Microbiology and surgical indicators of necrotizing fasciitis in a tertiary hospital of southwest Taiwan

International Journal of Infectious Diseases 16 (2012) e159–e165

Contents lists available at SciVerse ScienceDirect

International Journal of Infectious Diseases journal homepage: www.elsevier.com/locate/ijid

Microbiology and surgical indicators of necrotizing fasciitis in a tertiary hospital of southwest Taiwan Yao-Hung Tsai a,b,*, Kuo-Chin Huang a,b, Shih-Hsun Shen a,b, Wei-Hsiu Hsu a,b, Kuo-Ti Peng a,b, Tsung-Jen Huang a,b a b

Department of Orthopaedic Surgery, Chang Gung Memorial Hospital at Chia-Yi, 6, West Sec, Chia-Pu Rd, Putz City, Chia-Yi, 613, Taiwan College of Medicine, Chang Gung University at Taoyuan, Taiwan

A R T I C L E I N F O

S U M M A R Y

Article history: Received 28 June 2011 Received in revised form 12 October 2011

Background: Necrotizing fasciitis is a true surgical emergency. This study was undertaken to determine whether clinical indicators could be used to initiate early surgery, and to compare the characteristics observed on initial examination of necrotizing fasciitis in patients who died and those who survived. Methods: We retrospectively reviewed the medical records of 143 patients with surgically confirmed necrotizing fasciitis of the extremities over a period of 3.5 years at a tertiary hospital of southwest Taiwan. Differences in mortality, patient characteristics, laboratory findings, and hospital course were compared between patients who died and those who survived, and between patients with Grampositive infections and those with Gram-negative infections. Results: A patient with a fungal infection died. Nine of the 58 patients in the Gram-positive group (15.5%) and 12 of the 60 cases in the Gram-negative group (20%) died. Hence a total of 22 patients died, giving a mortality rate of 15.4%. Hypotension, lower counts of total and segmented leukocytes, higher counts of banded leukocytes, and lower levels of serum albumin were significantly associated with mortality. Monomicrobial infections had a stake of 70.6%, and Vibrio spp were the predominant causative agents (26.6%). Conclusions: Hypotensive shock, severe hypoalbuminemia, and increased counts of banded leukocytes can be considered the clinical and laboratory risk indicators to initiate early surgery and to predict mortality for all types of necrotizing fasciitis. The clinical characteristics of Gram-negative infections were more fulminant than those of Gram-positive infections. ß 2011 International Society for Infectious Diseases. Published by Elsevier Ltd. All rights reserved.

Corresponding Editor: William Cameron, Ottawa, Canada Keywords: Necrotizing fasciitis Gram-positive Gram-negative Vibrio Ceftriaxone

1. Introduction Necrotizing fasciitis is a rapid life-threatening soft tissue infection and a true surgical emergency.1–5 The clinical features of necrotizing fasciitis, including hypotension, fever, hemorrhagic bullae, subcutaneous bleeding, purpura, necrosis, and gangrene, need to be diagnosed early at the time of presentation in the emergency room. Necrotizing fasciitis is categorized into four types: type I is mixed infection with aerobic and anaerobic bacteria (70–80% cases); type II is group A b-hemolytic Streptococcus spp and Staphylococcus aureus (20–30% cases); type III is Gramnegative monomicrobial necrotizing fasciitis, including those caused by Vibrio spp, Aeromonas spp, and Klebsiella spp; and type

* Corresponding author. Tel.: +886 5 3621000x2885; fax: +886 5 3623005. E-mail addresses: [email protected], [email protected] (Y.-H. Tsai).

IV is fungal infection.1–5 Emergent surgical fasciotomy and broadspectrum antibiotic therapy are the most effective management strategies to reduce the mortality rate of necrotizing fasciitis.1–7 In Europe and the USA, Gram-positive aerobic bacteria are the major causative organisms of necrotizing fasciitis. Necrotizing fasciitis caused by b-hemolytic Streptococcus spp is often associated with toxic shock syndrome and is reported to account for 43–58% of the mortality rate.5,8 S. aureus, a Gram-positive coccus, is the most common agent causing skin and soft tissue infections worldwide, and methicillin-resistant S. aureus (MRSA) has been reported to have a prevalence of 59–75% in soft tissue infections and 29–39% in necrotizing fasciitis in emergency departments.9–11 Our institution is situated on the western coast of southern Taiwan, and nearly 40 patients present with necrotizing fasciitis annually. Vibrio spp are the leading causative agents of necrotizing fasciitis and related fatality, as a result of patient occupational exposure to seawater or raw seafood.6,7,12,13 We have established that severe hypoalbuminemia, severe thrombocytopenia, and

1201-9712/$36.00 – see front matter ß 2011 International Society for Infectious Diseases. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijid.2011.11.001

e160

Y.-H. Tsai et al. / International Journal of Infectious Diseases 16 (2012) e159–e165

increased banded leukocyte counts are laboratory risk indicators of Vibrio necrotizing fasciitis that could be used to initiate early surgery and predict mortality.14 We have also established a treatment strategy that includes emergency fasciotomy or amputation, antibiotic therapy with a third-generation cephalosporin plus tetracycline, and admission to the intensive care unit (ICU) for patients suspected of having fulminant necrotizing fasciitis. With this strategy for treating Vibrio necrotizing fasciitis, the mortality rate of 32% observed during the 4-year period from 2002 to 2005 was reduced to a mortality rate of 19% during the 3year period from 2006 to 2008.14 Nevertheless, it is yet to be determined whether these laboratory indicators and treatment strategies for Vibrio necrotizing fasciitis are suitable for treating other types of necrotizing fasciitis. This retrospective study was undertaken to determine whether clinical and laboratory risk indicators, such as hypotension, fever, total leukocyte counts, counts of banded and segmented forms of leukocytes, platelet counts, and serum albumin levels, can be used to initiate early surgery and predict mortality for all types of necrotizing fasciitis, and to investigate the association between mortality and different organisms. We also compared the clinical characteristics observed on initial examination of necrotizing fasciitis caused by Gram-positive and Gram-negative pathogens, and identified the risk factors that influence the outcomes in these two groups. 2. Materials and methods We retrospectively reviewed the medical records of 143 patients with surgically confirmed necrotizing fasciitis of the extremities admitted between January 2007 and June 2010. Patients who did not undergo surgery were excluded. Of the 143 patients, 110 were men and 33 were women, and their mean age was 64.5 years (range 15–90 years). The most common symptoms were pain and swelling of the involved limbs with edematous, patchy, erythematous, and hemorrhagic bullous skin lesions at the time of admission to the emergency room or at the time of consultation on the ward. Broad-spectrum antibiotics were initially administered to all the patients, and an emergency fasciotomy or immediate limb amputation was performed in cases where necrotizing fasciitis was diagnosed. Age, gender, co-morbidities, signs and symptoms, affected sites, microbiological findings, predisposing factors, laboratory findings at the time of admission, the duration between contact and admission, the interval between diagnosis and first surgery, length of stay, and clinical outcomes were reviewed for each patient. We categorized the patients into two groups: those who died during hospitalization and those who survived. Specimens obtained from the surgical wounds or blood were classified as Gram-positive bacteria, Gram-negative bacteria, fungus, and ‘no growth’, according to the microbiological findings. We further categorized those patients with positive bacterial cultures into two groups: the Gram-positive group and the Gramnegative group. Differences in mortality, patient characteristics, clinical presentation, underlying chronic diseases, infection sites, first operative procedure, laboratory data, and hospital course were compared between the death and survival groups of patients and between the Gram-positive and Gram-negative groups. We used the Wilcoxon rank-sum test for continuous variables and the Chi-square test and Fisher’s exact test for categorical variables to examine significant relationships between these factors among the groups. A p-value of <0.05 was considered significant. 3. Results Twenty-two patients died, resulting in a mortality rate of 15.4% (Table 1). Bacterial growth was absent in 24 patients. A

Table 1 Patient characteristics at first consultation in the emergency room and patient treatment compared between patients who died and those who survived Variable

Died (n = 22)

Age (years), mean Sex, n Male Female Timing from symptoms to presentation in ER (days), mean Timing from treatment in ER to first operation (h), mean Underlying chronic disease, n Hepatic dysfunction and diabetes mellitus Hepatic dysfunction alone Diabetes mellitus alone Diabetes mellitus with others Steroid intake Chronic renal insufficiency Heart disease Hepatitis with steroid intake Gout Gout with steroid intake Other Nil Wound location, n Upper extremity Lower extremity First operation, n Fasciotomy Amputation Final operation after first fasciotomy, n Amputation Split-thickness skin graft Flap Debridement Without secondary operation Microbiology results, n No growth (16.8%) With growth Monomicrobial (70.6%) Polymicrobial (12.6%) Hospital days, mean

71.81

63.17

0.012a

16 6 3

94 27 3.18

0.591

4.45

Survived (n = 121)

6.5

2 7 1 6 1 2 1 1 0 1 0 0

12 28 26 7 7 3 6 5 8 3 5 11

6 16

43 78

21 1 21 3 2 0 7 9

118 3 118 9 48 7 45 9

0

24

p-Value

0.517 0.16

0.452

0.491

20 2 24.54

81 16 34.23

0.307 0.001a

ER, emergency room. a p < 0.05; significant difference.

monomicrobial infection was confirmed in 101 patients and a polymicrobial infection was confirmed in 18 patients. Cases in which the pathogen was present in polymicrobial cultures were included in this study only if heavy and predominant growth was observed. Culture findings confirmed that the cause of infection was Gram-positive bacteria in 58 patients, Gram-negative bacteria in 60 patients, and fungus in one patient. Broad-spectrum antibiotics were initially administered to patients with suspected necrotizing fasciitis in the emergency room (Table 2). These antibiotics were continued after surgery and were then changed to antibiotics specifically targeting cultured bacteria a few days later. 3.1. Patient characteristics in the death and survival groups The group of patients who died included 16 men and 6 women, with a mean age of 71.8 years (range 46–90 years). The estimated period from injury to presentation at the emergency room ranged from 1 to 7 days (mean 3 days) prior to admission. The mean timeinterval from treatment in the emergency room to the first operation was 4.45 h (range 2–18 h). Six patients had upper limb skin lesions and 16 patients had lower limb skin lesions. Twenty-one patients initially underwent fasciotomy and debridement, and one patient underwent an immediate above-the-knee amputation because of severe skin

Y.-H. Tsai et al. / International Journal of Infectious Diseases 16 (2012) e159–e165

e161

Table 2 Initial usage of antibiotics in the emergency room Antibiotic

Gram-positive group

Gram-negative group

No growth

Fungus

Died

Survived

Died

Survived

Survived

Died

Ceftriaxone Ceftriaxone + tetracycline Ceftriaxone + vancomycin Ceftriaxone + gentamicin Ceftriaxone + clindamycin Oxacillin + gentamicin Penicillin + gentamicin Cefazolin + gentamicin Vancomycin + gentamicin Vancomycin Oxacillin Amoxicillin/clavulanate Total

4 0 3 1 0 0 0 0 1 0 0 0 9

16 2 2 5 0 17 1 3 0 1 0 2 49

7 3 2 0 0 0 0 0 0 0 0 0 12

26 11 2 3 1 2 0 1 0 0 0 2 48

12 8 1 0 2 0 0 0 0 0 1 0 24

0 1 0 0 0 0 0 0 0 0 0 0 1

lesions and sepsis at the time of presentation. Three patients underwent an above-the-knee amputation a few days after fasciotomy due to progressive uncontrolled sepsis. One hundred and twenty-one patients survived and were alive during the 6-month follow-up period. This group included 94 men and 27 women, with a mean age of 63.17 years (range 15–90 years). The estimated period from injury to presentation at the emergency room ranged from 1 to 21 days (mean 3.18 days) prior to admission. The mean time-interval from treatment in the emergency room to first operation was 6.5 h (range 1–24 h). Forty-three of the survivors had upper limb involvement and 78 had lower limb involvement. An initial fasciotomy with debridement was performed in 118 patients. Two patients underwent an immediate below-the-knee amputation, and one patient underwent an immediate above-the-knee amputation initially. Owing to progressive skin involvement after the fasciotomy, two patients underwent an above-the-knee amputation, six had above-theelbow amputations, and one had a finger amputation. Twentythree patients received hyperbaric oxygen therapy, and vacuumassisted closure devices for treating large wounds were used in 13 patients when they were stable and the infections were under control after leaving the ICU. 3.2. Comparison between the death and survival groups Gender, interval between symptoms and admission, interval between diagnosis of necrotizing fasciitis and first surgery, wound location, nature of first surgery, fever, counts of the lymphocyte forms of leukocytes, and platelet counts were similar between the two groups. Advanced age (p = 0.012) and a systolic blood pressure of 90 mmHg at the time of presentation to the emergency room (p = 0.000) were significantly associated with mortality. The patients who initially received ceftriaxone with/without other regimens had a significantly higher survival rate (Table 2). A body temperature of <38.5 8C was noted in 108 patients (75.5%). Patients who died had lower counts of total (p = 0.001) and segmented leukocytes (p = 0.01), higher counts of banded leukocytes (p = 0.000), and lower levels of serum albumin (p = 0.000) compared to the survivors. Those patients with a white blood cell count of 11  109 cells/l (p = 0.007), a segmented white blood cell count of 74% (p = 0.001), a >0% count of banded forms of white blood cells (p = 0.005), and a serum albumin level of 2.5 g/dl (p = 0.000) had significantly higher mortality (Table 3). 3.3. Microbiology and mortality The pathogens identified are listed in Table 4. The Gramnegative Vibrio spp were the most frequently identified pathogens, however patients with Klebsiella pneumoniae infections had the

Total

65 25 10 9 3 19 1 4 1 1 1 4 143

highest mortality rate (60%). Additional aerobic pathogens identified in the polymicrobial cultures were Streptococcus spp, coagulase-negative staphylococci (CoNS), Escherichia coli, Enterobacter cloacae, and Pseudomonas aeruginosa. The additional anaerobic pathogens identified in the polymicrobial infections were Clostridium spp in four patients, Peptostreptococcus anaerobius in two, Prevotella in one, and Bacteroides fragilis in one. The pathogen causing the fungal infection in the patient who died was identified from the wound as Penicillium species; the contact mechanism was unknown for this patient. Twenty-four patients without bacterial growth survived. Fourteen patients had a history of contact with raw seafood or seawater. One patient had a cutting injury, one had acquired an abrasion wound, one had a chronic ulcer, one had worked in dirty water, and one had an insect bite. Five patients did not recall any injuries. 3.4. Patient characteristics in the Gram-positive group The Gram-positive group included 41 men and 17 women (mean age 64.2 years). Nine patients died a mean of 25.1 days (range 7–88 days) after admission, and the mortality rate was 15.5%. The interval from injury to presentation at the emergency room ranged from 1 to 21 days (mean 3.89 days). The mean time interval between treatment in the emergency room and the first operation was 5.91 h (range 2–24 h). Two patients had both hepatitis B and C with diabetes mellitus. One patient who died had liver cirrhosis and diabetes mellitus. Eight patients had a history of hepatic dysfunction alone; four patients had hepatitis B and C, three patients had liver cirrhosis and hepatitis B or C, and one patient who had hepatitis C died. Fourteen patients had diabetes mellitus alone, and one of them died. One patient who had diabetes mellitus and steroid intake died. Two patients had both diabetes mellitus and gout, and two patients had both diabetes mellitus and chronic renal insufficiency. One patient had steroid intake alone. Three patients had chronic renal insufficiency, of whom two died; three patients had valvular heart disease, of whom one died. Six patients had gout alone. Three patients had both hepatitis and steroid intake, of whom one died, and one patient with both gout and steroid intake died. One patient had a history of urinary tract infection, one patient had psoriasis, and one patient had an HIV infection. Eight patients were healthy before the infection occurred. Nine patients (15.5%) were febrile (body temperature >38.5 8C) and 19 patients (32.8%) had a systolic blood pressure of 90 mmHg at presentation in the emergency room. Thirty-four patients were admitted to the ICU. The mean duration of hospital stay for the patients in the Gram-positive group was 33.3 days (range 6–90 days).

e162

Y.-H. Tsai et al. / International Journal of Infectious Diseases 16 (2012) e159–e165

Table 3 Clinical findings and laboratory data at first consultation in the emergency room compared between patients who died and those who survived No. of patients (%) (N = 143) Systolic blood pressure (mmHg) 90 >90 Body temperature (8C) 38.5 <38.5 White blood cell count ( 109/l) Mean 11 >11 Band forms (%) Mean =0 >0 Segmented forms (%) Mean 74 >74 Lymphocyte forms (%), mean Platelet count ( 109/l) Mean 15 >15 Albumin (g/dl) Mean <2.5 2.5 a

Died (n = 22), n

Survived (n = 121), n

p-Value 0.000a

56 (39.2) 87 (60.8)

18 4

38 83

35 (24.5) 108 (75.5)

3 19

32 89

38 (26.6) 105 (73.4)

10.9086  7.9701 11 11

16.0107  6.6676 27 94

0.001a 0.007a

28 (19.6) 115 (80.4)

12.45  9.53 0 22

5.03  7.11 28 93

0.000a 0.005a

70.61  18.67 11 11 8.02  5.41

80.2  10.14 22 99 7.74  4.36

0.01a 0.001a

0.388

33 (23.1) 110 (76.9)

73 (51.0) 70 (49.0)

89 (62.2) 54 (37.8)

148.318  103.394 13 9

0.904

171.830  91.652 60 61

1.84  0.39 21 1

0.204 0.412

0.000a 0.000a

2.51  0.63 68 53

p < 0.05; significant difference.

3.5. Patient characteristics in the Gram-negative group The Gram-negative group included 48 men and 12 women (mean age 64.8 years). Twelve patients died a mean 23 days (range 2–78 days) after admission, resulting in a mortality rate of 20%. The estimated period from contact or injury to presentation at the emergency room ranged from 1 to 7 days (mean 2.63 days) prior to admission. The mean time-interval from treatment in the emergency room to first operation was 6 h (range 2–24 h). Eight patients had both hepatic dysfunction and diabetes mellitus, and one of them died. Twenty-one patients had a history of hepatic dysfunction alone, such as liver cirrhosis, hepatitis B or C, and alcoholic liver disease, and six died. Seven patients had diabetes mellitus without hepatic disease. One

patient had diabetes mellitus with gout. Three patients who had diabetes mellitus with chronic renal insufficiency and one patient who had diabetes mellitus with myocardial infarction died. Five patients had steroid intake alone, of whom one died. One patient had chronic renal insufficiency, two had hypertension, one patient had valvular heart disease, three patients had both hepatitis and steroid intake, two had gout alone, and one patient had both gout and steroid intake. Two patients had been in good health. Twenty patients (33.3%) had a body temperature of >38.5 8C. Thirty-two patients (53.3%) were hypotensive with a systolic blood pressure of 90 mmHg. Forty-four patients were initially admitted to the ICU. The mean hospital stay for patients with a Gramnegative infection was 34.5 days (range 2–87 days).

Table 4 Microbiology and mortality in patients with necrotizing fasciitis Pathogens

No. of patients (%) (N = 143)

Died

Survived

Gram-positive aerobic pathogens Beta-hemolytic Streptococcus MRSA MSSA Viridans Streptococcus CoNS Gram-negative aerobic pathogens Vibrio vulnificus Vibrio cholerae non-O1 Klebsiella pneumoniae Aeromonas hydrophila Aeromonas sobria Pseudomonas aeruginosa Proteus mirabilis Escherichia coli Morganella morganii Chryseobacterium meningosepticum Fungus Penicillium sp No growth

58 18 18 14 1 7 60 36 2 5 4 3 3 1 3 2 1

(40.6) (12.6) (12.6) (9.8) (0.7) (4.9) (42.0) (25.2) (1.4) (3.5) (2.8) (2.1) (2.1) (0.7) (2.1) (1.4) (0.7)

9 3 5 0 0 1 12 4 1 3 2 1 1 0 0 0 0

49 15 13 14 1 6 48 32 1 2 2 2 2 1 3 2 1

1 (0.7) 24 (16.8)

1 0

0 24

Mortality rate of pathogen (%) 15.5 16.7 27.8 0 0 14.3 20 11.1 50 60 50 33.3 33.3 0 0 0 0 100 0

MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-sensitive Staphylococcus aureus; CoNS, coagulase-negative Staphylococcus. a p < 0.05; significant difference.

Y.-H. Tsai et al. / International Journal of Infectious Diseases 16 (2012) e159–e165

3.6. Comparison between the Gram-positive and Gram-negative groups Age, sex, mortality rate, interval between diagnosis and first surgery, body temperature of >38.5 8C, duration of ICU stay, days of hospitalization, total white blood cell counts, number of lymphocyte forms of white cells, and serum albumin levels did not differ significantly between the two groups (Tables 5 and 6). However, we found significant differences between the two groups with regard to the interval between contact and admission (p = 0.000), systolic blood pressure of 90 mmHg (p = 0.027), segmented white blood cell counts (p = 0.006), banded white cell counts (p = 0.000), and platelet counts (p = 0.000). There were significant differences between the survivors in the Gram-positive and Gram-negative groups with regard to systolic blood pressure of 90 mmHg (p = 0.037), body temperature of >38.5 8C (p = 0.048), interval between contact and admission (p = 0.01), banded white cell counts (p = 0.000), and platelet counts (p = 0.000). Also, the probability of Gram-negative infection was significantly higher in patients with hepatic dysfunction alone. 4. Discussion Necrotizing fasciitis has been reported to have a cumulative mortality rate of 34% (range 6–76%).15 Angoules et al. performed a systematic review of necrotizing fasciitis affecting upper and lower extremities and found a mortality rate of 21.9%.2 Thus, early recognition of necrotizing fasciitis followed by an immediate decision for surgical intervention is necessary in the emergency department to reduce mortality. The laboratory risk indicator for necrotizing fasciitis (LRINEC) score has often been described as a diagnostic tool to distinguish necrotizing fasciitis from non-necrotizing soft tissue infections.1,5,16,17 However, our previous study suggested that the LRINEC scoring system is inappropriate for determining the early management of patients with suspected Vibrio necrotizing fasciitis, which has been categorized as type III necrotizing fasciitis due to its fulminant clinical presentation on admission to the emergency room and high mortality rate (25–100%).14 We identified a systolic blood pressure of 90 mmHg, severe hypoalbuminemia, severe thrombocytopenia, and increased count of banded forms of leukocytes as the clinical and laboratory risk indicators of Vibrio necrotizing fasciitis that can be used to initiate early surgery and predict mortality.14 In this study, we used these positive predictors and their parameters to investigate all types of necrotizing fasciitis. We found that the patients who died had lower systolic blood pressure, lower total and segmented leukocyte counts, higher banded leukocyte counts, and lower serum albumin levels than the patients who survived. Therefore, we consider hypotensive shock, severe hypoalbuminemia, and increased count of banded leukocytes as the risk indicators that can be used to determine the timing of operative intervention and predict mortality for patients suspected of having fulminant necrotizing fasciitis. Although type I polymicrobial necrotizing fasciitis has been the cause of nearly 80% of necrotizing fasciitis in some studies, a significant increase in the incidence of monomicrobial necrotizing fasciitis has been reported recently (up to 60%).5,18–21 S. aureus and group A Streptococcus spp were the predominant pathogens causing necrotizing fasciitis in the USA and Europe; however, monomicrobial Gram-negative aerobic pathogens, such as Vibrio vulnificus, K. pneumoniae, Aeromonas hydrophila, and E. coli, were the most frequently isolated microorganisms in Asia.2,5,6,22–26 In our series, monomicrobial infections (70.6%) were found to be more common than in the previously reported studies, and Vibrio spp were the predominant pathogens (26.6%). Also, the patients

e163

Table 5 Patient characteristics at first consultation and patient treatment compared between the Gram-positive group and the Gram-negative group Variable

Grampositive group

Number 58 Age (years), mean 64.2 Sex, n Male 41 17 Female Mortality rate (%) 15.5 Died, n 9 Survived, n 49 Timing from contact to 3.89  3.05 presentation at ER (days), mean  SD Died 3.33  2.23 Survived 4  3.19 Timing from treatment in 5.91  4.84 ER to first operation (h), mean  SD 3.33  0.86 Died 6.38  5.11 Survived Underlying chronic disease, n (n died) Hepatic dysfunction and 3 (1) diabetes mellitus 8 (1) Hepatic dysfunction alone Diabetes mellitus alone 14 (1) Diabetes mellitus with 5 (1) other conditions 1 (0) Steroid intake Chronic renal insufficiency 3 (2) Heart disease 3 (1) Hepatitis with steroid intake 3 (1) Gout 6 (0) Gout with steroid intake 1 (1) Other 3 (0) Nil 8 (0) Contact mechanism, n (n died) Seawater or seafood contact 3 (0) Cutting or nail injury 2 (0) Bamboo or wood injury 1 (0) Abrasion 28 (4) Previous surgical 3 (2) wound infection 5 (0) Chronic ulcer Snake or insert bite 0 Dirty water or drain contact 2 (0) Unknown mechanism 14 (3) Wound location, n (n died) Upper extremity 21 (4) Lower extremity 37 (5) First operation, n (n died) Fasciotomy 58 (9) Amputation 0 (0) Final operation after first 58 fasciotomy, n (n died) 5 (1) Amputation Split-thickness skin graft 21 (1) Flap 2 (0) Debridement 24 (4) Without secondary operation 6 (3) Hospital days, mean  SD 33.34  21.2 Died 25.11  25.95 Survived 34.86  20.16 Positive culture, n (n died) Wound 43 (6) Blood 1 (0) Wound and blood 14 (3) ER, emergency room; SD, standard deviation. a p < 0.05; significant difference.

Gramnegative group 60 64.8 48 12 20 12 48 2.63  1.87

p-Value

0.836 0.287

0.156

0.000a

2.66  2.23 2.63  1.81 6  5.05

0.286 0.01a 0.924

5.33  4.33 6.17  5.24

0.19 0.834

8 (1)

0.127

21 (6) 7 (0) 5 (4) 5 1 3 3 2 3 0 2

(1) (0) (0) (0) (0) (0)

39 3 2 10 0

(4) (0) (1) (4) (0)

1 1 1 3

(0) (0) (0) (3)

0.007a 0.077

(0)

19 (2) 41 (10) 57 (11) 3 (1) 57 7 (2) 22 (1) 3 (0) 15 (2) 10 (6) 34.55  19.96 23  23.65 37.44  18.07 35 (8) 5 (0) 20 (4)

0.75 0.83 0.51

2.49  0.68 2.5  0.59 0.777 213.983  95.258 134.646  76.108 0.000a 7.81  4.26 7.16  4.3 0.36 81.41  12.83 79.07  8.21 0.287 3.82  8.57 6.5  6.19 0.000a 17.071  6.469 15.122  7.242 0.126 37 27 12 21 0.037a

BP, blood pressure; ICU, intensive care unit; WBC, white blood cell; SD, standard deviation. a p < 0.05; significant difference.

24 16 25 32 0.087 41 30

2 1 7 11 0.379

8 18 0.048a

194.444  128.760 105.083  59.949 0.081 7.89  4.48 8.21  6.41 0.942 78.89  7.70 63.21  22.08 0.081 9.11  7.47 15.75  10.18 0.154 10.232  5.749 11.167  9.797 0.915 0 0 9 12 1 8 10

7.82  4.25 7.37  4.75 0.36 81.02  12.16 75.9  13.62 0.006a 4.63  8.57 8.35  7.98 0.000a 16.010  6.791 14.331  7.887 0.131 24 16 34 44 0.119 49 40 9 20 0.053 39 28

No Yes <38.5 >38.5 >90

19 32 0.027a

All patients (N = 118) Gram-positive group (n = 58) Gram-negative group (n = 60) p-Value Died (n = 21) Gram-positive group (n = 9) Gram-negative group (n = 12) p-Value Survived (n = 97) Gram-positive group (n = 49) Gram-negative group (n = 48) p-Value

90

1 2 0.725

2.37  0.69 2.37  0.61 0.812 210.952  100.096 128.733  73.666 0.000a

Platelet count ( 109/l), mean  SD Lymphocyte forms (%), mean  SD Segmented forms (%), mean  SD Band forms (%), mean  SD WBC count ( 109/l), mean  SD ICU stay, n Body temperature (8C), n Systolic BP (mmHg), n

Table 6 Clinical findings and laboratory data at first consultation in the ER compared between the Gram-positive group and the Gram-negative group

1.72  0.31 1.86  0.38 0.518

Y.-H. Tsai et al. / International Journal of Infectious Diseases 16 (2012) e159–e165 Albumin (g/dl), mean  SD

e164

with K. pneumoniae infections had the highest mortality rate, which was 60%. In a previous report, we found the clinical characteristics of V. vulnificus infections were more rapidly progressive and fulminant than those of the S. aureus infections.27 In addition, for patients with V. vulnificus infections there was a significant association with hepatic dysfunction, whereas for patients with S. aureus infections there was a significant association with diabetes mellitus.27 In this study, the intervals between contact and admission to the emergency room for the Gram-negative group were significantly shorter than those for the Gram-positive group. Patients with infections of Gram-negative organisms significantly differed with regard to hypotensive shock, higher counts of banded leukocyte forms, and decreased platelet counts compared to patients with infections of Gram-positive organisms. Although we performed immediate surgery within a mean interval of 6 h between diagnosis and the first operation, we observed that the clinical course of Gram-negative infection was more rapid and fulminant than that of Gram-positive infection. Patients with hepatic dysfunction, such as liver cirrhosis, hepatic cell carcinoma, hepatitis B or C, and alcoholic liver disease, were significantly more likely to have Gram-negative infections. Clindamycin plus imipenem, meropenem, ampicillin/sulbactam, tigecycline, and piperacillin/tazobactam have been recommended as the initial antibiotic regimen for treating necrotizing fasciitis.4,5,17,28 An initial combination of clindamycin and betalactam penicillin is recommended by many authors for the treatment of necrotizing fasciitis caused by group A Streptococcus spp and Clostridium species.4,5,17,28,29 S. aureus, especially MRSA, has been reported to be the most common cultured pathogen in the past decade.9–11 However, at our institution, only four patients had clostridial infections, and Vibrio spp were the leading cause of necrotizing fasciitis. We started third-generation cephalosporin plus tetracycline or gentamicin antibiotic therapy in the emergency room for patients with a history of contact with seawater or raw seafood. The clinical signs and symptoms of necrotizing fasciitis caused by these organisms are characteristically indistinguishable at the time of presentation, and it takes 3– 4 days to obtain the results of microbiological analysis and antimicrobial sensitivity of the specimens. In this study, we primarily used ceftriaxone to cover all types of necrotizing fasciitis in 112 patients, because the clinical course of Gramnegative infection seemed more rapid and fulminant than that of Gram-positive infection. The use of ceftriaxone with/without other regimens can be advised at the first suspicion of necrotizing fasciitis to cover Gram-positive, Gram-negative, and anaerobic organisms. There were several limitations to our study. First, we did not observe organism growth in the specimens from 24 patients. Because 14 of them had a history of contact with seawater or raw seafood, they were suspected of and treated for Vibrio necrotizing fasciitis. PCR can be considered for early detection and accurate confirmation of the pathogens. The second limitation of our study was that we did not use hyperbaric oxygen therapy in the initial stage of necrotizing fasciitis owing to the unstable vital signs and need for stay in the ICU. Some studies have reported good results from the use of hyperbaric oxygen therapy, and we often use it to enhance wound healing.1,5 In conclusion, hypotensive shock, severe hypoalbuminemia, and increased banded leukocyte forms can be considered as clinical and laboratory risk indicators to initiate early surgery and predict mortality for all types of necrotizing fasciitis. The clinical characteristics of Gram-negative infections were more rapid and fulminant than those of Gram-positive infection. The initial usage of ceftriaxone with/without other regimens can be advised at the first suspicion of necrotizing fasciitis.

Y.-H. Tsai et al. / International Journal of Infectious Diseases 16 (2012) e159–e165

Ethics issues: The study protocol was approved by Institutional Review Board of Chang Gung Medical Foundation. Conflict of interest: None of the authors reports a potential conflict of interest. The study was not supported financially by any source. References 1. Sarani B, Strong M, Pascual J, Schwab CW. Necrotizing fasciitis: current concepts and review of the literature. J Am Coll Surg 2009;208:279–88. 2. Angoules AG, Kontakis G, Drakoulakis E, Vrentzos G, Granick MS, Giannoudis PV. Necrotizing fasciitis of upper and lower limb: a systematic review. Injury 2007;38S:18–25. 3. Anaya DA, McMahon K, Nathens AB, Sullivan SR, Foy H, Bulger E. Predictors of mortality and limb loss in necrotizing soft tissue infections. Arch Surg 2005;140:151–7. 4. Bellapianta JM, Ljungquist K, Tobin E, Uhl R. Necrotizing fasciitis. J Am Acad Orthop Surg 2009;17:174–82. 5. Morgan MS. Diagnosis and management of necrotizing fasciitis: a multiparametric approach. J Hosp Infect 2010;75:249–57. 6. Tsai YH, Hsu RW, Huang TJ, Hsu WH, Huang KC, Li YY, et al. Necrotizing soft tissue infections and sepsis caused by Vibrio vulnificus compared with those caused by Aeromonas species. J Bone Joint Surg Am 2007;89:631–6. 7. Huang KC, Hsieh PH, Huang KC, Tsai YH. Vibrio necrotizing soft-tissue infection of the upper extremity: factors predictive of amputation and death. J Infect 2008;57:290–7. 8. Haywood CT, McGeer A, Low DE. Clinical experience with 20 cases of group A Streptococcus necrotizing fasciitis and myonecrosis: 1995 to 1997. Plast Reconstr Surg 1999;103:1567–73. 9. Lee TC, Carrick MM, Scott BG, Hodges JC, Pham HQ. Incidence and clinical characteristics of methicillin-resistant Staphylococcus aureus necrotizing fasciitis in a large urban hospital. Am J Surg 2007;194:809–13. 10. Miller LG, Perdreau-Remington F, Rieg G, Mehdi S, Perlroth J, Bayer AS, et al. Necrotizing fasciitis caused by community-associated methicillin-resistant Staphylococcus aureus in Los Angeles. N Engl J Med 2005;352:1445–53. 11. Moran GJ, Krishnadasn A, Gorwitz RJ, Fosheim GE, McDougal LK, Carey RB, et al. Methicillin-resistant S. aureus infections among patients in the emergency department. N Engl J Med 2006;355:666–74. 12. Tsai YH, Hsu RW, Hung KC, Chen CH, Cheng CC, Peng KT, et al. Systemic Vibrio infection presenting as necrotizing fasciitis and sepsis—a series of thirteen cases. J Bone Joint Surg Am 2004;86:2497–502.

e165

13. Tsai YH, Huang TJ, Hsu RW, Weng YJ, Hsu WH, Huang KC, et al. Necrotizing softtissue infections and primary sepsis caused by Vibrio vulnificus and Vibrio cholerae non-O1. J Trauma 2009;66:899–905. 14. Tsai YH, Hsu RW, Huang KC, Huang TJ. Laboratory indicators for early detection and surgical treatment of Vibrio necrotizing fasciitis. Clin Orthop Relat Res 2010;468:2230–7. 15. McHenry CR, Piotrowski JJ, Petrinic D, Malangoni MA. Determinants of mortality for necrotizing soft-tissue infections. Ann Surg 1995;221:558–63. 16. Wong CH, Khin LW, Heng KS, Tan KC, Low CO. The LRINEC (laboratory risk indicator for necrotizing fasciitis) score: a tool for distinguishing necrotizing fasciitis from other soft tissue infections. Crit Care Med 2004;32:1535–41. 17. Anaya DA, Dellinger EP. Necrotizing soft-tissue infection: diagnosis and management. Clin Infect Dis 2007;44:705–10. 18. Anaya DA, Bulger EM, Kwon YS, Kao LS, Evans H, Nathens AB. Predicting death in necrotizing soft tissue infections: a clinical score. Surg Infect 2009;10:517–22. 19. Bair MJ, Chi H, Wang WS, Hsiao YC, Chiang RA, Chang KY. Necrotizing fasciitis in southeast Taiwan: clinical features, microbiology, and prognosis. Int J Infect Dis 2009;13:255–60. 20. Frazee BW, Fee C, Lynn J, Wang R, Bostrom A, Hargis C, et al. Communityacquired necrotizing soft tissue infections: a review of 122 cases presenting to a single emergency department over 12 years. J Emerg Med 2008;34:139–46. 21. Oncul O, Erenoglu C, Top C, Kucukardali Y, Karabudak O, Kurt Y, et al. Necrotizing fasciitis: a life-threatening clinical disorder in uncontrolled type 2 diabetic patients. Diabetes Res Clin Pract 2008;80:218–23. 22. Brook I, Frazier EH. Clinical and microbiological features of necrotizing fasciitis. J Clin Microbiol 1995;33:2382–7. 23. Olilivie CM, Miclau T. Necrotizing soft tissue infections of the extremities and back. Clin Orthop Relat Res 2006;447:179–86. 24. Cheng NC, Tai HC, Tang YB, Chang SC, Wang JT. Necrotizing fasciitis: clinical features in patients with liver cirrhosis. Br J Plast Surg 2005;58:702–7. 25. Lee CC, Chi CH, Lee NY, Lee HC, Chen CL, Chen PL, et al. Necrotizing fasciitis in patients with liver cirrhosis: predominance of monomicrobial Gram-negative bacillary infections. Diagn Microbiol Infect Dis 2008;62:219–25. 26. Salvador VB, San Juan MD, Salisi JA, Consunji RJ. Clinical and microbiological spectrum of necrotizing fasciitis in surgical patients at a Philippine university medical centre. Asian J Surg 2010;33:51–8. 27. Tsai YH, Hsu RW, Hung KC, Huang TJ. The comparison of necrotizing fasciitis and sepsis caused by Vibrio vulnificus and Staphylococcus aureus. J Bone Joint Surg Am 2011;93:274–84. 28. Rogers RL, Perkins J. Skin and soft tissue infections. Prim Care 2006;33:697–710. 29. Cainzos M, Gonzalez-Rodriguez FJ. Necrotizing soft tissue infections. Curr Opin Crit Care 2007;13:433–9.