Risk Factors for Predicting Mortality and Amputation of Patients with Necrotizing Soft-Tissue Infections: Retrospective Analysis of 111 Cases from a Single Medical Center

Objective Necrotizing soft-tissue infections (NSTIs) are rare clinical infections with surgical emergencies having a high mortality rate. This study aimed to investigate risk factors for mortality and amputation of patients with NSTI. Methods We retrospectively analyzed critical factors for outcomes of 111 patients with NSTI hospitalized in our department from 1 January 1999 to 31 December 2018. NSTI diagnosis was based on the patient's clinical characteristics, laboratory risk indicator for necrotizing fasciitis (LRINEC) score, laboratory test data, and microbiological findings in blood and wound culture. The risk factors for mortality and amputation of NSTI were determined using univariate or multivariate logistic regression analysis, receiver operating characteristics (ROC), and the area under the ROC curve (AUC) at 90 days after admission. Results Diagnosis of 111 patients with NSTI was confirmed according to clinical features, LRINEC score, image data, laboratory findings, and microorganism culture in blood and wounds. The mortality rate was 9.91% (11/111) at day 90 follow-up. High white blood cell (WBC), low hematocrit (HCT), and multiple surgeries were identified to be critical risk factors for NSTI mortality in univariate and multivariate logistic analyses. AUCs, 95% confidence intervals (CI), and P values of risk factors were 0.699, 0.54–0.95, and P = 0.0117 for high WBC; 0.788, 0.63–0.97, and P = 0.0006 for low HCT; and 0.745, 0.59–0.90, and P = 0.0018 for multiple surgeries, respectively. These patients also had high LRINEC scores. Amputation occurred in 34.23% (38/111) of patients. Risk factors for amputation were higher age, low hemoglobin (Hb), and multiple wounds. AUCs, 95% confidence intervals (CI), and P values were 0.713, 0.11–0.32, and P < 0.0001 for higher age; 0.798, 0.08–0.29, and P=0.0007 for low Hb; and 0.757, 0.17–0.34, and P  < 0.0001 for multiple lesion sites, respectively. Conclusions High LRINEC scores, high WBC, low HCT, and multiple surgeries were relevant to increased mortality. Higher age, low Hb, and multiple wounds were associated with amputation risk. These clinical features must be paid attention to when patients are diagnosed with NSTI.


Introduction
Necrotizing soft-tissue infections (NSTIs) are a group of rare and severe infectious diseases with a high mortality rate [1][2][3].Te typical features of NSTI include extensive soft tissues necrotizing, including necrotizing fasciitis (NF), myositis, gas gangrene, Fournier gangrene (FG), and systemic signs of toxicity from sepsis to infectious shock [4,5].Multiple organs in NSTI are frequently damaged with varying degrees [6][7][8].NSTI can be divided into type I NSTI (polymicrobial) and type II NSTI (monomicrobial) according to microbial infection types [1].Te estimated NSTI incidence varies between 0.3 and 15.5/100000 inhabitants from Europe or the USA to Asia [5,[7][8][9][10][11].In addition, NSTIs have up to 20-35% mortality rate [9,[12][13][14].Most patients with NSTI must be admitted into the intensive care unit (ICU) for treatments.Te therapeutic methods for NSTI include urgent surgical debridement of infected tissues, broad-spectrum antibiotics use, and protection of organ functions as soon as possible [1,7].In addition, inhibitory peptide administration of pathogen binding to CD28 is now under clinical trial, which is a very promising therapeutic method [15].However, the prognosis of most patients with NSTI is poor.Up to 20% of NSTI require amputation and 30% of patients have mild to severe organ dysfunctions [16,17].
Outcomes of NSTI signifcantly vary depending on initial diagnosis time and therapeutic methods.Misdiagnosis or delayed treatments of NSTI at an early stage frequently occurred because NSTI symptoms sometimes resemble clinical features of other soft-tissue infections and have no specifc diagnostic biomarkers [12,18,19].Previous reports have presented some risk factors for the outcome of NSTI, including higher age, underlying diseases, aberrant laboratory fndings, and clinical characteristics [20][21][22].However, real risk factors of NSTI outcome remain undefned.Te present study was conducted to assess the risk factors for mortality and amputation in NSTI.

Materials and Methods
2.1.Patients.We retrospectively reviewed and analyzed the medical records of 111 NSTI patients who were diagnosed and treated in our hospital from 1 January 1999 to 31 December 2018 for this study.Study variables for outcome evaluation included age, gender, underlying diseases, imaging data, microbiological culture of blood or wounds, laboratory fndings, organ functions, and therapeutic approaches.Te inclusion criteria are as follows: (1) all patients were 18 years or older and (2) all patients were hospitalized in our department.Te exclusion criteria are as follows: (1) patients were diagnosed with abscesses and phlegmons and (2) the outpatients were not included in this study.Risk factors evaluation for mortality and amputation was also determined according to LRINEC score [23], which was calculated based on serum C-reaction protein (CRP), WBC, Hb, serum sodium, creatinine, and glucose levels.

Microorganism Culture in Wound or
Blood.Bacterial culture from peripheral blood and wound soft tissues covered Baumannii, Staphylococcus, Klebsiella pneumonia, and Escherichia coli strains growth.

Imaging Examinations.
Te image studies included echocardiography, computed tomography angiography (CTA), computed tomography (CT), magnetic resonance imaging (MRI), and routine ultrasound in the heart, chest, and arteries and veins of the extremities.

Managements.
Managements of NSTI included surgical debridement intervention, empirical broad-spectrum antimicrobial use, and protection of dysfunctional organs.Intravenous immunoglobulin (IVIG), hyperbaric oxygen therapy (HBOT), continuous renal replacement therapy (CRRT) or hemodialysis, and ventilator use were based on disease status following NSTI management protocol.
2.6.Outcome.We focused on the chances of mortality and amputation at 90 days after surgery, antibiotic therapy, and protective treatments of organs.

Statistical Analyses.
Quantitative data are presented as median (interquartile range, IQR) or mean ± standard deviations (SD).Qualitative data are reported as percentages (%).Te prevalence of events presents with frequency and cumulative frequency.Variable comparisons among diferent groups were performed using the Student's t-test or the Mann-Whitney test.For risk factors evaluation of mortality and amputation in NSTI, we frst screened using the univariate logistic regression analysis.Ten, multivariate logistic regression analysis was used combined with P value and 95% CI for the calculation of variables at optimal AUC cut-of value with high sensitivity and specifcity to determine the fnal risk factors.P < 0.05 was considered as a signifcant diference.

Laboratory and Imaging Findings.
Peripheral blood counts and biochemistry data at admission are shown in Table 2. Tese biomarkers were used to compare later therapeutic efects.We also documented imaging data at admission (Table 3), including echocardiography, computed 2 Emergency Medicine International tomography angiography (CTA) in the extremities, regular or enhanced CT, MRI, chest or extremities X-ray, and ultrasound.Te results showed that there were positive fndings in less than 50% of patients.

Microorganism Culture in
Wound or Blood.NSTIs are frequently involved in various microorganism infections.Terefore, we routinely performed microorganism cultures in wounds or blood.As shown in Table 4, positive rates in Baumannii, Staphylococcus, Klebsiella pneumonia, and Escherichia coli were 10.91% (12 cases), 16.36% (18 cases), 11.82% (13 cases), and 8.18% (9 cases), respectively.Infectious chances with more than one kind of bacteria were 70%, and the rate of noninfection was 30%.Tese results revealed that an antibiotic regimen may be necessary because most patients have at least one bacterial infection.5.A total of 111 patients have experienced surgical debridement in hospitalization.Some patients had more than one surgery, ventilator, or norepinephrine use or lived in the ICU.

Risk Factor Analysis for NSTI Mortality.
Te mortality rate of NSTI patients is high.We performed univariate and multivariable logistic analyses for risk factors.LRINEC score was a key factor (Table 6).In this study, 11 of 111 patients died at 90 days after admission.LRINEC score of live patients was signifcantly lower than that of dead patients (P < 0.05).Laboratory data such as CRP, WBC, lactate, and PCT of live patients were dramatically lower than that of the dead patients (P < 0.05).In contrast, Hb and HCT of live patients were higher than that of the dead patients (Table 7).
We also performed ROC analysis of patients with mortality  9).Tese results indicated that high WBC, low HCT value, and surgery at admission were critical factors for mortality prediction.

Risk Factors of Amputation for NSTI Patient.
Amputation is a severe outcome of NSTI.To assess risk factors for amputation, we analyzed diferent data in clinical factors and laboratory fndings.We found that higher age, low Hb, and multiple wound sites are positively correlated with amputation (Figure 4).Te AUC value can predict amputation using age, hemoglobin, and wound sites as variables with the logistic model (Figure 4(a)) or without the logistic model (Figure 4(b)).Te AUCs of age, hemoglobin, and wound sites were 0.713, 0.681, and 0.757 in the logistic model, respectively (Table 10).We also performed a ROC comparison and the Mann-Whitney test.P values and 95% CI of age, hemoglobin, and wound sites for amputation prediction were 0.11-0.32 and P < 0.0001; 0.08-0.29 and P � 0.0007; and 0.17-0.34and P < 0.0001, respectively (Table 11).Tese results revealed that old age, low hemoglobin, and multiple wound sites at admission were critical markers for amputation prediction.

Discussion
NSTIs are severe and life-threatening diseases, which are frequently misdiagnosed for other diseases because of no typical symptoms [5,16,19].Terefore, the identifcation of risk factors for outcomes of NSTI is a critical issue.Here, we performed a single center cohort study of 111 patients and reported that 11 of 111 NSTI (9.9%) patients were dead and  Emergency Medicine International    NSTIs, necrotizing soft-tissue infections; SD, standard derivation; CRP, C-response protein; WBC, white blood cell; Hb, hemoglobin; HCT, hematocrit; BUN, blood urea nitrogen; Cr, creatinine; PCT, procalcitonin.admission.LRINEC score, WBC, HCT, and surgery were key indicators for mortality prediction.Higher age, hemoglobin, and multiple wound sites were critical factors for predicting amputation.

Emergency Medicine International
Te studies showed that the overall mortality rate of NSTI was up to 35% [24,25].Te reasons for the high mortality rate vary from one hospital to another [20,26,27] but one factor may be severe sepsis [28].Te mortality rate in this cohort study was 9.9% (11/111), which was lower than that of the other reports [29][30][31] but close to other results [7,8,32,33].Our low mortality may be because of early surgical intervention, active antibiotics use, IVTG, and HBOT administration.We identifed that the LRINEC score was an important marker.Tis result is consistent with other studies [32,34,35].Hua et al. and the other team also reported that patients aged >75 years old had a high mortality rate [3,7,8].Te other team found that patients with underlying diseases such as diabetes, cardiovascular disease, and higher lactate levels were associated with a high   mortality rate [7,36,37].Our univariate logistic analysis also showed that lactate in dead patients was signifcantly higher than that in survival patients (Table 7; P < 0.05).High WBC, low HCT, and multiple surgeries may refect that patients have experienced extensive severe sepsis.Indeed, Madsen et al. [7] reported that higher lactate levels may be associated with septic shock.Terefore, if patients with NSTI have these abnormal parameters, then physicians must pay more attention to these patients.
Other interesting fndings in this study revealed that higher age, low hemoglobin, and multiple wounds were strongly associated with a high amputation rate.Previous reports indicated that higher age was relevant to high mortality, especially for patients who had underlying diseases [3,7,28].Tese apparent biomarkers showed that the patients had low immunity because of old age.Low hemoglobin and multiple wounds indicated that NSTI had severe infections.Terefore, early radical surgical debridement and antibiotic use are key factors for protecting amputation.Other reports also indicated that early surgery and antibiotic therapy were very important for amputation [1,38,39].Te study also revealed that inhibitory peptide   10 Emergency Medicine International administration of pathogen binding to CD28 can protect organ functions in NSTI [15].Here, our study showed that Baumannii, Staphylococcus, Klebsiella pneumonia, and Escherichia coli strains had a high positive infectious rate.
Te study indicated that group A. streptococcus and Staphylococcus aureus infections were especially associated with the lower extremities [7,31].Tese results confrmed that early antibiotic therapy and surgery are critical methods for avoiding amputation and death [31].Tis study had few limitations which are as follows: (I) current results were from a single hospital with a retrospective study.Tis may cause a selection bias and limit the evaluation of relationships between risk factors and outcomes.(II) Te total patients were 111 cases in this study.Small sample sizes may afect multivariate regression analysis.Sample size limitation was also reported in other studies because NSTI is a rare disease [1,2,40].To confrm the present fndings, we plan to expand our study to multiple hospitals with more participants.(III) NSTI is a severe disease.Its treatment protocols are hard and identical because of complex clinical symptoms and various laboratory fndings.Tis limitation may afect outcomes of NSTI.(IV) Te present study followed up for 90 days.We counted dead and amputated patients at this cut-of time.Tis follow-up time may afect study conclusions.

Conclusion
Te present study revealed that high LRINEC scores, elevated WBC, low HCT, and multiple surgeries were associated with mortality.Higher age, low hemoglobin, and multiple wounds are key markers for indicating amputation.Early antibiotic therapy and surgery are important procedures for avoiding amputation and mortality.

3. 4 .
Managements.After patients were admitted to the hospital, we performed radical surgical debridement intervention and antibiotics administration within 24 hours.Surgical procedures are shown in Figures1 and 2. Te frst patient was diagnosed with Fournier gangrene (FG) because he had extensive necrosis of soft tissues in the scrotum (Figure1(a)).We performed radical surgical debridement (Figures1(b) and 1(c)), and the damaged area gradually recovered (Figures1(d) and 1(f )).Te second patient was a 56-year-old man with diabetes mellitus.He had necrosis of skin and soft tissue in the lower extremities (Figures2(a) and 2(b)) because of a car accident.We also found necrosis of tendons and muscles during debridement (Figures2(c) and 2(d)), fresh wounds (Figures2(e) and 2(f )), and local exposure tibia (Figure2(g)).After 3 negative pressure therapies, he experienced skin grafting over the tibia (Figures2(g) and 2(h)) and was fnally recovered (Figure2(j)).Te detailed therapeutic methods and times from emergency or hospitalized clinics are shown in Table

Figure 1 :
Figure 1: Surgical debridement procedure in a patient with NSTI.(a-f ) All steps from initial surgery to end of surgery.

Figure 2 :
Figure 2: Surgical management of a patient with necrosis of skin and soft tissue in the lower extremity after avulsion.(a, b) Necrosis of skin and soft tissue in the left lower extremity.(c, d) Necrosis of tendons and muscles during debridement.(e, f ) Fresh wound and local exposure tibia.(g) After 3 negative pressure therapy.(h, i) Skin grafting over tibia.(j) Recovery of the wound.

Figure 3 :
Figure 3: AUC of WBC, HCT, and surgery at admission in nonsurvivor patients with NSTI.(a) ROC plots show results of WBC, HCT, and surgery for predicting mortality in the logistic model.(b) ROC plots show results of WBC, HCT, and surgery for predicting mortality without the logistic model.NSTI: necrotizing soft-tissue infections; ROC: receiver operating characteristics; AUC: area under the ROC curve; WBC: white blood cell; HCT: hematocrit.

Figure 4 :
Figure 4: ROC from patient's age, hemoglobin, and wounds for predicting amputation.(a) ROC plots show results of age, hemoglobin, and wounds for predicting amputation in the logistic model.(b) ROC plots show results of age, hemoglobin, and wounds for predicting amputation without the logistic model.NSTI: necrotizing soft-tissue infections; ROC: receiver operating characteristics.

Table 1 :
Demographic characteristics of the patients with NSTI.

Table 3 :
Image studies in 111 NSTI patients.

Table 4 :
Bacterial culture features of NSTI patients.

Table 5 :
Management of the patients with NSTI.

Table 6 :
LRINEC score comparison of live and dead patients.

Table 7 :
Laboratory fndings of live and dead patients with NSTI.

Table 9 :
Comparison of ROC in diferent parameters of dead patients.

Table 10 :
ROC and AUC of patients with NTSI for amputation.