Diagnostic Key Points and Surgical Management of Necrotizing Fasciitis: A Retrospective Study

Abstract

Necrotizing fasciitis (NF) is a fatal disease with a high mortality rate that can be easily misdiagnosed. The aim of this study was to improve the diagnostic rate of NF and overall survival. We conducted a single-center, retrospective, noncontrolled study involving 36 patients who were admitted to our department between December 2017 and October 2019, and summarized the diagnostic key points and timing of surgical treatment. All patients were diagnosed at our department and underwent multiple courses of treatment. The records included information regarding underlying diseases, bacterial culture results, laboratory risk indicator for necrotizing fasciitis (LRINEC) score, number of procedures, and type of antibiotics. All 36 cases of NF were cured and showed good patient condition on follow-up; the mean number of surgeries was three, and the mean duration of hospitalization was 37 days (range, 21-83 days). The LRINEC scores of 16 patients were ≥8 points. Seventeen patients with underlying diabetic disease had higher inflammatory index scores than those without diabetes. The LRINEC scores of patients with (n = 17) and without (n = 19) DM were 7.40 ± 2.99 and 3.80 ± 2.39, respectively (P < .01). Cases of NF that were treated with early incision and surgical abscess drainage required fewer surgeries and a shorter length of hospitalization. Thus, surgeons should be more aware of NF and aim to make an early and accurate diagnosis using various approaches. Complete surgical debridement plays an essential role in NF treatment, and diabetes mellitus is a significant adverse factor that exacerbates the severity of NF. Negative-pressure techniques are useful in cases involving nonanaerobic infections and cause minimal complications.

Necrotizing fasciitis (NF) is a specialized infection that spreads rapidly along the fascia and leads to soft tissue necrosis. The three most important early clinical symptoms of NF are fever, gradually worsening pain after tissue injury, and rapidly progressing erythema and swelling. The infected area usually extends beyond the area of original skin involvement, and erythema and swelling can be observed on palpation of the lesion. There is also disproportionate pain in the area of the involved skin. Since the infection site is the fascia, the occurrence of severe inflammatory symptoms may precede skin changes. ¹ Therefore, misdiagnosis or diagnosis failure usually occurs during the early stages of NF,^2–4 with a current misdiagnosis rate of 75%. ⁵ Delayed treatment may cause severe systemic infections, multiple organ dysfunction syndromes, and a high mortality rate. ⁶ NF shows an incidence of ∼0.3 new cases per million-persons/year and a mortality rate of up to 29%.^2,7 Other atypical symptoms include hypotension, skin necrosis, ecchymosis, and skin ischemia, which is accompanied by blisters or bullae. The presence of bullae showed a moderate positive correlation with amputation and mortality rate, suggesting that this symptom reflects a turning point in the prognosis of the disease. ³

The determination of a gold standard for the diagnosis of NF has been difficult. The “bedside finger test” is an overt significant auxiliary examination of NF that is performed at the bedside or during procedures to assist diagnosis. ⁸ In addition, many auxiliary methods are used for NF examinations, including B-scan ultrasound, enhanced computed tomography (CT), and magnetic resonance imaging (MRI). Early diagnosis and urgent surgical intervention are essential in cases of NF, ⁹ especially since the mortality rate has been shown to be lower in cases undergoing early initial debridement. ¹⁰ Thus, even if the condition has not been conclusively diagnosed, early initial debridement should be performed if NF is suspected.

This study summarizes the diagnostic methods, surgical treatments, and results of NF cases treated at our department. The analysis of diagnosis pitfalls may reveal clinical features that are important for early diagnosis. Early diagnosis and treatment are crucial against NF, and the importance of early surgery, even if incomplete with simple surgical drainage, was explored to facilitate the improvement of NF. Further, the effects of diabetes on NF were examined and the use of negative-pressure wound therapy (NPWT) related to the clinical experience of NF is summarized.

Patients and Methods

Patients

This was a single-center, retrospective, noncontrolled study involving 36 patients admitted to our department between December 2017 and October 2019 on the basis of the following inclusion criteria: diagnosis of NF or non-NF at the first hospital visit, but NF confirmed after admission to our department. The exclusion criteria were as follows: NF associated with osteomyelitis; NF of head/neck with intracranial infection; and age under 18 years.

Treatment Protocol

The surgical procedure included complete debridement and drainage. The “finger test” was conducted pre- and intraoperatively to predict the area of necrosis. This test involved probing the necrotic hole or incision and drainage opening with fingers to determine if the lesion site can be easily separated without resistance. Intraoperatively, necrotic tissue, including necrotic fascia, soft tissue, and skin, was removed entirely. In patients with a normal skin surface along with pus and necrosis in the deep fascia layers, the epidermis was incised. Removal of ischemic and necrotic tissue was expected to allow better control of infection. Failure to separate the inferior fascial layer and the absence of fascial necrosis were considered as endpoints of debridement. Finally, a 3% hydrogen peroxide solution at 37 °C, 0.25% iodophor solution, and balanced salt solution were used sequentially in alternating washes, and complete drainage was performed (Figures 1 and 2).

Figure 1.

One week following lesion development, the skin temperature of the right lower extremity was significantly elevated, and the skin was red and swollen. The hole in the popliteal fossa was a preadmission incision and drainage opening, with exudation of purulent secretions. Subcutaneous fluid fluctuation could be palpated in the thigh. The skin surface was scattered with blisters.

Figure 2.

During the debridement, the “finger test” was performed along the original upper popliteal fossa drainage opening, and the normal skin on the surface was incised while exploring, revealing extensive necrosis of the deep fascial layer, partial myofascial necrosis, and microvascular embolization of the skin, which was accompanied by a large amount of purulent discharge and malodor. The peripheral circulation was good.

After the necrotic fascia was excised, complete hemostasis was achieved. The NPWT technique was used for wounds free of anaerobic infections. The V.A.C. GranuFoam dressing (KCI) was cut according to the wound's shape and size to ensure that there was no cavity in the wound and to facilitate adequate drainage. The continuous negative suction pressure of −125 mm Hg was used to ensure optimal perfusion. ¹¹ A compressed gauze dressing was applied to all the outliers of V.A.C. GranuFoam™ dressing for at least 48 h to prevent bleeding.

During the second operation, repair surgery was considered if there was no necrotic tissue and fresh granulomatous tissue was visible. Purulent secretions and necrotic tissue were still present in some patients after the first surgery, and the dead spaces could even be seen disseminated along with the deep fascia layer and/or sarcolemma. In such cases, complete debridement and drainage were performed until the infected areas were controlled. During debridement, some wounds recovered whereas others showed progressive necrosis; therefore, wound repair and further debridement was performed simultaneously to reduce the wound size. Repair surgery included either direct closure or skin grafting. The incision was sutured, and the sutures were removed 7 to 12 days later. If the patients showed signs of infection at the incision site, the sutures surrounding the infection site were immediately released. A regular local dressing change was conducted until the wound healed or secondary suturing was performed.

Diagnosis and General Treatment

Specimens for bacterial culture and drug-susceptibility analyses were collected before surgery. MRI, CT or B-mode ultrasound data were collected in the ER to facilitate the assessment of the infection. The records collected included the results of laboratory tests, including routine blood tests, C-reactive protein (CRP) and procalcitonin (PCT) measurements, and biochemical examinations. The Laboratory Risk Indicator for Necrotizing Fasciitis (LRINEC) was used to diagnose and differentiate NF from other soft tissue infections. ¹²

The patients’ vital signs were continuously monitored, and their systemic condition was comprehensively evaluated. Antibiotics were conventionally administered as empirical treatment. The patient was switched to an antibacterial-sensitive antibiotic after the pathogen drug sensitivity results were obtained. The patients’ underlying diseases were treated simultaneously. For example, the blood glucose levels of patients with diabetes were maintained within the range for optimal glycemic control. We maintained the water and electrolyte balance to perform systemic correction of anemia and hypoproteinemia.

Data Analysis

Patient data of vital signs and symptoms, location of the infection, underlying disease, laboratory findings on admission, bacteriological status, number of procedures, and outcomes were retrospectively recorded. Laboratory data were scored according to the LRINEC scale, and wound healing was based on the absence of laceration after suture removal. After discharge from the hospital, all patients were followed-up for at least 6 months in the outpatient clinic to ascertain any complications.

Statistical Methods

Categorical variables were expressed as counts and percentages. An independent Student's t-test was used to analyze the numerical data. Descriptive data were presented as means with standard deviations for quantitative variables and frequencies for categorical variables. Statistical significance was set at P < .05. SPSS 26.0 for Windows was used for all statistical analyses.

Results

Patient Population

We reviewed 37 cases in our department, excluding one with associated intracranial infection. These 36 patients included 24 men (66.67%) and 12 women (33.33%) with ages ranging from 25 to 85 years and a mean age of 53 ± 4.6 years. The causes of NF included folliculitis (11 [30.56%] cases), trauma (10 [27.78%] cases), burns/scalding (6 [16.67%] cases), foreign body (1 [2.78%] cases), cesarean delivery (1 [2.78%] cases), and idiopathic causes (7 [19.44%] cases). Twenty-eight patients had underlying diseases, of which 17 (47.22%) had type 2 diabetes mellitus. Five patients had multiple comorbidities. The infection sites included the lower limbs (19 [52.78%] cases), face (2 [5.56%] cases), abdomen (6 [16.67%] cases), arms (2 [5.56%] cases), trunk (6 [16.67%] cases), and hands (1 [2.78%] cases). Thirty-six patients were not diagnosed with NF at the first evaluation in the hospital or emergency care unit, and were instead diagnosed as showing an abscess (13 cases), soft tissue infection (10 cases), erysipelas (3 cases), cellulitis (6 cases), and gangrene (4 cases). If erysipelas, cellulitis, and gangrene were considered as misdiagnoses, then the misdiagnosis rate was above 36% (Table 1). The bacteriological results were negative in 9 patients admitted to our department, and 1 patient was infected with two different bacterial strains (Table 2). Additional patient characteristics and antibiotic use are presented in Table 3.

Table 1.

Demographic Characteristics, Eetiology, Comorbidities, and Infection Site.

	N	%
Sex
Male	24	66.67
Female	12	33.33
Etiology
Folliculitis	11	30.56
Trauma	10	27.78
Burn/scald	6	16.67
Ink injection (foreign body)	1	2.78
Caesarean section	1	2.78
Idiopathic causes	7	19.44
Comorbidity
Diabetes	14	38.89
Diabetes + lower limb venous and lesser saphenous vein thrombosis	1	2.78
Diabetes + arteritis	1	2.78
Diabetes + acute lymphoblastic leukemia	1	2.78
Measles	1	2.78
Hyperuricemia	4	11.11
Hyperuricemia + hypertension	1	2.78
Hypertension	5	13.89
No comorbidity	8	22.22
Infection site
Lower limbs	19	52.78
Face	2	5.56
Abdomen	6	16.67
Arms	2	5.56
Hands	1	2.78
Trunk	6	16.67
First diagnosis
Abscess	13	36.11
Soft tissue infection	10	27.78
Erysipelas	3	8.33
Cellulitis	6	16.67
Gangrene	4	11.11

Table 2.

Bacteriological Results.

Strains	N	%
Staphylococcus aureus	9	25.00
Klebsiella pneumoniae	5	13.89
Klebsiella pneumoniae + Citrobacter braakii	1	2.78
Acinetobacter baumannii	5	13.89
Pseudomonas aeroginosa	3	8.33
Escherichia coli	3	8.33
Coccus	1	2.78
Negative	9	25.00

Table 3.

Additional Patient Characteristics.

Patient	First diagnosis	Antibiotics	Early treatment	Surgical treatment	Operations	LRINEC
1	Abscess	L + ICS + LF	/	FD + AMP	9	10
2	Abscess	PSTS	Drainage	FD	2	8
3	Abscess	PSTS + CSSS + T	Drainage	FD	3	10
4	Abscess	C	/	FD	3	2
5	Abscess	PSTS + CSSS + T	/	FD	3	10
6	Erysipelas	PSTS + LF	/	FD	2	3
7	Abscess	PSTS + T	/	FD	4	9
8	Erysipelas	C	/	FD	3	8
9	Cellulitis	CSSS + LF	Drainage	FD	2	3
10	Soft tissue infection	C	/	FD	2	2
11	Soft tissue infection	C	/	FD	2	5
12	Soft tissue infection	C + T	/	FD	3	8
13	Soft tissue infection	C	/	FD	3	2
14	Cellulitis	PSTS + CSSS + V	Drainage	FD	2	3
15	Abscess	PSTS + C	/	FD	2	4
16	Abscess	L + LF	/	FD	2	8
17	Gangrene	C + V + L + ICS	/	FD	2	3
18	Abscess	L + CSSS	Drainage	FD	3	1
19	Gangrene	L + PSTS	/	FD	4	4
20	Cellulitis	L + PSTS	/	FD	8	9
21	Abscess	C	Drainage	FD	2	2
22	Soft tissue infection	PSTS + CSSS + T	/	FD	3	9
23	Soft tissue infection	PSTS + CSSS + T	Drainage	FD	3	10
24	Abscess	L + CSSS	/	FD	2	3
25	Abscess	C	/	FD	2	2
26	Erysipelas	PSTS + CSSS + T	Drainage	FD	4	9
27	Soft tissue infection	C	/	FD	2	2
28	Cellulitis	L + CSSS	Drainage	FD	2	4
29	Gangrene	PSTS + CSSS + T	Drainage	FD	5	10
30	Abscess	C	/	FD	2	2
31	Gangrene	PSTS + CSSS + T	/	FD	3	8
32	Cellulitis	C + V + L + ICS	Drainage	FD	3	8
33	Soft tissue infection	L + CSSS	Drainage	FD	2	3
34	Soft tissue infection	CSSS + LF	/	FD	2	5
35	cellulitis	PSTS + CSSS + T	/	FD	4	9
36	Soft tissue infection	C	/	FD	2	3

Abbreviations: L, linezolid; C, cefodizime; ICS, imipenem and cilastatin sodium; T, tinidazole; PSTS, piperacillin sodium and tazobactam sodium; LF, levofloxacin; V, vancomycin; CSSS, cefoperazone sodium and sulbactam sodium; FD, fasciotomy and debridement; AMP, amputation.

Before the surgical procedure, 8 patients underwent CT, 10 underwent MRI, and 12 underwent B-mode ultrasound examinations to indicate soft tissue edema and fluid accumulation in the deep fascia. All patients showed pathological results that were characterized by focal amounts of acute and chronic inflammatory cell infiltration and necrotic abscess formation in the subcutaneous fibrous connective tissue.

All patients in our hospital underwent surgical debridement as soon as possible. All patients underwent at least two procedures and an average of three procedures; one patient underwent nine procedures and subsequently underwent partial amputation of the foot (Table 3). Twelve patients underwent early incision and simple abscess drainage in the primary hospital, but these procedures did not completely stop NF progression. For the 12 patients who underwent simple incision and drainage and the 24 patients who did not receive surgical management before our surgery, the total number of surgeries was 2.60 ± 0.55 and 3.33 ± 2.19, respectively (P = .476). Moreover, the length of hospitalization was significantly shorter in patients who underwent early surgical drainage (26.40 ± 5.68 vs. 40.06 ± 17.92 days; P = .116) (Table 4).

Table 4.

Previous Surgical Drainage Early in the Disease Versus the Number of Operations and Days in the Hospital.

Characteristic	Drainage (n = 12)	Undrained (n = 24)	P
Number of operations	2.6 ± 0.55	3.33 ± 2.19	.476
HOD	26.40 ± 5.68	40.06 ± 17.92	.116

All patients underwent clinical examination and laboratory blood analyses. The LRINEC [8] was used for grading, and 16 patients (44.44%) had scores of ≥8 points (Table 3). LRINEC scores were higher in the 17 patients with diabetes than in the 19 patients without diabetes (P < .05). Surgery resulted in a significant decrease in the LRINEC score, especially the inflammatory index CRP, the white blood cell component of the LRINEC score, postoperatively (P < .05). In the comparison of patients with and without diabetes, the CRP level and white blood cell count components of the LRINEC scores were significantly higher in the patients with diabetes, and significant differences were observed. Similar to preoperative and postoperative CRP, the white blood cell count component of the LRINEC score also showed a significant difference (P < .05) (Table 5). The mean duration of hospitalization was 37 days (range, 21-83 days). All patients recovered and were discharged. No recurrence was observed in any of the patients within 6 months of follow-up.

Table 5.

The Impact of Diabetes on Laboratory Data.

Group	CRP (points)	WBC (points)	CRP + WBC (points)	LRINEC scale (points)
DM (n = 17)	2.80 ± 1.93	1.00 ± 0.82	3.70 ± 2.26	7.40 ± 2.99
Non-DM (n = 19)	0.80 ± 1.69	0.40 ± 0.70	1.20 ± 1.87	3.80 ± 2.39
P	.024*	.095	.015*	.008*
Preoperative (n = 36)	2.00 ± 2.06	0.81 ± 0.83	2.81 ± 2.51	5.94 ± 3.09
Postoperative (n = 36)	0	0.06 ± 0.25	0.06 ± 0.25	2.63 ± 1.20
P	0*	.002*	.001*	.001*

*P < .05.

Discussion

This study involved a review of the diagnosis and treatment of cases in which NF was successfully cured. The bacterial culture-positive cases in this group did not involve fungal infections. There are no specific and sensitive markers for NF. ¹³ In this study, most cases in the early clinical stage were not diagnosed with NF at the primary hospitals. The misdiagnosis rate was as high as 36% in the primary hospitals, excluding abscess and soft tissue infection diagnoses. We argue that a combined assessment of clinical symptoms, bacterial culture, and laboratory examination findings can assist in the comprehensive diagnosis of NF.

The earliest clinical symptoms of NF may be overt and visible, or they may be covert and hidden. Skin manifestations are often accompanied by general symptoms such as redness, and fever. However, the progression of symptoms is rapid, and redness can spread rapidly. In fact, before the skin changes, the underlying fascial tissue is already infected and necrotic, which can interfere with early diagnosis. At this point, the soft tissue condition can be determined by MRI, which can identify the level of soft tissue necrosis and the presence of abscess formation before surgical intervention.

A commonly used adjunctive diagnostic method is the “bedside finger test,” ⁸ in which the fingers are used to examine the necrotic fascia layer before and during the operation. The lesion site can easily separate without resistance, and bleeding is often accompanied by foul-smelling, dishwater pus with a fluid-like or purulent discharge. The above symptoms and the findings of the “bedside finger test” are important aspects to facilitate diagnosis. In all the cases we reviewed, the findings of the “bedside finger test” were determined to be positive.

Bacterial culture results can facilitate diagnosis. ¹³ However, since nonsteroidal anti-inflammatory drugs, hormones, or antibiotics may be used in the early stage, some patients do not show fever and even test negative in the bacterial culture. In such cases, NF cannot be excluded, and clinical symptoms should be carefully evaluated. ¹⁴ In primary hospitals, all patients were treated with one or more antibiotics, and 10 of them were confirmed to show negative bacterial cultures. However, after reviewing the entire course and outcome of treatment, these cases still showed NF.

Wong et al ¹² were the first to propose that LRINEC can be used to facilitate the diagnosis of NF and its differentiation from other soft tissue infections. The LRINEC evaluates the white blood cell count and the CRP, hemoglobin, blood sodium, serum creatinine, and blood glucose levels. The highest score is 13 points, and 50%-75% of the patients with suspected NF had scores of >6 points, while >75% of the patients had scores of >8 points. However, two recent studies^15,16 reported that LRINEC had low accuracy and poor sensitivity. Specifically, the sensitivity of NF was 43% in patients with a score >6 points and 27% in those with a score of >8 points. In our retrospective study, 16 patients (44.44%) had scores of ≥8 points, and only one patient did not have diabetes. The literature^17,18 also suggests the inconsistencies in the reported usefulness of LRINEC may have been due to differences in the characteristics of the study population, infection site, and type of infection, and various factors may have led to errors in the results. In patients diagnosed with diabetes or highly suspected to show diabetes, laboratory findings are also nonspecific. ¹⁹ Nevertheless, the use of laboratory tests, MRI,^20,21 CT, and B-mode ultrasound can assist in diagnosing NF.^22–24Delayed surgical debridement has been generally suggested to be associated with increased mortality. Meta-analyses have recommended that surgery should be performed within 12 h of the suspected diagnosis.^10,25 In a comparison of early (<12 h) initial debridement versus late (≥12 h) initial debridement, the overall mortality rate was 14% in the early group and 25.8% in the late group. In our study, among patients at the early stage of the disease, those who underwent simple surgical drainage in the primary hospital required fewer operations and a shorter hospital stay than patients who did not receive any surgical treatment before treatment at our department. Due to the limited number of cases, the differences between the two groups are not significant. Nevertheless, these findings suggest the importance of early surgical management.

From a clinical perspective, thorough debridement may have more potent therapeutic effects than antibiotic therapy. ²⁶ In our study, all patients who could be traced back to the primary hospital or emergency department received antibiotic treatment. However, necrosis still progressed rapidly, and the patients showed higher scores for inflammatory indicators and LRINEC. However, when we compared the preoperative levels of inflammatory indicators with those recorded after debridement, the results showed a significant difference. Therefore, we believe that debridement is critically important for NF treatment.

Notably, diabetes has been recognized as an important risk and aggravating factor for NF.^27,28 Cheng et al ²⁹ conducted a retrospective study of 165 patients with NF and found that patients with diabetes were more prone to NF because they had greater susceptibility to bacterial infections and because their amputation rate (28.6%) was significantly higher than that of patients with NF but without diabetes (12.6%, P < .05). The results of laboratory tests in this group showed that patients with diabetes had higher levels of inflammatory markers than those without diabetes. These results reiterate that patients with diabetes are more susceptible to infections and NF development in general. ³⁰

In our therapeutic experience, NPWT is a useful and highly efficient treatment for NF. The advantages of NPWT include persistent and complete drainage, effective infection control, reduced postoperative secondary necrosis, formation of new granulomatous tissue, and significant short-term improvement in wound status.^31–33 Some researchers have also suggested that NPWT reduces the incidence of septic shock, and organ failure.^31,34,35

Clinical diagnosis is particularly important in the treatment of NF, and a comprehensive diagnosis should be made in conjunction with the results of laboratory tests. Drainage during the early stages plays an important role in relieving NF progression. Complete debridement is a key treatment strategy for NF and reduces the incidence of severe complications and the mortality rate. In contrast, NPWT can also control infection and promote wound healing. Diabetes is an adverse factor in patients with NF, and much attention should be focused on diabetes in clinical practice. As a single-center, short-term case study, our sample size is small; however, given the lack of mortality, the experiences of these patients can provide a clinical reference for the diagnosis and treatment of surgical soft tissue infections.