbapenem-resistant Pseudomonas aeruginosa were frequent in sputum specimens. In patients with diabetes, HbA1c > 8.0% was associated with increased odds of MDRO-positive cultures after adjustment for age, sex, and specimen type (adjusted odds ratio, 2.82; 95% confidence interval 1.84–4.32; p < 0.001). Routine lipid markers and uric acid were not significantly associated with MDRO status.
Conclusions
Poor glycemic control was associated with MDRO-positive cultures in adults with diabetes. These findings support specimen-specific MDRO surveillance and suggest that glycemic-control information may help contextualize MDRO-positive culture findings in clinical microbiology interpretation and antimicrobial stewardship.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12941-026-00877-0.
Keywords: Diabetes mellitus, Glycated hemoglobin, Multidrug-resistant organisms, Antimicrobial resistance, Clinical microbiology, Vancomycin-resistant enterococcus, Methicillin-resistant Staphylococcus aureus
Background
Diabetes mellitus (DM) is a chronic metabolic disorder associated with increased susceptibility to bacterial infections and infection-related complications. Chronic hyperglycemia can impair host defense mechanisms, including neutrophil chemotaxis, phagocytosis, intracellular killing, cytokine regulation, and cellular immune responses [1, 2]. Diabetic neuropathy, peripheral vascular disease, delayed wound healing, and tissue barrier dysfunction may further increase infection susceptibility [3]. Consequently, urinary tract infections, skin and soft tissue infections, respiratory tract infections, and bloodstream infections are frequently encountered in patients with DM [1, 4–6].
The convergence of diabetes, infection, and antimicrobial resistance has been increasingly recognized as an important clinical problem [7]. Patients with DM may experience recurrent infections, repeated healthcare contact, and prolonged or repeated antimicrobial exposure, all of which may contribute to colonization or infection by multidrug-resistant organisms (MDROs). Previous studies have reported MDRO-related concerns in diabetic foot infections, resistant Gram-negative bacterial infections, and other high-risk diabetic populations [8–12]. However, many studies have focused on specific infection syndromes or individual organisms rather than examining MDRO-positive cultures across multiple routine clinical specimen types.
MDROs are clinically important because they are associated with limited therapeutic options, delayed appropriate antimicrobial therapy, prolonged hospitalization, increased healthcare costs, and adverse outcomes [13–16]. Although MDROs have traditionally been considered a healthcare-associated problem, community-onset MDRO infections are also increasingly recognized, particularly among patients with comorbidities, prior hospitalization, urinary catheterization, or previous antimicrobial exposure [17]. These observations highlight the need to better characterize patient- and specimen-level patterns of MDRO-positive cultures in routine clinical microbiology settings.
The distribution of MDRO phenotypes may differ substantially according to specimen type and anatomical site. Methicillin-resistant Staphylococcus aureus is commonly associated with skin and soft tissue infections and bloodstream infections, whereas resistant Gram-negative bacilli are frequently encountered in respiratory, bloodstream, and healthcare-associated infections [8, 12, 18, 19]. Vancomycin-resistant Enterococcus may also be isolated from urinary and bloodstream specimens, particularly in patients with repeated healthcare exposure or prior antimicrobial treatment. Because resistance patterns vary by region, hospital level, patient population, and specimen source, local clinical microbiology data remain essential for empirical antimicrobial therapy, antimicrobial stewardship, and infection-control planning [18–20].
Glycated hemoglobin (HbA1c) is a widely used marker of long-term glycemic control in patients with DM. Poor glycemic control may impair immune function and create local conditions that favor microbial persistence [1, 2]. However, the relationship between HbA1c and MDRO-positive cultures across routine clinical specimen types, including abscess/pus, urine, sputum, and blood, remains insufficiently characterized. Therefore, this retrospective clinical microbiology study aimed to compare MDRO-positive culture patterns between adults with and without DM, characterize specimen-specific profiles of predefined clinically monitored MDRO phenotypes, and evaluate the association between HbA1c-based glycemic control and MDRO-positive cultures among adults with DM in a regional teaching hospital in southern Taiwan.
Methods
Study design and setting
This retrospective cohort study was reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. This retrospective clinical microbiology study was conducted at Pingtung Christian Hospital, a regional teaching hospital with 676 beds in southern Taiwan. The hospital provides acute care and clinical microbiology laboratory services for patients from Pingtung County and surrounding areas. Clinical microbiology and electronic medical record data were reviewed for adult outpatients and inpatients who had culture-positive bacterial isolates between January 2023 and December 2023. The analytical unit was the eligible, non-duplicate clinical isolate. Patient-level variables, including diabetes status, age, sex, and HbA1c, were linked to the included isolates. All data were extracted from the hospital Healthcare Information System and Electronic Medical Record database. Patient identifiers were removed before analysis.
Patient and isolate selection
Eligible records were culture-positive bacterial isolates obtained from adult patients aged 20 years or older during the study period. The analysis was restricted to isolates recovered from abscess/pus, urine, sputum, or blood specimens and belonging to one of six prespecified clinically relevant bacterial groups: Staphylococcus aureus, Enterococcus species, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Escherichia coli. These organisms were selected because they corresponded to the clinically monitored MDRO phenotypes routinely reported by the hospital clinical microbiology laboratory and infection-control program.
To avoid duplicate counting, only the first eligible non-duplicate isolate from each patient during the study period was included. The first eligible isolate was determined according to the specimen collection date and culture record sequence in the laboratory information system. Repeated culture reports from the same patient during the study period were excluded. Therefore, each patient contributed only one eligible isolate to the final analysis, and no clustering adjustment for multiple isolates per patient was required.
The exclusion criteria were as follows: age younger than 20 years; gestational diabetes; repeated culture reports from the same patient during the study period; specimens other than abscess/pus, urine, sputum, or blood; and bacterial species outside the six prespecified pathogen groups.
Diabetes status was determined from the electronic medical record based on documented physician diagnosis and/or available laboratory evidence consistent with the diagnostic criteria used by the Health Promotion Administration, Ministry of Health and Welfare, Taiwan. When laboratory data were available, diabetes-range values such as HbA1c ≥ 6.5% were considered supportive evidence. Isolates linked to patients meeting these criteria were classified as the DM group, whereas isolates linked to patients without documented diabetes diagnosis or diabetes-range laboratory evidence were classified as the non-DM group. Isolates were linked to patient-level demographic and laboratory variables, including diabetes status, age, sex, and available HbA1c values.
Definitions of MDROs
In this study, MDROs were operationally defined according to the phenotypic resistance categories routinely monitored by the hospital clinical microbiology laboratory and infection-control program. The analysis was restricted to six clinically monitored MDRO phenotypes: methicillin-resistant Staphylococcus aureus (MRSA), vancomycin-resistant Enterococcus (VRE), carbapenem-resistant Klebsiella pneumoniae (CRKP), carbapenem-resistant Escherichia coli (CREC), carbapenem-resistant Acinetobacter baumannii (CRAB), and carbapenem-resistant Pseudomonas aeruginosa (CRPA). This operational definition was selected because these organisms represent the major MDRO phenotypes used for local infection-control surveillance and routine clinical microbiology reporting at the study hospital. Therefore, MDRO classification in this study should not be interpreted as an organism-wide multidrug-resistance definition based on non-susceptibility across all antimicrobial classes.
Antimicrobial susceptibility results were interpreted according to the Clinical and Laboratory Standards Institute (CLSI) M100 breakpoints used by the laboratory during the study period [21]. Carbapenem resistance was defined using organism-specific CLSI M100 criteria. In particular, organism-specific imipenem disk diffusion breakpoints were applied for non-glucose-fermenting Gram-negative bacilli; A. baumannii and P. aeruginosa were not interpreted using a uniform inhibition-zone cutoff. After applying these operational MDRO definitions, eligible isolates were categorized as MDRO-positive or non-MDRO for analysis.
Specimen processing and microbiological identification
All clinical specimens were processed according to the standard operating procedures of the hospital clinical microbiology laboratory. Eligible specimens included abscess/pus, urine, sputum, and blood cultures. Routine culture, organism identification, and antimicrobial susceptibility testing were performed as part of standard clinical care.
Bacterial identification and antimicrobial susceptibility testing were performed using the VITEK-2 Compact automated system (bioMérieux, Marcy-l’Étoile, France). Antimicrobial susceptibility results were interpreted according to CLSI M100 breakpoints used by the laboratory during the study period. Isolates with suspected MDRO phenotypes were confirmed by disk diffusion testing when indicated. Quality control procedures were performed according to the laboratory standard operating protocols.
Clinical and biochemical variables
Demographic variables, including age and sex, were retrieved from electronic medical records. Encounter status at the time of culture collection was classified as inpatient or outpatient and used in an additional sensitivity analysis. Microbiological variables included specimen type, bacterial species, antimicrobial susceptibility results, and MDRO status.
For patients with DM, biochemical data were extracted from available laboratory records. The variables included glycated hemoglobin (HbA1c), total cholesterol, triglycerides, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, and uric acid. Poor glycemic control was defined as HbA1c > 8.0%, whereas HbA1c ≤ 8.0% was classified as relatively well-controlled glycemia. The HbA1c cutoff of >8.0% was selected a priori as an operational definition of poor glycemic control, reflecting clinically meaningful suboptimal long-term glycemic status in patients with diabetes. HbA1c-specific analyses were conducted only for isolates linked to patients with available HbA1c data. Biochemical analyses were performed using available-case analysis; isolates linked to patients without a recorded value for a specific biochemical variable were excluded from the analysis of that variable.
Statistical analysis
Statistical analyses were performed using SPSS version 25.0 (IBM Corp., Armonk, NY, USA). Categorical variables were summarized as frequencies and percentages, and continuous variables were summarized as means with standard deviations. The chi-square test was used to compare categorical variables, including specimen type, sex, age category, DM status, MDRO status, and HbA1c category. Fisher’s exact test was considered when expected cell counts were small. Continuous biochemical variables were compared between MDRO-positive and non-MDRO groups using the independent-samples Student’s t test.
Univariable binary logistic regression was initially used to estimate crude odds ratios and 95% confidence intervals for selected factors associated with MDRO-positive cultures. Among isolates linked to patients with DM and available HbA1c data, multivariable binary logistic regression was performed to estimate the adjusted association between HbA1c category and MDRO-positive cultures. The primary multivariable model included HbA1c category as the main explanatory variable and adjusted for age, sex, and specimen type. An additional sensitivity model further adjusted for inpatient status at the time of culture collection to assess whether the association between HbA1c category and MDRO-positive cultures was influenced by inpatient healthcare exposure. Results were reported as crude odds ratios or adjusted odds ratios with 95% confidence intervals. Because MDRO-positive cultures were not rare, an additional sensitivity analysis was performed using modified Poisson regression with robust standard errors to estimate adjusted risk ratios for the association between HbA1c category and MDRO-positive cultures. The modified Poisson models used the same covariates as the logistic regression models, including age, sex, specimen type, and, in an additional model, inpatient status at the time of culture collection.
To address the potential loss of information caused by dichotomizing HbA1c, sensitivity analyses were performed among isolates linked to patients with DM and available HbA1c data. First, HbA1c was analyzed as a continuous variable, and the association with MDRO-positive cultures was expressed per 1% increase in HbA1c. Second, HbA1c was categorized into clinically relevant strata (≤7.0%, 7.1–8.0%, 8.1–9.0%, and >9.0%) to explore whether the proportion of MDRO-positive cultures increased across worsening glycemic control categories. A test for trend was performed by entering the ordered HbA1c categories as an ordinal variable in the logistic regression model. Sensitivity analyses were adjusted for age, sex, and specimen type where applicable.
A two-sided p value <0.05 was considered statistically significant. To assess potential selection bias related to missing HbA1c data, demographic and microbiological characteristics were compared between isolates linked to patients with DM who had available HbA1c data and those without available HbA1c data. To evaluate whether specimen or organism composition differed according to glycemic-control category, specimen type and bacterial species distributions were compared between isolates linked to patients with HbA1c ≤ 8.0% and those linked to patients with HbA1c > 8.0%. No multiple imputation was performed because HbA1c and biochemical measurements were obtained as part of routine clinical care, and the availability of HbA1c data was likely influenced by clinical testing patterns and healthcare utilization. HbA1c-specific analyses were therefore conducted using complete-case analysis. Because each patient contributed only one eligible isolate to the final analysis, no statistical adjustment for within-patient clustering was performed.
Results
Study population and isolate selection
During the 12-month study period, 10,153 culture-positive bacterial records were initially screened from the clinical microbiology database. After applying the predefined exclusion criteria, 1989 eligible non-duplicate clinical isolates from adult patients were included in the final analysis. Of these, 902 isolates were obtained from patients with diabetes mellitus (DM group), and 1087 isolates were obtained from patients without diabetes mellitus (non-DM group). The isolate selection process and cohort classification are illustrated in Fig. 1.
Fig. 1.
Flow diagram of isolate selection and cohort classification. A total of 10,153 culture-positive bacterial records were screened from the clinical microbiology database between January and December 2023. After exclusion of patients aged <20 years, gestational diabetes, repeated culture reports from the same patient, specimens other than abscess/pus, urine, sputum, and blood, and bacterial species outside the six prespecified pathogen groups, 1989 eligible non-duplicate clinical isolates were included in the final analysis. These isolates were classified into the non-diabetes mellitus group (non-DM; n = 1087), including 190 MDRO-positive and 897 non-MDRO isolates, and the diabetes mellitus group (DM; n = 902), including 176 MDRO-positive and 726 non-MDRO isolates. DM, diabetes mellitus; MDRO, multidrug-resistant organism
The baseline microbiological characteristics of the included isolates are summarized in Table 1. The overall proportion of MDRO-positive isolates was 18.4% (366/1989). The MDRO-positive proportion was 19.5% in the DM group and 17.5% in the non-DM group, with no statistically significant difference between the two groups (p = 0.244).
Table 1.
Baseline microbiological characteristics according to diabetes status
| Characteristic | Total n = 1989 | Non-DM n = 1087 | DM n = 902 | p value |
|---|---|---|---|---|
| Specimen type, n (%) | <0.001 | |||
| Abscess/pus | 352 (17.7) | 236 (21.7) | 116 (12.9) | |
| Urine | 791 (39.8) | 392 (36.1) | 399 (44.2) | |
| Sputum | 363 (18.3) | 196 (18.0) | 167 (18.5) | |
| Blood | 483 (24.3) | 263 (24.2) | 220 (24.4) | |
| Bacterial species, n (%) | ||||
| Escherichia coli | 847 (42.6) | 470 (43.2) | 377 (41.8) | – |
| Staphylococcus aureus | 371 (18.7) | 205 (18.9) | 166 (18.4) | – |
| Klebsiella pneumoniae | 258 (13.0) | 142 (13.1) | 116 (12.9) | – |
| Enterococcus species | 234 (11.8) | 124 (11.4) | 110 (12.2) | – |
| Pseudomonas aeruginosa | 202 (10.2) | 113 (10.4) | 89 (9.9) | – |
| Acinetobacter baumannii | 77 (3.9) | 33 (3.0) | 44 (4.9) | – |
| MDRO status, n (%) | 0.244 | |||
| MDRO-positive | 366 (18.4) | 190 (17.5) | 176 (19.5) | |
| Non-MDRO | 1623 (81.6) | 897 (82.5) | 726 (80.5) | |
DM diabetes mellitus, MDRO multidrug-resistant organism
The dash (–) indicates that no statistical test was performed or reported for that row. Species-level distributions are presented descriptively. Individual species-level comparisons were exploratory and were not emphasized because they were not adjusted for multiple testing
The distribution of specimen types differed significantly between patients with and without DM. Urine was the most frequent specimen source in the DM group, accounting for 399 of 902 isolates (44.2%), whereas urine accounted for 392 of 1087 isolates (36.1%) in the non-DM group. Abscess/pus specimens were less frequent in the DM group than in the non-DM group (12.9% vs. 21.7%). Blood and sputum specimen proportions were similar between the two groups. Escherichia coli was the most frequently isolated bacterial species overall, followed by Staphylococcus aureus, Klebsiella pneumoniae, Enterococcus species, Pseudomonas aeruginosa, and Acinetobacter baumannii.
MDRO-positive cultures by specimen type and diabetes status
The proportion of MDRO-positive cultures varied by specimen source, as shown in Table 2. Among abscess/pus specimens, 40 of 116 isolates from the DM group were MDRO-positive (34.5%), compared with 72 of 236 isolates from the non-DM group (30.5%). This difference was not statistically significant.
Table 2.
MDRO-positive cultures by specimen type and diabetes status
| Specimen type | Non-DM MDRO-positive/total (%) | DM MDRO-positive/total (%) | OR (95% CI) | p value |
|---|---|---|---|---|
| Abscess/pus | 72/236 (30.5) | 40/116 (34.5) | 1.20 (0.75–1.92) | 0.452 |
| Urine | 41/392 (10.5) | 61/399 (15.3) | 1.55 (1.01–2.36) | 0.043 |
| Sputum | 42/196 (21.4) | 40/167 (24.0) | 1.15 (0.71–1.89) | 0.567 |
| Blood | 35/263 (13.3) | 35/220 (15.9) | 1.23 (0.74–2.05) | 0.419 |
CI confidence interval, DM diabetes mellitus, MDRO multidrug-resistant organism, OR odds ratio
Odds ratios were calculated separately for each specimen type using MDRO-positive versus non-MDRO isolates in the DM and non-DM groups. An OR > 1 indicates higher odds of MDRO-positive cultures in the DM group compared with the non-DM group within the same specimen type
In urine specimens, the MDRO-positive proportion was significantly higher in the DM group than in the non-DM group (61/399, 15.3% vs. 41/392, 10.5%; p = 0.043). In specimen-stratified 2 × 2 comparisons, DM was associated with a modest but statistically significant increase in the odds of MDRO-positive urine cultures (OR, 1.55; 95% CI 1.01–2.36). For abscess/pus, sputum, and blood specimens, the ORs were above 1.0 but the corresponding comparisons were not statistically significant. Overall, the association between DM and MDRO-positive cultures was most evident in urine specimens.
Specimen-specific distribution of MDRO phenotypes
The distribution of MDRO phenotypes differed markedly according to specimen type. The detailed phenotype distribution is presented in Table 3 and visually summarized in Fig. 2.
Table 3.
Specimen-specific distribution of MDRO phenotypes
| Specimen type | Total MDROs | MRSA | VRE | CRKP | CRAB | CRPA | CREC |
|---|---|---|---|---|---|---|---|
| Abscess/pus | 112 | 102 | 6 | 2 | 0 | 2 | 0 |
| Urine | 102 | 9 | 76 | 2 | 3 | 9 | 3 |
| Sputum | 82 | 46 | 0 | 3 | 18 | 14 | 1 |
| Blood | 70 | 60 | 7 | 2 | 1 | 0 | 0 |
MRSA methicillin-resistant Staphylococcus aureus, VRE vancomycin-resistant Enterococcus, CRKP carbapenem-resistant Klebsiella pneumoniae, CRAB carbapenem-resistant Acinetobacter baumannii, CRPA carbapenem-resistant Pseudomonas aeruginosa, CREC carbapenem-resistant Escherichia coli, MDRO multidrug-resistant organism
The overall association between specimen type and MDRO phenotype was statistically significant (chi-square test, p < 0.001)
Fig. 2.
Specimen-specific distribution of multidrug-resistant organism phenotypes. Stacked bar chart showing the distribution of MDRO phenotypes among abscess/pus, urine, sputum, and blood specimens. MRSA predominated in abscess/pus and blood specimens, VRE predominated in urine specimens, and CRAB and CRPA were relatively enriched in sputum specimens. The total numbers of MDRO-positive isolates were 112 in abscess/pus specimens, 102 in urine specimens, 82 in sputum specimens, and 70 in blood specimens. CRAB, carbapenem-resistant Acinetobacter baumannii; CREC, carbapenem-resistant Escherichia coli; CRKP, carbapenem-resistant Klebsiella pneumoniae; CRPA, carbapenem-resistant Pseudomonas aeruginosa; MDRO, multidrug-resistant organism; MRSA, methicillin-resistant Staphylococcus aureus; VRE, vancomycin-resistant Enterococcus
In abscess/pus specimens, MRSA was the dominant MDRO phenotype, accounting for 102 of 112 MDRO-positive abscess/pus isolates. MRSA was also the predominant MDRO phenotype in blood specimens, accounting for 60 of 70 MDRO-positive blood isolates.
Urine specimens showed a distinct MDRO pattern. Although E. coli was the most common organism isolated from urine overall, VRE was the predominant MDRO phenotype in urine, accounting for 76 of 102 urinary MDRO-positive isolates.
In sputum specimens, MRSA was the most frequent MDRO phenotype by absolute count; however, CRAB and CRPA were more prominently represented in sputum than in other specimen types. The association between specimen type and MDRO phenotype was statistically significant.
Glycemic control, biochemical markers, and MDRO-positive cultures in patients with diabetes
Among the 902 isolates linked to patients with DM, 615 were linked to patients with available HbA1c data. The relationship between HbA1c category, biochemical markers, and MDRO status is shown in Table 4.
Table 4.
Glycemic control and biochemical markers according to MDRO status in patients with diabetes
| Variable | Non-MDRO | MDRO-positive | Crude OR (95% CI) | Adjusted OR (95% CI) | p value |
|---|---|---|---|---|---|
| HbA1c category, n (%) | <0.001 | ||||
| HbA1c ≤ 8.0% | 343/396 (86.6) | 53/396 (13.4) | Reference | Reference | – |
| HbA1c > 8.0% | 148/219 (67.6) | 71/219 (32.4) | 3.10 (2.07–4.65) | 2.82 (1.84–4.32) | <0.001 |
| Biochemical markers, n; mean ± SD | |||||
| Total cholesterol (mg/dL) | n = 237; 156.7 ± 41.8 | n = 37; 149.8 ± 49.3 | – | – | 0.373 |
| HDL cholesterol (mg/dL) | n = 194; 40.5 ± 14.2 | n = 31; 40.8 ± 15.4 | – | – | 0.371 |
| LDL cholesterol (mg/dL) | n = 216; 90.8 ± 35.6 | n = 35; 93.2 ± 34.5 | – | – | 0.782 |
| Triglycerides (mg/dL) | n = 299; 142.3 ± 81.4 | n = 36; 139.4 ± 86.0 | – | – | 0.502 |
| Uric acid (mg/dL) | n = 218; 7.0 ± 3.1 | n = 40; 6.2 ± 2.2 | – | – | 0.081 |
CI confidence interval, HbA1c glycated hemoglobin, HDL high-density lipoprotein, LDL low-density lipoprotein, MDRO multidrug-resistant organism, OR odds ratio, SD standard deviation
HbA1c analysis included patients with available HbA1c data. The crude odds ratio for HbA1c category was calculated using MDRO-positive versus non-MDRO isolates among patients with available HbA1c data. The adjusted odds ratio was estimated using multivariable logistic regression adjusted for age, sex, and specimen type. “Reference” indicates the reference category. The dash (–) indicates not applicable. Biochemical marker analyses were performed using available data for each variable; therefore, sample sizes differ across markers
Among isolates linked to patients with HbA1c ≤ 8.0%, 53 of 396 isolates (13.4%) were MDRO-positive. In contrast, among isolates linked to patients with HbA1c > 8.0%, 71 of 219 isolates (32.4%) were MDRO-positive. Poor glycemic control was significantly associated with MDRO-positive cultures (p < 0.001). Based on the 2 × 2 comparison of MDRO-positive versus non-MDRO isolates, HbA1c > 8.0% was associated with substantially increased odds of MDRO-positive cultures compared with HbA1c ≤ 8.0% (crude OR, 3.10; 95% CI 2.07–4.65). In multivariable logistic regression adjusted for age, sex, and specimen type, HbA1c > 8.0% remained significantly associated with increased odds of MDRO-positive cultures among isolates linked to patients with DM (adjusted OR, 2.82; 95% CI 1.84–4.32; p < 0.001).
In sensitivity analysis, HbA1c analyzed as a continuous variable showed that each 1% increase in HbA1c was associated with increased odds of MDRO-positive cultures after adjustment for age, sex, and specimen type (adjusted OR, 1.22; 95% CI 1.10–1.36; p < 0.001). When HbA1c was categorized into multiple clinically relevant strata, the proportion of MDRO-positive cultures increased progressively across worsening glycemic control categories, supporting a dose–response pattern (p for trend <0.001; Supplementary Table S1).
In an additional sensitivity model further adjusted for inpatient status at the time of culture collection, HbA1c > 8.0% remained significantly associated with increased odds of MDRO-positive cultures among isolates linked to patients with DM (adjusted OR, 2.55; 95% CI 1.63–3.98; p < 0.001; Supplementary Table S2).
To assess potential selection bias related to missing HbA1c data, demographic and microbiological characteristics were compared between isolates with and without available HbA1c measurements. These comparisons are presented in Supplementary Table S3. Among the 902 isolates linked to patients with DM, 615 had available HbA1c data and 287 did not. The MDRO-positive proportion was similar between isolates with available HbA1c data and those without available HbA1c data (124/615, 20.2% vs. 52/287, 18.1%; p = 0.471). Age distribution, sex distribution, specimen type, and bacterial species distribution were also broadly comparable between the two groups. However, isolates with available HbA1c data were more frequently obtained during inpatient encounters than those without available HbA1c data (330/615, 53.7% vs. 125/287, 43.6%; p = 0.006), suggesting that HbA1c availability may partly reflect clinical testing patterns or healthcare utilization.
Because MDRO-positive cultures were not rare, modified Poisson regression with robust standard errors was performed as an additional sensitivity analysis to estimate adjusted risk ratios. In this analysis, HbA1c > 8.0% remained associated with a higher MDRO-positive proportion after adjustment for age, sex, and specimen type (adjusted RR, 2.18; 95% CI 1.57–3.03; p < 0.001). After additional adjustment for inpatient status at the time of culture collection, the association remained statistically significant (adjusted RR, 2.05; 95% CI 1.46–2.88; p < 0.001; Supplementary Table S4). The adjusted risk ratios were smaller in magnitude than the corresponding adjusted odds ratios, but the direction and statistical significance of the association were consistent.
Because MDRO-positive proportions differed by specimen type, we further compared specimen type and bacterial species distributions between isolates linked to patients with HbA1c ≤ 8.0% and those linked to patients with HbA1c > 8.0%. Specimen type and bacterial species distributions were broadly comparable between the two HbA1c categories (p = 0.848 and p = 0.983, respectively; Supplementary Table S5), whereas the MDRO-positive proportion differed significantly between the two HbA1c categories (p < 0.001). Because specimen type may influence MDRO-positive proportions, specimen type was included as an adjustment covariate in the primary multivariable model.
Discussion
This retrospective clinical microbiology study analyzed 1989 culture-positive clinical isolates from a regional teaching hospital in southern Taiwan and identified several clinically relevant findings. First, the overall proportion of MDRO-positive isolates was not significantly different between patients with and without diabetes mellitus (DM), suggesting that diabetes status alone may not fully explain the occurrence of MDRO-positive cultures. Second, the association between DM and MDRO-positive cultures was most evident in urine specimens; compared with non-DM, DM was associated with a modest but statistically significant increase in the odds of MDRO-positive urine cultures. Third, among patients with DM, poor glycemic control, defined as HbA1c > 8.0%, was associated with increased odds of MDRO-positive cultures. Finally, MDRO phenotypes showed clear specimen-specific patterns: MRSA predominated in abscess/pus and blood specimens, VRE predominated in urine specimens, and CRAB and CRPA were relatively enriched in sputum specimens.
A notable finding of this study is that the overall MDRO-positive proportion was similar between the DM and non-DM groups. This result suggests that DM should not be considered a uniform risk factor for all MDRO-positive cultures across all specimen types. Instead, the likelihood of MDRO-positive cultures may be influenced by clinical context, specimen source, metabolic control, healthcare exposure, and local microbial ecology. This interpretation is consistent with the concept that antimicrobial resistance in patients with diabetes is multifactorial rather than solely attributable to the presence of diabetes [2, 7]. The absence of a significant overall difference between DM and non-DM groups also suggests the value of examining MDRO-positive proportions according to specimen source and glycemic-control status. However, these stratified findings should be interpreted descriptively, because observed differences may still be influenced by organism distribution, healthcare exposure, and patient characteristics.
The urinary tract was the specimen source in which the association between DM and MDRO-positive cultures was most evident. In urine specimens, patients with DM had a higher MDRO-positive proportion than patients without DM, although the magnitude of association was modest. This finding is biologically plausible. Patients with DM are prone to urinary tract infections because of glucosuria, impaired neutrophil function, autonomic neuropathy-related bladder dysfunction, urinary stasis, and incomplete bladder emptying [1, 5, 6]. Previous studies have shown that urinary tract infections occur more frequently and recur more often in patients with type 2 DM than in individuals without DM [5, 6, 22]. In addition, diabetic urinary tract infections may involve more complicated clinical courses and broader antimicrobial exposure, both of which can increase the likelihood of isolating resistant organisms [23].
An important microbiological finding was that VRE was the predominant MDRO phenotype in urine specimens, despite E. coli being the most common urinary isolate overall. This discrepancy is clinically relevant. In routine practice, empirical therapy for urinary tract infection is often designed around common Gram-negative uropathogens, especially E. coli. However, the predominance of VRE among urinary MDRO-positive isolates suggests that, in patients with diabetes who have urine cultures positive for resistant organisms, enterococcal resistance should be carefully considered. This does not imply that empirical anti-VRE therapy is warranted for all patients with diabetes with suspected urinary tract infection. Rather, it supports the need for specimen-specific interpretation of culture results, careful review of prior microbiology, and timely antimicrobial susceptibility reporting. For clinical microbiology laboratories, highlighting local urinary MDRO patterns may help clinicians interpret preliminary culture results and adjust therapy earlier when VRE is suspected.
The predominance of VRE among urinary MDRO-positive isolates should be interpreted in the context of the operational MDRO definition used in this study. Although E. coli was the most common urinary isolate overall, ESBL-producing Enterobacterales without carbapenem resistance were not included in the predefined MDRO phenotypes analyzed here. Therefore, the high proportion of VRE among urinary MDRO-positive isolates does not indicate that VRE was the most common urinary pathogen overall, but rather that VRE predominated within the monitored urinary MDRO phenotypes. Several local factors may have contributed to this finding, including endemic VRE circulation, repeated healthcare exposure, urinary catheterization, and antimicrobial selective pressure from broad-spectrum agents. However, this retrospective study did not include ward-level temporal clustering, institutional antibiotic consumption data, or molecular typing. Therefore, we could not determine whether the observed urinary VRE pattern reflected a localized outbreak, clonal spread, or endemic institutional circulation.
Another finding of this study was the association between poor glycemic control and MDRO-positive cultures among isolates linked to patients with available HbA1c data. This association remained significant after adjustment for age, sex, and specimen type and was supported by sensitivity analyses using continuous HbA1c, multiple HbA1c strata, additional adjustment for inpatient status, and modified Poisson regression estimating adjusted risk ratios. Chronic hyperglycemia can impair host defense and may promote bacterial persistence, providing biological plausibility for the observed association [1, 2]. However, patients with poor glycemic control may differ from those with better glycemic control in healthcare contact, chronic disease severity, prior antimicrobial exposure, device use, and comorbidity burden. Therefore, the HbA1c-related findings should be interpreted as associations rather than proof of causality or fully independent predictive effects.
The specimen-specific distribution of MDRO phenotypes has practical implications for clinical microbiology interpretation and antimicrobial stewardship. MRSA predominated in abscess/pus specimens, consistent with the established role of S. aureus in skin and soft tissue infections [18]. In patients with DM, neuropathy, peripheral vascular disease, impaired skin barrier function, delayed wound healing, and repeated antimicrobial exposure may increase the complexity of wound infections [3, 8–10]. Previous studies have reported a substantial burden of resistant organisms in diabetic foot and wound infections, although the relative contribution of MRSA and resistant Gram-negative organisms varies by region and patient population [9, 11, 24, 25]. MRSA was also the dominant MDRO phenotype in blood specimens. Because bloodstream infections caused by resistant organisms are associated with delayed appropriate therapy, increased healthcare burden, and adverse outcomes [13–16], local bloodstream MDRO patterns should be considered when interpreting blood culture results, particularly in patients with severe infection, catheter-related infection, prior MRSA colonization, or healthcare exposure [14, 26, 27].
In sputum specimens, MRSA was the most frequent MDRO phenotype by absolute count, but CRAB and CRPA were relatively more prominent than in other specimen types. This pattern is consistent with the recognized importance of non-fermenting Gram-negative bacilli in respiratory and healthcare-associated infections [19]. A. baumannii and P. aeruginosa are well adapted to hospital environments and may be isolated from patients with prior antimicrobial exposure, prolonged hospitalization, mechanical ventilation, or impaired host defense. Because the present study did not include detailed clinical variables such as ventilator use, ICU admission, or prior antibiotic history, respiratory isolates should be interpreted cautiously, with attention to the distinction between colonization and true infection. Nevertheless, the relative enrichment of CRAB and CRPA in sputum specimens supports the value of specimen-specific MDRO surveillance rather than reliance on pooled institutional resistance summaries alone.
From a clinical microbiology and stewardship perspective, these findings suggest that antimicrobial decision-making should incorporate specimen type, prior culture history, local resistance patterns, glycemic-control information, and final susceptibility results. Urine cultures from patients with DM may warrant careful attention to VRE, abscess/pus and blood cultures may require awareness of local MRSA epidemiology, and sputum cultures may require attention to carbapenem-resistant non-fermenting Gram-negative bacilli. These observations should not be interpreted as prescriptive empirical treatment recommendations; rather, they highlight how routine culture and susceptibility data can support specimen-specific interpretation and communication between clinical microbiology laboratories, infection-control teams, and frontline clinicians, particularly in regional hospitals where resources for molecular surveillance may be limited [20].
Several limitations should be acknowledged. First, this was a single-center retrospective study, and the findings may not be generalizable to other hospitals, regions, or healthcare systems with different patient populations and antimicrobial resistance patterns. In addition, this study used an operational MDRO definition restricted to six clinically monitored resistance phenotypes routinely reported by the study hospital. Therefore, the findings may not be generalized to other MDRO definitions, organism groups, or institutions with different surveillance priorities and resistance epidemiology. Second, the analysis was based on culture-positive isolates rather than prospectively adjudicated clinical infections. Therefore, some isolates may represent colonization rather than true infection, particularly in specimens such as sputum and urine. For this reason, the term MDRO-positive cultures is more appropriate than MDRO infections when interpreting the findings. Third, the study lacked detailed clinical variables such as prior antimicrobial exposure, hospitalization history, ICU admission, catheter use, urinary tract instrumentation, wound severity, duration of diabetes, comorbidity burden, and indicators of chronic disease severity. These factors are known or plausible contributors to MDRO-positive cultures and may differ between patients with poor and relatively well-controlled glycemia. Therefore, residual confounding by healthcare exposure, disease severity, device use, and antimicrobial selective pressure may remain despite adjustment for age, sex, specimen type, and inpatient status in sensitivity analysis. Fourth, molecular testing for resistance determinants, such as mecA for MRSA or vanA/vanB for VRE, was not performed. As a result, the study could not evaluate clonal transmission, resistance gene distribution, or molecular epidemiology. In addition, ward-level temporal clustering, institutional antibiotic consumption data, and molecular typing were not available; therefore, the predominance of urinary VRE could not be attributed to a localized outbreak, clonal spread, endemic institutional circulation, or antimicrobial selective pressure with certainty. Finally, HbA1c and biochemical variables were available only when clinically measured; therefore, complete-case analyses may be affected by selection bias. Although the MDRO-positive proportion was similar between isolates with and without available HbA1c data, isolates with available HbA1c data were more frequently obtained during inpatient encounters. This finding suggests that HbA1c availability may be influenced by healthcare utilization or clinical testing patterns. Therefore, HbA1c-related findings should be interpreted as associations rather than causal or fully independent predictive effects.
Future studies should validate these findings in multicenter cohorts and incorporate more detailed clinical variables, including prior antibiotic exposure, hospitalization, device use, infection severity, and clinical outcomes. Prospective studies could help distinguish colonization from true infection and determine whether incorporating HbA1c information into clinical microbiology interpretation improves empirical antimicrobial selection. Molecular typing and resistance gene analysis would also help clarify whether the observed specimen-specific MDRO patterns reflect local transmission, endogenous colonization, or independent acquisition events.
Conclusions
This retrospective clinical microbiology study showed that diabetes status alone was not associated with a significantly higher overall proportion of MDRO-positive isolates. However, urine specimens from patients with DM showed a higher MDRO-positive proportion, with VRE as the predominant urinary MDRO phenotype. Poor glycemic control, defined as HbA1c > 8.0%, was associated with increased odds of MDRO-positive cultures among patients with DM. These findings support specimen-specific MDRO surveillance and the integration of glycemic control information into clinical microbiology interpretation and antimicrobial stewardship strategies.
Supplementary Information
Acknowledgements
Not applicable.
Abbreviations
- AST
Antimicrobial susceptibility testing
- ATCC
American Type Culture Collection
- CI
Confidence interval
- CLSI
Clinical and Laboratory Standards Institute
- CRAB
Carbapenem-resistant Acinetobacter baumannii
- CREC
Carbapenem-resistant Escherichia coli
- CRKP
Carbapenem-resistant Klebsiella pneumoniae
- CRPA
Carbapenem-resistant Pseudomonas aeruginosa
- DM
Diabetes mellitus
- EMR
Electronic medical record
- HbA1c
Glycated hemoglobin
- HDL
High-density lipoprotein
- HIS
Healthcare Information System
- ICU
Intensive care unit
- IRB
Institutional Review Board
- LDL
Low-density lipoprotein
- MDRO
Multidrug-resistant organism
- MRSA
Methicillin-resistant Staphylococcus aureus
- OR
Odds ratio
- SD
Standard deviation
- STROBE
Strengthening the Reporting of Observational Studies in Epidemiology
- VRE
Vancomycin-resistant Enterococcus
Author contributions
JSC and CSL conceived and designed the study. MYC collected and curated the clinical microbiology and laboratory data. JSC, MYC, SRH, and CSL contributed to data interpretation. JSC and CSL drafted the manuscript. SRH and CSL critically revised the manuscript for important intellectual content. CSL supervised the study and obtained funding. All authors reviewed and approved the final manuscript and agree to be accountable for all aspects of the work.
Funding
This work was supported by the National Science and Technology Council, Taiwan (grant no. 109-2222-E-126-002) and I-Shou University (grant no. ISU-114-IUC-04). The funder had no role in study design, data collection, analysis, interpretation, manuscript preparation, or the decision to submit the article for publication.
Data availability
No datasets were generated or analysed during the current study. The datasets used and/or analyzed during the current study are not publicly available due to patient privacy and institutional restrictions but are available from the corresponding author on reasonable request and with permission from the study institution.
Use of AI-assisted technologies
During the preparation of this work the authors used AI-assisted technologies solely to improve readability and language editing. After using these tools, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication. Use of AI-assisted technologies during manuscript preparation is described in the “Declarations” section.
Declarations
Ethics approval and consent to participate
This study was approved by the Institutional Review Board of Pingtung Christian Hospital, Taiwan (IRB No. IRB734A), and was conducted in accordance with the ethical principles of the Declaration of Helsinki. Because this retrospective study used de-identified clinical and laboratory data, the requirement for informed consent was waived by the Institutional Review Board.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Casqueiro J, Casqueiro J, Alves C. Infections in patients with diabetes mellitus: a review of pathogenesis. Indian J Endocrinol Metab. 2012;16(Suppl 1):S27–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Akash MSH, Rehman K, Fiayyaz F, Sabir S, Khurshid M. Diabetes-associated infections: development of antimicrobial resistance and possible treatment strategies. Arch Microbiol. 2020;202:953–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Tesfaye S, Selvarajah D. Advances in the epidemiology, pathogenesis and management of diabetic peripheral neuropathy. Diabetes Metab Res Rev. 2012;28(Suppl 1):8–14. [DOI] [PubMed] [Google Scholar]
- 4.Calliari LE, Almeida FJ, Noronha RM. Infections in children with diabetes. J Pediatr. 2020;96:39–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hirji I, Guo Z, Andersson SW, Hammar N, Gomez-Caminero A. Incidence of urinary tract infection among patients with type 2 diabetes in the UK General Practice Research Database. J Diabetes Complications. 2012;26:513–6. [DOI] [PubMed] [Google Scholar]
- 6.Fu AZ, Iglay K, Qiu Y, Engel S, Shankar R, Brodovicz K. Risk characterization for urinary tract infections in subjects with newly diagnosed type 2 diabetes. J Diabetes Complicat. 2014;28:805–10. [DOI] [PubMed] [Google Scholar]
- 7.Sharma S, Chakravarti VK, Rani S, Parveen M, Yadav K, Chandra E, et al. The triple threat in diabetes: interconnection of multidrug resistance and systemic complications. Vasc Endovasc Rev. 2025;8(1s):119–28. [Google Scholar]
- 8.Mohammad AM, Ahmad J. Clinico-microbiological study and antimicrobial drug resistance profile of diabetic foot infections in North India. Foot. 2011;21:6–14. [DOI] [PubMed] [Google Scholar]
- 9.Ji X, Jin P, Chu Y, Feng S, Wang P. Clinical characteristics and risk factors of diabetic foot ulcer with multidrug-resistant organism infection. Int J Low Extrem Wounds. 2014;13:64–71. [DOI] [PubMed] [Google Scholar]
- 10.Dai J, Jiang C, Chen H, Chai Y. Assessment of the risk factors of multidrug-resistant organism infection in adults with type 1 or type 2 diabetes and diabetic foot ulcer. Can J Diabetes. 2020;44:342–9. [DOI] [PubMed] [Google Scholar]
- 11.Huang J, Huang D, Tian D. Retrospective study on the distribution of pathogens, antibiotic resistance patterns, and risk factors of multidrug-resistant organism infections in patients with type 2 diabetic foot ulcers. Medicine (Baltimore). 2026;105(5):e45888. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Menon A, Bhat S, Iqbal N, Kandasamy R. Clinical and epidemiological profile of multidrug-resistant Klebsiella pneumoniae infections in patients with type 2 diabetes mellitus. Cureus. 2026;18(1):e101724. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Oxman D, Lohr K, Gupta E, Madara J, Len E, Hseih J. Incidence of multidrug resistant infections in emergency department patients with suspected sepsis. Am J Med Sci. 2020;360:650–5. [DOI] [PubMed] [Google Scholar]
- 14.Stewardson AJ, Allignol A, Beyersmann J, Graves N, Schumacher M, Meyer R, et al. The health and economic burden of bloodstream infections caused by antimicrobial-susceptible and non-susceptible Enterobacteriaceae and Staphylococcus aureus in European hospitals, 2010 and 2011: a multicentre retrospective cohort study. Euro Surveill. 2016;21:30319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Adrie C, Garrouste-Orgeas M, Ibn Essaied W, Schwebel C, Darmon M, Mourvillier B, et al. Attributable mortality of ICU-acquired bloodstream infections: impact of the source, causative micro-organism, resistance profile and antimicrobial therapy. J Infect. 2017;74:131–41. [DOI] [PubMed] [Google Scholar]
- 16.Shi N, Kang J, Wang S, Song Y, Yin D, Li X, et al. Bacteriological profile and antimicrobial susceptibility patterns of Gram-negative bloodstream infection and risk factors associated with mortality and drug resistance: a retrospective study from Shanxi. China Infect Drug Resist. 2022;15:3561–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Matzaras R, Biros D, Sakkou SF, Lymperatou D, Filippas-Ntekouan S, Prokopidou A, et al. Risk factors associated with community-onset infections due to multidrug-resistant organisms. Antibiotics (Basel). 2025;14(11):1073. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Odell CA. Community-associated methicillin-resistant Staphylococcus aureus skin infections. Curr Opin Pediatr. 2010;22:273–7. [DOI] [PubMed] [Google Scholar]
- 19.Jiang W, Li L, Wen S, Song Y, Yu L, Tan B. Gram-negative multidrug-resistant organisms were dominant in neurorehabilitation ward patients in a general hospital in southwest China. Sci Rep. 2022;12:11087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Chen YP, Liang CC, Chang R, Kuo CM, Hung CH, Liao TN, et al. Detection and colonization of multidrug resistant organisms in a regional teaching hospital of Taiwan. Int J Environ Res Public Health. 2019;16:1104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Clinical and Laboratory Standards Institute. Performance standards for antimicrobial susceptibility testing. CLSI supplement M100. 33rd ed. Wayne: Clinical and Laboratory Standards Institute; 2023. [Google Scholar]
- 22.Lye WC, Chan RKT, Lee EJC, Kumarasinghe G. Urinary tract infections in patients with diabetes mellitus. J Infect. 1992;24:169–74. [DOI] [PubMed] [Google Scholar]
- 23.Nath T, Das SK, Hazra S. Pattern of uropathogens and antibiotic sensitivity in diabetes patients attending to out-patient department and diabetes clinic of a teaching hospital: a cross-sectional study. J Fam Med Prim Care. 2021;10:3638. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Trivedi U, Parameswaran S, Armstrong A, Burgueno-Vega D, Griswold J, Dissanaike S, et al. Prevalence of multiple antibiotic resistant infections in diabetic versus nondiabetic wounds. J Pathog. 2014;2014:173053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Prashanth SP, Vishwanath M, Murgesh JV. A comparative study of proportion of resistance to multiple antibiotics in infected wounds of diabetic versus non-diabetic patients in a tertiary care hospital. Natl J Physiol Pharm Pharmacol. 2022;12:684–8. [Google Scholar]
- 26.Thaden JT, Li Y, Ruffin F, Maskarinec SA, Hill-Rorie JM, Wanda LC, et al. Increased costs with multidrug-resistant Gram-negative bloodstream infections are primarily due to patients with hospital-acquired infections. Antimicrob Agents Chemother. 2017;61:e1709–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Grundmann H, Schouls LM, Aanensen DM, Pluister GN, Tami A, Chlebowicz M, et al. The dynamic changes of dominant clones of Staphylococcus aureus causing bloodstream infections in the European region: results of a second structured survey. Euro Surveill. 2014;19:20987. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
No datasets were generated or analysed during the current study. The datasets used and/or analyzed during the current study are not publicly available due to patient privacy and institutional restrictions but are available from the corresponding author on reasonable request and with permission from the study institution.
During the preparation of this work the authors used AI-assisted technologies solely to improve readability and language editing. After using these tools, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication. Use of AI-assisted technologies during manuscript preparation is described in the “Declarations” section.