Abstract
Introduction:
Fungal necrotizing soft tissue infections (NSTIs) are often diagnosed late due to the low likelihood, resulting in poor patient outcomes. Traditionally NSTI are treated with emergent surgical debridement followed by serial interventions. Adjunctive antibiotics are aimed at common bacterial infections along with toxin reduction; however, fungal treatment is often delayed. This study aimed to identify predictors, treatment patterns, and outcomes in fungal NSTIs.
Methods:
A retrospective review of patients with NSTIs (n= 634) from a single tertiary academic center was conducted from 2011 to 2023, with patients stratified by fungal presence in wound cultures. Primary outcomes were mortality along with the number of serial debridement. Secondary outcomes were length of stay, intensive care requirements, acute kidney injury, and 30-days readmission. Logistic regression was utilized to identify predictors of fungal infection.
Results:
Of 634 patients, 110 (17.4%) had fungal-positive wound cultures, with Candida albicans being the most common (47.3%). Patients with positive fungal cultures more frequently had Type 2 diabetes mellitus (74.5% vs. 61.6%, p = 0.014). Patients with positive fungal wound cultures were more likely to require ≥ 3 operative debridement (p < 0.001), had a longer length of stay (19 vs. 12 d, p < 0.001), require mechanical ventilation (p < 0.001), and had greater 90-days mortality (20% vs. 12%, p = 0.03). Female gender (OR: 1.73, 95% CI: [1.09–2.76], p= 0.021), diabetes mellitus (OR: 1.75, 95% CI: [1.07–2.93], p = 0.030), vasopressor use (OR: 1.71, 95% CI: [1.04–2.81], p = 0.034), and VRE infection (OR: 12.2, 95% CI: [4.16–41.3], p < 0.001) were predictive of fungal infection.
Conclusion:
Fungal NSTIs are associated with worse patient outcomes. Early empirical antifungal therapy may be considered in high-risk patients, particularly those with diabetes, as part of comprehensive NSTI management.
Keywords: critical care, fungal infection, NSTI, predictors, surgical outcomes
Introduction
Necrotizing soft tissue infections (NSTIs) are rapidly progressive, life- or limb-threatening infections characterized by the development of necrosis of the skin, subcutaneous tissue, and superficial fascia.1 Emergent surgical debridement and broad-spectrum antimicrobials are critical to halting disease progression.2 Despite advances in early intervention, inhospital mortality remains as high as 20% for the past three decades.2-4
Fungal NSTIs, though rare, carry significantly greater mortality rates, up to three times that of bacterial NSTIs, yet remain poorly characterized in the literature.5-7 Early intervention is essential in NSTI management; however, fungal infections complicate this approach.2 Antifungals are often excluded from initial empirical treatment regimens, and fungal infections are difficult to diagnose due to overlapping clinical features with bacterial infections, limited culture sensitivity, and delayed time to positivity.7,8 Critically ill or immunocompromised patients, who are more susceptible to fungal NSTIs, may present atypically, further hindering timely diagnosis.9 Notably, misdiagnosis and underdiagnosis of fungal infections lead to an underestimation of their economic burden, which already exceeded $11.5 billion in 2019, comprising direct medical costs, productivity losses, and premature deaths.10,11
Despite this, there are currently no guidelines for empirical antifungal initiation in NSTIs, even for high-risk populations. Thus, we sought to characterize the outcomes, treatment patterns, and predictors of NSTIs with positive fungal cultures. A clearer understanding of these infections may enable earlier diagnosis, guide empirical treatment strategies, and ultimately improve clinical outcomes.
Methods
Patients and study design
Retrospective analysis of institutional data from a tertiary academic center for patients diagnosed with NSTIs (n = 634) was performed between October 2011 and October 2023 in an institutional review board (IRB)-approved REDCap database under a waiver of consent. IRB approval #2032H0132 through the Ohio State University Biomedical Sciences IRB. The included patients were aged ≥18 years and underwent surgical debridement for a diagnosis of NSTI. Exclusion criteria were incarcerated, pregnant and <18 years of age patients. Transferred patients were excluded if we were unable to obtain necessary admission data points through our electronic medical record (EMR) system. Diagnosis of NSTI was on the basis of International Classification of Diseases (ICD) codes ICD-9 728.86, and ICD-10 M72.6. An in-depth chart review was performed to confirm the diagnosis, and exclude patients without infection or necrosis of subcutaneous tissue/superficial fascia.12 This cohort was split into two groups on the basis of the presence of fungi in wound cultures. Positive tissue cultures were defined as quantitative wound cultures with > 100,000 colony-forming units (cfu)/milliliter.
Data collection and outcome measures
We collected baseline demographic and comorbidity data, including Charlson Comorbidity Index, as well as socioeconomic data, such as insurance payor type, primary care physician (PCP) status, and National Area Deprivation Index (ADI). ADI was evaluated in quartiles, with greater scores corresponding to low socioeconomic status.13 We accounted for whether patients were admitted through the emergency department or from a medical facility (i.e., outside hospital, skilled nursing facility). Once admitted, we collected clinical/surgical presentation including primary site of NSTI, baseline lab values, acute physiology and chronic health evaluation (APACHE) II score, sequential organ failure assessment (SOFA) score, blood culture results, wound culture results, and primary outcomes (in-patient mortality and serial debridement) and secondary outcomes such as length of stay (days), length of intensive care unit (ICU) stay (days), mechanical respiratory support, ventilator support days, vasopressor requirements, need for amputation, acute kidney injury (AKI) (at presentation or developed during admission), need for renal replacement therapy (RRT) due to AKI, discharge destination (home/home health, medical facility, other), 30-days readmission, 90-day mortality). In the fungus-positive cohort, data were collected on the type of fungus identified in wound cultures, timing of positive fungal culture results, time to detection of fungal growth (in days), administration of empirical antifungal treatment, specific antifungal therapies used, interval to initiation of antifungal therapy (in days), discharge prior to positive fungal culture, and mortality occurring before a positive fungal culture.
Statistical analysis
All statistical analyses were performed using R Statistical Software version 4.4.3 (R Core Team, 2025), with p-values of less than 0.05 considered statistically significant. Missing data were imputed using the missForest package in R, which utilizes a non-parametric imputation method via random forest models to predict missing values. This approach is capable of handling both continuous and categorical variables without requiring distributional assumptions. For each variable with missing data, a random forest model is trained using the available observed data and is then used to predict the missing values. These predictions replace the initial estimates, and the process continues until convergence. Missing data affected less than 2% of the total values and occurred in the following variables: Race, insurance status, presence of a PCP, ADI, source of hospital admission, body mass index (BMI), APACHE score, SOFA score, pre-operative white blood cell (WBC) count, pre-operative sodium, and pre-operative hemoglobin levels. Imputation models incorporated baseline demographics, comorbid conditions, and the anatomical location of NSTI to inform predictions.
Normally distributed values were compared using Student’s t-test, and non-parametric continuous variables were compared using Mann–Whitney U test. Chi-squared or Fisher Exact tests were utilized to compare categorical variables, depending on the sample size. A uni-variable logistic regression analysis was conducted to identify characteristics associated with a positive fungal wound culture. Variables evaluated included demographics, socioeconomics, comorbidities, admission source, NSTI site, meets sepsis criteria, SOFA score, vasopressor requirement, AKI at presentation, blood and wound culture results, and pre-operative white count, hemoglobin, sodium, and creatinine. Variables with a p value less than 0.2 in the uni-variable analysis or deemed clinically significant were then included in a multi-variable regression model. This threshold was intentionally liberal to reduce the risk of excluding potentially important confounders or predictors early on. Multicollinearity was evaluated using the variance inflation factor, with variables with values > 5 excluded from the multi-variable model. Adjusted odds ratios (OR) were calculated to model a positive fungal wound culture for the fungal-positive cohort. No backward or forward selection was conducted for the models presented.
Results
Baseline characteristics: Fungal versus non-fungal
Of 634 patients, 110 (17.4%) had fungal-positive wound cultures. Patients with positive fungal cultures were more likely to have Type 2 diabetes mellitus (74.5% vs. 61.6%, p = 0.010). No other baseline characteristics were found to be significantly different between those with positive cultures and those without. There was no difference in the primary site of NSTI between cohorts (Table 1).
Table 1.
Patient Characteristics by Fungal Culture Status
| Characteristic |
No fungus (n = 524)a |
Positive fungal Cx (n = 110)a |
pb |
|---|---|---|---|
| Age | 53.9 (14.5) | 53.5 (13.2) | 0.79 |
| Gender | 0.051 | ||
| Male | 287 (54.8%) | 49 (44.5%) | |
| Female | 237 (45.2%) | 61 (55.5%) | |
| Race | 0.7 | ||
| Non-Hispanic White | 437 (83.4%) | 89 (80.9%) | |
| Black | 76 (14.5%) | 18 (16.4%) | |
| Other | 11 (2.1%) | 3 (2.7%) | |
| Insurance payor | 0.7 | ||
| Government | 293 (55.9%) | 63 (57.3%) | |
| Private | 218 (41.6%) | 43 (39.1%) | |
| Uninsured | 13 (2.5%) | 4 (3.6%) | |
| Patient has PCP | 301 (57.4%) | 61 (55.5%) | 0.7 |
| National Area Deprivation Index Quartile | 0.7 | ||
| First | 9 (1.7%) | 1 (0.9%) | |
| Second | 61 (11.6%) | 12 (10.9%) | |
| Third | 164 (31.3%) | 29 (26.4%) | |
| Fourth | 290 (55.3%) | 68 (61.8%) | |
| Live 50 miles or greater from hospital | 198 (37.8%) | 43 (39.1%) | 0.8 |
| Body mass index category | 37.1 (14.0) | 35.6 (12.0) | 0.30 |
| Diabetes | 323 (61.6%) | 82 (74.5%) | 0.010 |
| CKD | 69 (13.2%) | 20 (18.2%) | 0.2 |
| ESRD | 13 (2.5%) | 5 (4.5%) | 0.2 |
| Liver disease | 49 (9.4%) | 8 (7.3%) | 0.5 |
| COPD | 73 (13.9%) | 20 (18.2%) | 0.3 |
| CHF | 66 (12.6%) | 15 (13.6%) | 0.8 |
| TIA | 36 (6.9%) | 11 (10.0%) | 0.3 |
| MI | 54 (10.3%) | 13 (11.8%) | 0.6 |
| Steroid use | 15 (2.9%) | 5 (4.5%) | 0.4 |
| IVDU | 56 (10.7%) | 11 (10.0%) | 0.8 |
| Hemiplegia | 24 (4.6%) | 5 (4.5%) | >0.9 |
| PUD | 116 (22.1%) | 22 (20.0%) | 0.6 |
| PVD | 63 (12.0%) | 17 (15.5%) | 0.3 |
| Cancer | 61 (11.6%) | 13 (11.8%) | >0.9 |
| Charlson Comorbidity Index | 0.8 | ||
| None | 54 (10.3%) | 8 (7.3%) | |
| Mild | 136 (26.0%) | 32 (29.1%) | |
| Moderate | 136 (26.0%) | 28 (25.5%) | |
| Severe | 198 (37.8%) | 42 (38.2%) | |
| Transferred for care | 295 (56.3%) | 68 (61.8%) | 0.3 |
| Primary Site of NSTI | 0.093 | ||
| Head/neck/trunk | 107 (20.4%) | 26 (23.6%) | |
| Groin | 201 (38.4%) | 49 (44.5%) | |
| Upper limb | 57 (10.9%) | 4 (3.6%) | |
| Lower limb | 159 (30.3%) | 31 (28.2%) |
n (%); Median (SD).
Pearson’s Chi-squared test; Fisher exact test, exact test; Wilcoxon rank sum test.
CKD = Chronic kidney disease; COPD = Chronic obstructive pulmonary disease; ESRD = End stage renal failure; IVDU = Intravenous drug use; MI = myocardial infarction; PCP = primary care physician; PUD = Peptic ulcer disease; PVD = Peripheral vascular disease; TIA = transient ischemic attack.
Clinical and surgical presentation by fungal cohort
Patients in the positive fungal culture cohort were more likely to require debridement greater than 1000 cm2 (18.1% vs. 29.1%, p = 0.009), require more than three debridement operations (19.5% vs. 43.6%, p < 0.001), and have NSTIs that grew Vancomycin-resistant Enterococcus (VRE, 1.0% versus 11.8%, p < 0.001) and gram-negative organisms (62.4% vs. 73.6%, p= 0.025). There was no difference in other aspects of the hospital course, including developing sepsis, initial SOFA and APACHE II score, and pre-operative WBC and lactate. (Table 2).
Table 2.
Clinical and Surgical Presentation by Fungal Culture Status
| Characteristic |
No fungus (n = 524)a |
Positive fungal Cx (n = 110)a |
pb |
|---|---|---|---|
| Developed sepsis | 61 (11.6%) | 19 (17.3%) | 0.11 |
| SOFA score | 2.00 (1.00, 4.00) | 2.00 (1.00, 4.00) | 0.087 |
| APACHE II score | 10 (7–15) | 12 (8–17) | 0.05 |
| Pre-operative WBC | 16.9 (12.2–23.1) | 18.0 (13.3–24.5) | 0.10 |
| Pre-operative Lactate | 1.9 (1.4–2.7) | 2.2 (1.4–2.7) | 0.43 |
| Bacteremia | 88 (16.8%) | 22 (20.0%) | 0.42 |
| Fungemia | 0 (0%) | 4 (3.6%) | 0.001 |
| Debridement within 12 h | 291 (55.5%) | 67 (60.9%) | 0.30 |
| Debridement size > 1000 cm2 | 95 (18.1%) | 32 (29.1%) | 0.009 |
| Required 3+ Debridements | 102 (19.5%) | 48 (43.6%) | <0.001 |
| Required amputation | 73 (13.9%) | 15 (13.6%) | >0.9 |
| MRSA infection | 67 (12.8%) | 15 (13.6%) | 0.8 |
| VRE infection | 5 (1.0%) | 13 (11.8%) | <0.001 |
| ESBL infection | 20 (3.8%) | 5 (4.5%) | 0.8 |
| Wound culture—gram + | 471 (89.9%) | 95 (86.4%) | 0.3 |
| Wound culture—gram − | 327 (62.4%) | 81 (73.6%) | 0.025 |
| Wound culture-Anaerobes | 325 (62.0%) | 73 (66.4%) | 0.4 |
n (%); Median (Q1,Q3).
Pearson’s Chi-squared test; Fisher exact test, exact test; Wilcoxon rank sum test.
ESBL = extended spectrum beta lactamase; MRSA = Methicillin resistant staphylococcus aureus; SOFA = sequential organ failure assessment; VRE = Vancomycin-resistant Enterococcus.
Outcomes by fungal culture status
Patients with positive NSTI fungal cultures were more likely to require vasopressors (52.7% vs. 38.0%, p = 0.004), more often require ventilatory support (54.5% vs. 35.3%, p < 0.001), and experienced more ventilatory days (3 [1–6] vs. 4.5 [2–8.5], p = 0.03). They also experienced increased length of stay (18 [12–34] vs. 11 [7–18] days, p < 0.005) and have longer intensive care length of stay (3.0 [0.0–7.0]. 5.0 [2.0–11.0], p < 0.001). Patients with fungal-positive cultures also had greater 30-day (9.9% vs. 18.2%, p = 0.01) and 90-day mortality rates (12.0% vs. 20.0%, p = 0.03). There was no difference between cohorts regarding in-hospital mortality or new RRT requirements (Table 3).
Table 3.
Outcomes by Fungal Culture Status
| Characteristic |
No fungus (n = 524)a |
Positive fungal Cx (n = 110)a |
pb |
|---|---|---|---|
| AKI requiring renal replacement therapy | 32 (10.8%) | 9 (12.7%) | 0.6 |
| Required vasopressors | 199 (38.0%) | 58 (52.7%) | 0.004 |
| Required mechanical ventilation | 185 (35.3%) | 60 (54.5%) | <0.001 |
| Days requiring intensive care | 3.0 (0.0–7.0) | 5.0 (2.0–11.0) | <0.001 |
| In-hospital mortality | 36 (6.9%) | 13 (11.8%) | 0.08 |
| 30-days mortality | 52 (9.9%) | 20 (18.2%) | 0.01 |
| 90-days mortality | 63 (12.0%) | 22 (20.0%) | 0.03 |
| Length of stay (total) | 11 (7–18) | 18 (12–34) | <0.005 |
| Days on the vent | 3 (1–6) | 4.5 (2–8.5) | 0.03 |
n (%); Median (Q1,Q3).
Pearson’s Chi-squared test; Fisher exact test, exact test; Wilcoxon rank sum test.
Clinical course of patients with fungal NSTI
Of the subjects who had a fungal-positive cultures (n = 110, 17.4%), the predominant types of fungi detected were Candida albicans (n = 52, 47.3%) and Candida glabrata/parapsilosis (n = 49, 44.5%). Other types of fungi, such as Trichosporon (n = 1, 0.9%), Mucor (n = 1, 0.9%), Dermatophytes (n = 3, 2.7%), and Aspergillus (n = 0, 0%), were far less prevalent. Most fungus-positive cultures were discovered during the index debridement (n = 71, 64.5%), and the remaining were found in subsequent debridements (n = 39, 35.5%), with a median time to positive fungal growth of two days (IQR 1.0, 4.0).
Only 12 (10.9%) of the 110 patients with fungal-positive cultures received empirical antifungal coverage. The most common antifungal therapies for all fungal-positive patients were Dakins in form of local wound care (n = 33, 30.0%), Fluconazole (n = 39, 35.5%), and Caspofungin (n = 36, 32.7%). A minority of patients were treated with Amphotericin (n = 1, 0.9%) or Polymyxin B (n = 1, 0.9%), and the median time before administering an antifungal treatment was three days (IQR 1.0, 5.0) after admission. Finally, 10.9% (n = 12) of these patients were discharged before a positive fungal culture, and 2.7% (n = 3) of them experienced mortality before having a positive fungal culture (Table 4).
Table 4.
Fungal Positive NSTI Cohort Characteristics
| Characteristic |
Positive fungal Cx (n = 110)a |
|---|---|
| Time to positive fungal growth (Days) | 2.0 (1.0–4.0) |
| Time to administered antifungal therapy (Days) | 3.0 (1.0–5.0) |
| Type of fungus | |
| Candida albicans | 52 (47.3%) |
| Candida glabrata/parapsilosis | 49 (44.5%) |
| Other | 9 (8.2%) |
| Antifungal therapy | |
| Dakins/Topical Therapy | 33 (30.0%) |
| Fluconazole | 39 (35.5%) |
| Caspofungin | 36 (32.7%) |
| Other | 2 (1.8%) |
| Positive fungal culture timing | |
| Present in Index Debridement | 71 (64.5%) |
| Subsequent Debridement | 39 (35.5%) |
| Died before fungal culture resulted | 3 (2.8%) |
| Received empirical antifungal coverage | 12 (10.9%) |
| Discharged before fungal culture resulted | 12 (11.0%) |
n (%); Median (Q1,Q3).
Predictors of obtaining a positive fungal culture in patients with NSTI
On multi-variable analysis, the following variables were determined to be significant: Gender, history of diabetes, NSTI site, sepsis development, SOFA score, gram-negative wound culture, VRE infection, and pre-operative WBC. BMI was also included in the multi-variable analysis on the basis of previous evidence.5,14 After multi-variable logistic regression, female gender (OR: 1.73, 95% CI: [1.09,2.76], p = 0.021), diabetes mellitus (OR: 1.75, 95% CI: [1.07, 2.93], p = 0.030), vasopressor use (OR: 1.71, 95% CI: [1.04, 2.81], p = 0.034), and VRE infection (OR: 12.2, 95% CI: [4.16, 41.3], p < 0.001) were predictive of positive fungal cultures. Interestingly, increasing BMI was a negative predictor of positive fungal cultures (OR: 0.98, 95% CI: [0.96, 1.00], p = 0.023) (Table 5).
Table 5.
Multi-variable Logistic Regression for Predictors of Having a Fungal NSTI
| Characteristic | Odds ratio | 95% confidence interval | p |
|---|---|---|---|
| Gender | |||
| Male | |||
| Female | 1.73 | 1.09, 2.76 | 0.021 |
| Body mass index (kg/m2) | 0.98 | 0.96, 1.00 | 0.023 |
| History of Diabetes | |||
| No | — | ||
| Yes | 1.75 | 1.07, 2.93 | 0.030 |
| Primary site of NSTI | |||
| Head/Neck/Trunk | — | — | |
| Groin | 1.11 | 0.63, 2.00 | 0.7 |
| Upper limb | 0.40 | 0.11, 1.13 | 0.11 |
| Lower limb | 0.80 | 0.42, 1.51 | 0.5 |
| Developed sepsis | |||
| No | — | — | |
| Yes | 0.96 | 0.48, 1.79 | 0.9 |
| Required vasopressors | |||
| No | |||
| Yes |
1.71 | 1.04, 2.81 | 0.034 |
| SOFA Score | 0.98 | 0.90, 1.06 | 0.6 |
| Wound culture-Gram negative | |||
| No | — | ||
| Yes | 1.39 | 0.86, 2.30 | 0.2 |
| VRE Infection | |||
| No | — | — | |
| Yes | 12.2 | 4.16, 41.3 | <0.001 |
| Pre-operative WBC | 1.01 | 1.00, 1.03 | 0.087 |
Values shown in bold indicate statistically significant P-values.
Discussion
Patients with NSTIs with positive fungal cultures in our cohort experienced significantly greater 30-day and 90-day mortality compared with those with non-fungal positive infections. This subset had a prolonged hospital course, marked by prolonged total and ICU length of stays, greater rates of respiratory failure requiring mechanical ventilation, and increased need for vasopressor use. Hence, it causes patient discomfort as well as financial constraints. These findings are consistent with prior literature demonstrating both the elevated mortality and increased severity of illness associated with fungal-positive NSTIs, particularly among critically ill patients.1,5,6,9 Patients with fungal infections also required more frequent surgical debridements and had larger areas of tissue removal. This may be due to poor source control as a result of delayed appropriate antifungal therapy, which, in our study, began a median of three days after fungal cultures were obtained.15
Importantly, we noted that NSTI patients with positive fungal cultures had greater morbidity and mortality from their infection compared with patients with NSTIs due to other organisms. However, there are currently no established guidelines for empirical antifungal therapy in NSTI patients. Although early initiation of broad-spectrum antibacterial agents is standard, including anti-MRSA and β-lactam coverage, empirical antifungal therapy is not routinely recommended, even for high-risk populations.2 The recommended time from presentation to antibiotic initiation for NSTIs is within three hours.3 Yet antifungal agents in our study were initiated at a median of three days after culture collection. Only 11% of patients with positive fungal cultures received empirical antifungal therapy. Although fungal NSTIs accounted for only 17% of our entire NSTI cohort, the delay in appropriate treatment potentially contributed to worse outcomes in this subgroup. Additionally, the most common initial antifungal agent chosen was fluconazole, whereas 45% of the infections were Candida glabrata/parapsilosis, which carry high rates of azole resistance. These resistant infections may factor into the worse outcomes seen for fungal infections and suggest that more broad-spectrum empirical fungal coverage should be chosen.
To address this gap, we examined potential factors that could independently predict positive fungal cultures in NSTIs. Vasopressor requirement, diabetes mellitus, female gender, and VRE infection were all independently associated with a fungal NSTI. Although the association between VRE and fungal NSTIs has not been previously described, VRE is more prevalent in immunocompromised patients, and the overlap in susceptible patient populations may be the reason for the association.16 Although we did not include immunosuppressive therapy use in our analysis, we did include comorbidities such as history of oncologic disease, steroid use, chronic kidney disease, and cirrhosis that are associated with ongoing and chronic immunosuppressive use. Given that immunocompromised patients have historically been at greater risk for fungal infections, future studies are warranted to explore the role of immunosuppression in pathogenesis of NSTIs with positive fungal cultures.15 Female population were twice as likely to have NSTIs with positive fungal cultures. Contributing factors, including genetic, hormonal, biological, or differences in immune response, may play a role in such findings. However, the exact reasons for this increased susceptibility would require further investigation and could involve a combination of genetic, environmental, and lifestyle factors.
Our findings support consideration of broad-spectrum, parenteral empirical antifungal initiation in critically ill patients, particularly those with a history of diabetes mellitus or who are immunocompromised. It is unclear from our data whether topical treatment is sufficient. Candida species accounted for the majority of fungal infections, and approximately half of our patients grew strains of Candida known to have elevated levels of fluconazole resistance. Given these resistance patterns, caspofungin or other echinocandins may provide better empirical coverage when fungal NSTI is suspected, especially in high-risk cohorts.17-20 However, it is unclear from these data if these infections are true, invasive fungal NSTIs. It may be that the growth represents colonization rather than an invasive infection, thus prospective investigation to examine outcomes with empirical antifungal therapy in high-risk patients may be warranted.
Limitations
This study is limited by the retrospective, single-center design and reliance on EMR documentation, which may affect data completeness and generalizability. Because of this study design, the causality of these positive fungal cultures remains a question, and additional prospective studies are needed to determine if empirical antifungal coverage could improve outcomes. Despite these limitations, our study is one of the largest single-center NSTI databases and provides novel insight into the NSTIs with positive fungal cultures, their clinical predictors, treatment patterns, and outcomes.
Conclusion
Necrotizing soft tissue infections with positive fungal cultures are associated with significantly greater morbidity and mortality compared with bacterial NSTIs, as well as more complex hospital courses involving prolonged ICU stays, increased surgical interventions, and delayed appropriate therapy. Despite these risks, empirical antifungal treatment remains uncommon and is often initiated only after substantial delays. In this study, vasopressor use, diabetes mellitus, female gender, and VRE infection were identified as independent predictors of NSTIs with positive fungal cultures. Given the high prevalence of azole-resistant Candida species, echinocandins may offer superior empirical coverage. These findings underscore the need to reevaluate current treatment protocols and support consideration of prospective evaluation of empirical antifungal therapy in critically ill NSTI patients, particularly those with identifiable risk factors. Early recognition and appropriate empirical treatment may reduce diagnostic delays and improve clinical outcomes in this high-risk population.
Funding Information
NIH R35GM150968 (JRW, CC), NIH T32AI106704 (JRW, CC), NIH K08 GM157658 (AJ).
Footnotes
Authors Disclosure Statement
The authors of this study have no financial conflicts to disclose.
References
- 1.Peetermans M, De Prost N, Eckmann C, et al. Necrotizing skin and soft-tissue infections in the intensive care unit. Clin Microbiol Infect 2020;26(1):8–17; doi: 10.1016/j.cmi.2019.06.031 [DOI] [PubMed] [Google Scholar]
- 2.McDermott J, Kao LS, Keeley JA, et al. Necrotizing soft tissue infections: A review. JAMA Surg 2024;159(11):1308–1315; doi: 10.1001/jamasurg.2024.3365 [DOI] [PubMed] [Google Scholar]
- 3.Nawijn F, Smeeing DPJ, Houwert RM, et al. Time is of the essence when treating necrotizing soft tissue infections: A systematic review and meta-analysis. World J Emerg Surg 2020;15(1):4; doi: 10.1186/s13017-019-0286-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Hedetoft M, Madsen MB, Madsen LB, et al. Incidence, comorbidity and mortality in patients with necrotising soft-tissue infections, 2005–2018: A Danish nationwide register-based cohort study. BMJ Open 2020;10(10):e041302; doi: 10.1136/bmjopen-2020-041302 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Horn CB, Wesp BM, Fiore NB, et al. Fungal infections increase the mortality rate three-fold in necrotizing soft-tissue infections. Surg Infect (Larchmt) 2017;18(7): 793–798; doi: 10.1089/sur.2017.164 [DOI] [PubMed] [Google Scholar]
- 6.Khoury MK, Heid CA, Cripps MW, et al. Antifungal therapy in fungal necrotizing soft tissue infections. J Surg Res 2020;256:187–192; doi: 10.1016/j.jss.2020.06.013 [DOI] [PubMed] [Google Scholar]
- 7.Bonne SL, Kadri SS. Evaluation and management of necrotizing soft tissue infections. Infect Dis Clin North Am 2017; 31(3):497–511; doi: 10.1016/j.idc.2017.05.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Azim A, Ahmed A. Diagnosis and management of invasive fungal diseases in non-neutropenic ICU patients, with focus on candidiasis and aspergillosis: A comprehensive review. Front Cell Infect Microbiol 2024;14:1256158; doi: 10.3389/fcimb.2024.1256158 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Webb BJ, Ferraro JP, Rea S, et al. Epidemiology and clinical features of invasive fungal infection in a US health care network. Open Forum Infect Dis 2018;5(8):ofy187; doi: 10.1093/ofid/ofy187 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Yadgar RJ, Bhatia N, Friedman A. Cutaneous fungal infections are commonly misdiagnosed: A survey-based study. J Am Acad Dermatol 2017;76(3):562–563; doi: 10.1016/j.jaad.2016.09.041 [DOI] [PubMed] [Google Scholar]
- 11.Benedict K, Whitham HK, Jackson BR. Economic burden of fungal diseases in the United States. Open Forum Infect Dis 2022;9(4):ofac097; doi: 10.1093/ofid/ofac097 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.2025 ICD-10-CM diagnosis code M72.6. 2025. Available from: https://www.icd10data.com/ICD10CM/Codes/M00-M99/M70-M79/M72-/M72.6 [Last accessed: April 9, 2025].
- 13.Robert Graham Center. Social deprivation index (SDI). 2025. Available from: https://www.graham-center.org/maps-data-tools/social-deprivation-index.html [Last accessed: August 7, 2025].
- 14.Jalilvand AD, Villarreal M, Schubauer K, et al. Characterization of necrotizing soft tissue infections in obese and non-obese patients. Surg Infect (Larchmt) 2022;23(5):475–482; doi: 10.1089/sur.2022.055 [DOI] [PubMed] [Google Scholar]
- 15.Wahab A, Sanborn D, Vergidis P, et al. Diagnosis and prevention of invasive fungal infections in the immunocompromised host. Chest 2025;167(2):374–386; doi: 10.1016/j.chest.2024.08.046 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Johnstone J, Chen C, Rosella L, et al. ; Ontario VRE Investigators. Patient- and hospital-level predictors of vancomycin-resistant enterococcus (VRE) bacteremia in Ontario, Canada. Am J Infect Control 2018;46(11):1266–1271; doi: 10.1016/j.ajic.2018.05.003 [DOI] [PubMed] [Google Scholar]
- 17.Talapko J, Juzbašić M, Matijević T, et al. Candida albicans—The virulence factors and clinical manifestations of infection. J Fungi (Basel) 2021;7(2):79; doi: 10.3390/jof7020079 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Seagle EE, Williams SL, Chiller TM. Recent trends in the epidemiology of fungal infections. Infect Dis Clin North Am 2021;35(2):237–260; doi: 10.1016/j.idc.2021.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Bueid A, Howard SJ, Moore CB, et al. Azole antifungal resistance in Aspergillus fumigatus: 2008 and 2009. J Antimicrob Chemother 2010;65(10):2116–2118; doi: 10.1093/jac/dkq279 [DOI] [PubMed] [Google Scholar]
- 20.Arastehfar A, Lass-Flörl C, Garcia-Rubio R, et al. The quiet and underappreciated rise of drug-resistant invasive fungal pathogens. J Fungi (Basel) 2020;6(3):138; doi: 10.3390/jof6030138 [DOI] [PMC free article] [PubMed] [Google Scholar]