Monomicrobial Necrotizing Fasciitis Caused by Aeromonas hydrophila and Klebsiella pneumoniae.

OBJECTIVE: To compare specific characteristics and clinical outcomes of monomicrobial necrotizing fasciitis caused by Aeromonashydrophila and Klebsiella pneumoniae.
MATERIAL AND METHODS: Cases of monomicrobial necrotizing fasciitis caused by A. hydrophila (n = 11) and K. pneumoniae (n = 7) over an 8-year period were retrospectively reviewed. Differences in mortality, patient characteristics, clinical presentations, and laboratory data were compared between the A. hydrophila and the K. pneumoniae groups.
RESULTS: The clinical signs and symptoms at the time of presentation did not differ significantly (p > 0.05) between the two groups. The A. hydrophila group had a significantly shorter interval between contact and admission (1.55 ± 0.52 vs. 5.14 ± 2.12 days, p < 0.001) and significant lower total white blood cell counts (10,245 ± 5,828 vs. 19,014 ± 11,370 cells/mm(3), p < 0.045) than the K. pneumoniae group in the emergency room. Hepatic dysfunction was associated with mortality in patients with A. hydrophila infection, while diabetes mellitus was associated with mortality in patients with K. pneumoniae infection. Overall, 5 (45.5%) patients in the A. hydrophila group and 3 (42.8%) in the K. pneumoniae group died.
CONCLUSION: The initial clinical course of A. hydrophila monomicrobial necrotizing fasciitis was characterized by more rapidly progressive disease than that of the K. pneumoniae infection. Patients with hepatic dysfunction and necrotizing fasciitis should be suspected of having A. hydrophila infection, and diabetic patients with necrotizing fasciitis should be suspected of having K. pneumoniae infection initially.

Introduction

Necrotizing fasciitis is a rapidly progressive, life-threatening soft-tissue infection that is a true medical and surgical emergency needing early diagnosis, emergent surgical debridement, and broad-spectrum antibiotic therapy when patients present to the emergency department [1,2,3,4]. Although necrotizing fasciitis is often caused by polymicrobial infection, the prevalence of monomicrobial necrotizing fasciitis has recently been reported to be as high as 60–80% [4,5,6,7]. Our previous study revealed that the clinical course of necrotizing fasciitis caused by Gram-negative microorganisms was more rapid and fulminant than that of Gram-positive infection, and Gram-negative aerobic pathogens, such as Vibrio vulnificus, Klebsiella pneumoniae, Aeromonas hydrophila, and Escherichia coli, were the most frequently isolated microorganisms causing necrotizing fasciitis [4].

Aeromonas spp. are Gram-negative bacilli that thrive in aquatic environments, especially in sewage, fresh or brackish water, soil, tap water, and nonfecal organic materials [8,9,10,11]. Necrotizing fasciitis caused by A. hydrophila often occurs after soft-tissue trauma with associated exposure to contaminated water or nonfecal organic materials and can produce skin lesions similar to those observed in necrotizing fasciitis caused by Vibrio species [8,9,10,11,12]. A. hydrophila is frequently associated with polymicrobial infections and can cause synergistic necrotizing fasciitis in patients coinfected with Clostridium species or Gram-negative bacilli, of which Klebsiella spp. have recently been reported to be the most common copathogens [11,12,13,14].

K. pneumoniae, a member of the Enterobacteriaceae, is a common cause of Gram-negative bacteremia and has frequently been described as a facultative organism in polymicrobial necrotizing fasciitis [15,16]. Monomicrobial necrotizing fasciitis caused by K. pneumoniae is extremely rare in the Western hemisphere [16], but community-acquired infections have been increasingly reported in the past decade, especially in the Asian regions, including Taiwan, Hong Kong, Japan, Singapore, and Malaysia, with fatality rates of nearly 50% [16,17,18,19,20,21].

Our previous studies [4,8,22,23] revealed that mortality rates in patients with Vibrio necrotizing fasciitis have decreased from 38 to 13% due to an effective program that includes early recognition, application of a treatment algorithm for emergency fasciotomy or amputation, treatment with a third-generation cephalosporin plus tetracycline or gentamicin, and an intensive unit care. Patients with Vibrio necrotizing fasciitis can be identified early because of a history of contact with seawater or raw seafood. However, patients with A. hydrophila and K. pneumoniae necrotizing fasciitis still have the highest mortality rates (50 and 60%, respectively) even after early diagnosis and surgical intervention in the emergency room [4]. Monomicrobial necrotizing fasciitis caused by A. hydrophila and K. pneumoniae are rarely reported [12,17,18]. To our knowledge, no publication or literature review has to date described or compared these two causative pathogens of fatal monomicrobial necrotizing fasciitis. Therefore, the purpose of this study was to compare the initial clinical features of A. hydrophila and K. pneumoniae necrotizing fasciitis, and the risk factors related to the outcomes.

Subjects and Methods

Study Design and Patient Selection

We reviewed the medical records of 33 patients with surgically confirmed necrotizing fasciitis caused by A. hydrophila and K. pneumoniae who were admitted to the emergency department of our hospital from June 2004 to December 2012. The enrolled patients were categorized into 2 groups: an A. hydrophila group and a K. pneumoniae group. Eight patients with A. hydrophila and 7 patients with K. pneumoniae infection who had confirmed polymicrobial necrotizing fasciitis were excluded (these patients survived after surgery).

Broad-spectrum antibiotics were initially administered to all patients, and excisional debridement of the necrotic fascia or immediate limb amputation was performed in all 33 patients diagnosed with necrotizing fasciitis.

Microbiology Laboratory Procedures

The identification of A. hydrophila and K. pneumoniae was based on standard phenotypic tests used in clinical microbiology laboratories, and the identification of A. hydrophila was further confirmed using the Api20E test (BioMerieux, Marcy l'Etoile, France). The antimicrobial susceptibility of A. hydrophila and K. pneumoniae was evaluated in the hospital microbiology laboratory using the standard disk diffusion technique at our institute. The susceptibility interpretative criteria for Enterobacteriaceae and Aeromonas spp. in our microbiological laboratory were established as recommended by the Clinical and Laboratory Standards Institute [24].

Clinical Assessment

Age, gender, comorbidities, signs and symptoms, infection site, results of bacteriological tests, predisposing factors, laboratory findings at the time of admission, interval between contact and admission, interval between diagnosis and first surgery, length of stay, and clinical outcomes were reviewed for each patient.

Differences in mortality, patient characteristics, clinical presentation, underlying chronic diseases, infection site, first operative procedure, laboratory data and hospital course were compared between the A. hydrophila group and the K. pneumoniae group.

Statistical Analysis

Statistical analyses were performed using SPSS version 12.0 statistical software (SPSS, Chicago, Ill., USA). Student's t test was used for continuous variables and Fisher's exact test was used for categorical variables to examine significant relationships between risk factors and outcomes in the two groups. p < 0.05 was considered statistically significant.

Results

Of the 18 patients, 15 were men and 3 were women, and the mean age was 63.6 ± 14.9 years (range 40–90). The most common complaints of patients with necrotizing fasciitis were hypotension and 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 in the hospital ward.

Culture findings confirmed that the cause of monomicrobial infection was A. hydrophila in 11 patients and K. pneumoniae in 8 patients. Eight patients died (5 in the A. hydrophila group and 3 in the K. pneumoniae group), resulting in an all-cause in-hospital mortality rate of 44.4%. A. hydrophila specimens were isolated from wounds in 8 cases, and from blood and wounds in 3 cases. K. pneumoniae was isolated from wounds in 5 cases and from blood and wounds in 2 cases. Broad-spectrum antibiotics were initially administered to these patients with necrotizing fasciitis in the emergency room. These antibiotics were continued after surgery and changed to antibiotics specifically targeting cultured bacteria a few days later.

Characteristics of the Patients in the A. hydrophila Group

Of the 11 patients in the A. hydrophila group, 9 were men and 2 were women, with a mean age of 65.9 ± 11.4 years (range 47–85). One patient reported having handled fish, 1 had acquired abrasion wounds while working, 1 had had contact with dirty water in a drain, and 1 had fallen into a ditch during a motor vehicle accident. Two patients had cutting wounds and 1 had been injured while working with bamboo on a farm. Four patients were farmers and did not recall any injuries. Five patients died (a mean of 6.20 ± 5.76 days after admission), resulting in an all-cause in-hospital mortality rate of 45.5% (tables 1, 2).

Table 1

Characteristics of patients with monomicrobial necrotizing fasciitis caused by A. hydrophila

Patient No.	Age, years	Gender	Chronic underlying diseases	Site	Interval A, days	Contact mechanism	Interval B, h	Operations (first, final)	Interval C, days	Result	Duration of hospitalization, days	Systolic blood pressure ≤90 mm Hg in the ER	Intensive care unit stay	Body temperature >38.5°C in the ER
1	78	M	HCC, LC	left leg	2	abrasion	18	Fas	0	death	16	Y	Y	Y
2	68	F	HB, steroid use	both legs	2	bamboo	2	Fas	0	death	2	Y	Y	N
3	55	M	LC, gout	left thigh	1	unclear	2	Fas	0	death	2	Y	Y	N
4	47	M	LC, HCC, DM	right leg	1	unclear	5	Fas	4	death	5	N	Y	N
5	58	M	Alcoholism	right leg	2	unclear	10	Fas, AK	4	death	6	Y	Y	N
6	64	M	Gout	left leg	1	cutting wound	10	Fas, STSG	20	discharge	63	N	N	N
7	77	M	Hypertension	right foot	2	drain	5	Fas	0	discharge	12	N	N	N
8	85	M	DM, CRI, steroid use	right leg	1	unclear	2	Fas, AK	11	discharge	90	Y	Y	N
9	73	M	DM, HB	left forearm	2	cutting wound	2	Fas, STSG	28	discharge	37	N	Y	Y
10	62	F	DM	left forearm	1	fish	2	Fas, Flap	20	discharge	14	N	Y	N
11	58	M	ESRD	right leg	2	fell into ditch	7	Fas, AK	10	discharge	24	N	Y	N
Mean	65.9				1.55		5.91		8.81		24.6

M = Male; F = female; HCC = hepatic cell carcinoma; LC = liver cirrhosis; HB = hepatitis B; CRI = chronic renal insufficiency; DM = diabetes mellitus; ESRD = end-stage renal disease; Interval A = time from contact to presentation in the emergency room; Interval B = time from the first consultation to the first operation; Interval C = time from the first operation to the final operation; Fas = fasciotomy; AK = above-the-knee amputation; STSG = split-thickness skin graft; Y = yes; N = no; ER = emergency room.

Table 2

Laboratory data of patients with monomicrobial necrotizing fasciitis caused by A. hydrophila

Patient No.	Result	White blood cell count, n/mm3	Banded forms, %	Segmented forms, %	Lymphocyte forms, %	Platelet count, n/mm3	Albumin, g/dl	Creatinine, μmol/l	Alanine aminotransferase, U/l	C-reactive protein, mg/l	Positive culture
1	death	16,500	16	78	2	151,000	2.2	1.61	30	43.5	B and W
2	death	5,600	15	67	6	150,000	1.7	2.08	45	93.4	W
3	death	3,100	27	57	8	57,000	2.9	2.12	23	65.77	B and W
4	death	5,100	34	45	8	64,000	1.6	1.4	121	30	W
5	death	2,600	9	61	6	90,000	1.7	2.8	189	100	W
6	discharge	11,200	4	88	4	158,000	3	1.7	27	26.2	B and W
7	discharge	18,800	2	81.5	7.5	131,000	2.5	1.14	25	115	W
8	discharge	7,200	4	71	18	140,000	2.1	2.33	35	27.5	W
9	discharge	12,100	9	85	3	170,000	2.4	1.49	20	140	W
10	discharge	13,500	1	80	13	79,000	2.5	0.88	72	49.5	W
11	discharge	17,000	12.5	71	9.5	150,000	1.5	5.7	19	89	W
Mean		10,245	12.1	71.3	7.73	121,818	2.19	2.11	55.1	70.9

W = Wound; B = blood.

The estimated period from exposure or injury to presentation at the emergency room was 1–2 days prior to admission. The mean interval from treatment in the emergency room to the first operation was 5.91 ± 5.1 h.

Two patients had upper-limb skin lesions and 9 had lower-limb skin lesions. All of the patients initially underwent fasciotomy and debridement. Three patients underwent above-the-knee amputation after a few days due to progressive skin involvement following fasciotomy. Two patients received skin grafts, 1 patient underwent flap reconstruction, 1 patient underwent debridement, and 1 received only wound care after the initial fasciotomy (fig. 1). Three patients did not undergo any surgery following fasciotomy; these patients died. Five (53.3%) patients were hypotensive with a systolic blood pressure ≤90 mm Hg. The mean hospital stay of the patients was 24.6 ± 28.3 days (range 2–90).

Fig. 1

A 73-year-old male with a history of hepatitis B and diabetes mellitus had left forearm and hand pain on the second day after a cutting injury. a Preoperative photographs of the left hand revealed skin erosion, vesicles, and subcutaneous bleeding. b After an emergency fasciotomy, a wound culture confirmed the presence of A. hydrophila. He received a skin graft on the thirtieth day after fasciotomy.

All of the A. hydrophila isolates were susceptible to ampicillin, amikacin, ceftazidime, ceftriaxone, cefuroxime, ciprofloxacin, gentamicin, imipenem, piperacillin, and tetracycline. Broad-spectrum antibiotic therapy with third-generation cephalosporin plus vancomycin or teicoplanin was initially administered to 9 patients, and antibiotic therapy with oxacillin plus gentamicin was administered to 2 patients (cases 7 and 9).

Characteristics of the Patients in the K. pneumoniae Group

Of the 7 patients in the K. pneumoniae group, 6 were men and 1 was a woman, with a mean age of 59.9 ± 19.6 years (range 40–90). Three patients had acquired abrasion wounds while working, and 1 had previous chronic ulcers. One patient had had contact with seawater, and 2 did not recall any injuries. Three patients died a mean of 27.7 ± 27.4 days after admission, and the all-cause in-hospital mortality rate was 42.8% (tables 3, 4).

Table 3

Characteristics of patients with monomicrobial necrotizing fasciitis caused by K. pneumoniae

Patient No.	Age, years	Gender	Chronic underlying diseases	Site	Interval A, days	Contact mechanism	Interval B, h	Operations (first, final)	Interval C, days	Result	Duration of hospitalization, days	Systolic blood pressure ≤90 mm Hg in the ER	Intensive care unit stay	Body temperature >38.5°C in the ER
1	84	M	DM, MI	right forearm	7	unknown	5	Fas	0	death	8	Y	Y	N
2	90	F	DM, gout, steroid use	left leg	2	abrasion	3	AK, debridement	12	death	16	Y	Y	N
3	58	M	DM, steroid use	left leg	4	abrasion	7	Fas, debridement	14	death	59	Y	Y	N
4	49	M	DM, LC, HC	right leg	7	seawater	12	Fas, STSG	30	discharge	44	N	Y	N
5	55	M	DM, HC, gout	right arm	6	abrasion	6	Fas, STSG	13	discharge	26	N	Y	N
6	43	M	DM	left leg	7	chronic ulcer	5	Fas, debridement	7	discharge	10	N	Y	N
7	40	M	DM, LC	right forearm	3	unknown	14	Fas, STSG	17	discharge	21	N	N	N
Mean	59.9				5.14		7.43		13.3		26.3

M = Male; F = female; DM = diabetes mellitus; MI = myocardial infarction; LC = liver cirrhosis; HC = hepatitis C; Interval A = time from contact to presentation to the emergency room; Interval B = time from the first consultation to the first operation; Interval C = time from the first operation to the final operation; AK = above-the-knee amputation; STSG = split-thickness skin graft; Y = yes; N = no; Fas = fasciotomy; ER = emergency room.

Table 4

Laboratory data of patients with monomicrobial necrotizing fasciitis caused by K. pneumoniae

Patient No.	Result	White blood cell count, n/mm3	Banded forms, %	Segmented forms, %	Lymphocyte forms, %	Platelet count, n/mm3	Albumin, g/dl	Creatinine, μmol/l	Alanine aminotransferase, U/l	C-reactive protein, mg/l	Positive culture
1	death	15,100	18	65.5	13.5	124,000	1.5	1.86	68	70	W
2	death	36,600	4	90	25	197,000	1.7	1.16	15	93	W
3	death	3,000	9	27.5	8	54,000	2	1.3	33	156	B and W
4	discharge	31,300	10.5	80	5	44,000	2.6	1.73	64	128	B and W
5	discharge	15,400	0	87	9	144,000	3	1.25	92	10	W
6	discharge	18,200	0	86	11	235,000	1.8	0.88	24	139	W
7	discharge	13,500	2	95	1	125,000	2	0.8	28	178	W
Mean		19,014	6.2	75.86	10.36	131,857.14	2.08	1.28	46.3	110.6

W = Wound; B = blood.

The interval from injury to presentation at the emergency room ranged from 2 to 7 days (mean 5.14 ± 2.12). The mean interval between treatment in the emergency room and the first operation was 7.43 ± 4.04 h.

Three patients had upper-limb skin lesions and 4 had lower-limb skin lesions. Six patients initially underwent fasciotomy with debridement and 1 patient underwent an immediate above-the-knee amputation due to progressive uncontrolled initial sepsis (fig. 2). Three patients received skin grafts, and 3 underwent debridement with direct closure. No Klebsiella patient was febrile. Three (14.5%) patients had a systolic blood pressure ≤90 mm Hg at presentation to the emergency room; these patients died. The mean duration of hospital stay was 26.3 ± 18.8 days (range 8–59).

Fig. 2

A 90-year-old female with a history of diabetes mellitus and chronic renal insufficiency had left lower leg pain for 2 days. a The left lower leg revealed patchy purpura and edema in the emergency room. b After fasciotomy, the lower leg showed yellowish pus accumulated in the fascia and muscular layer. Above-the-knee amputation was performed immediately. The cultured specimen confirmed K. pneumoniae; however, this patient died on the sixteenth day after admission owing to progressive septic shock and multiple organ failure.

All K. pneumoniae isolates were susceptible to amikacin, ceftazidime, ceftriaxone, cefuroxime, ciprofloxacin, aztreonam, gentamicin, and imipenem. Broad-spectrum antibiotics were administered initially in the emergency room to patients with K. pneumoniae infection; ceftriaxone alone was given to 3, cefazolin plus gentamicin to 2, and ceftriaxone plus vancomycin to 2 patients.

Comparison of the A. hydrophila and K. pneumoniae Groups

Age, sex, wound location, degree of hypotension, fever, nature of the first surgery, laboratory data and hospital course, and interval between diagnosis and the first surgery did not differ significantly (p > 0.05) between the two groups. However, the patients with Aeromonas infection had a significantly shorter interval between contact and admission (1.55 ± 0.52 vs. 5.14 ± 2.12 days, p < 0.001) and significantly lower total white blood cell counts (10,245 ± 5,828 vs. 19,014 ± 11,370 cells/mm3, p < 0.045) than patients with K. pneumoniae infection in the emergency room (tables 5, 6).

Table 5

Comparison between the A. hydrophila group and the K. pneumoniae group for characteristics at the first consultation and treatment

Variable	A. hydrophila group	K. pneumoniae group	p value
Patients, n	11	7
Age, years	65.9 ± 11.4	59.9 ± 19.6	0.42
Sex, n
Male	9	6	0.47
Female	2	1
Mortality rate, %	45.5	42.8	0.65
Death	5	3
Survival	6	4
Timing from contact to presentation to the ER, days	1.55 ± 0.52	5.14 ± 2.12	<0.001a
Death	1.6 ± 0.55	4.33 ± 2.52	0.049a
Survival	1.5 ± 0.55	5.75 ± 1.89	0.0007a
Timing from the first consultation to the first operation, h	5.91 ± 5.1	7.43 ± 4.04	0.52
Death	7.43 ± 6.77	5 ± 2	0.58
Survival	4.67 ± 3.33	9.25 ± 4.43	0.097
Underlying chronic disease, n
Hepatic dysfunction and DM	2 (1)	3 (0)
Hepatic dysfunction with or without other conditions	4 (4)	0 (0)	0.048b
Diabetes mellitus with or without other conditions	2 (0)	4 (3)	0.01c
End-stage renal disease	1 (0)	0 (0)
Hypertension	1 (0)	0 (0)
Gout	1 (0)	0 (0)
Wound location, n
Upper extremity	2 (0)	3 (1)
Lower extremity	9 (5)	4 (2)
Hospital stay, days	24.6 ± 28.3	26.3 ± 18.8	0.89
Death	6.20 ± 5.76	27.7 ± 27.4	0.13
Survival	40 ± 30.8	25.2 ± 14.2	0.4

Values are presented as means ± SD unless otherwise stated. Values in parentheses represent the number of deaths.

Mean p < 0.05 and the difference was significant.

Mean p < 0.05 and the difference was significant (association with mortality for hepatic dysfunction with or without other conditions, e.g. diabetes, gout, or steroid use).

Mean p < 0.05 and the difference was significant (association with diabetes mellitus with or without other conditions).

Table 6

Comparison between the A. hydrophila group and the K. pneumoniae group for laboratory data at the first consultation in the emergency room or ward

	White blood cell count, n/mm3	Banded forms, %	Segmented forms, %	Lymphocyte forms, %	Platelet count, n/mm3	Albumin, g/dl	Creatinine, μmol/l	Alanine aminotransferase, U/l	C-reactive protein, mg/l
All patients (n = 18)
A. hydrophila group (n = 11)	10,245 ± 5,828	12.1 ± 10.5	71.3 ± 13.1	7.73 ± 4.6	121,818 ± 41,080	2.19 ± 0.52	2.11 ± 1.31	55.1 ± 53.8	70.9 ± 39
K. pneumoniae group (n = 7)	19,014 ± 11,370	6.21 ± 6.64	75.9 ± 23.3	10.4 ± 7.62	131,857 ± 69,420	2.09 ± 0.53	1.28 ± 0.4	46.3 ± 28.5	110.6 ± 57.4
p value	0.045a	0.2	0.6	0.37	0.7	0.68	0.13	0.7	0.098

Mean p < 0.05 and the difference was significant.

Hepatic dysfunction with or without other conditions (diabetes, gout, or steroid use) was associated with mortality in patients with A. hydrophila infection (p = 0.048), while diabetes mellitus was associated with patients with mortality in K. pneumoniae infection (p = 0.01).

Discussion

In the present study, the interval between exposure and admission to the emergency room for patients with A. hydrophila infection was significantly shorter than that for patients with K. pneumoniae infection. The initial clinical course of A. hydrophila infection was characterized by more rapidly progressive disease than that of the K. pneumoniae infection. A. hydrophila infection often occurred in patients with liver cirrhosis, hepatitis, and hepatic malignancies, while K. pneumoniae was often associated with diabetes mellitus.

The more rapid progression of A. hydrophila compared to K. pneumoniae could be due to the fact that A. hydrophila can produce many virulence factors, including hemolysin, cytotoxin, aerolysin, enterotoxin, endotoxin, protease, adhesion, and lipases [12,14,25,26]. These factors are associated with extensive muscular necrosis and cause damage to the liver, kidneys, and pulmonary system, resulting in septic shock and multiple organ failure [25,26]. However, the virulence and pathogenicity of K. pneumoniae have been linked to its polysaccharide capsule envelope, and K1 and K2 are the most virulent serotypes [17,18,19,20]. These capsular serotypes and hypermucoviscosity phenotypes are highly resistant to serum killing and phagocytosis and have been implicated in the development of disseminated infection [18,27]. Necrotizing fasciitis caused by K. pneumoniae may be a consequence of transient bacteremia, or of gut bacterial translocation, followed by bacterial seeding at the extremities [17,18,27]. Equally important, Ko et al. [28] reported that A. hydrophila caused more rapid and intense local accumulation of inflammatory cells than K. pneumoniae and induced a more robust proinflammatory cytokine response when inoculated intramuscularly in mice [28].

The finding that patients with Aeromonas infection had a higher amount of banded leukocyte forms and significantly lower total white blood-cell counts than patients with K. pneumoniae infection in the emergency room in this study confirms the results of the animal study of Ko et al. [28]. They reported that the response of suppurative inflammation and aggregates of neutrophils might occur faster in mice with A. hydrophila infection than in those with K. pneumoniae infection [28]. We speculated that the response of suppurative inflammation might occur later with K. pneumoniae infection than that with A. hydrophila infection, and that the host can produce much more white blood cells to resist K. pneumoniae infection.

Finally, the finding of previous studies [10,12,14] that monomicrobial A. hydrophila infections such as necrotizing fasciitis and bacteremia have commonly been reported to be associated with liver cirrhosis and malignancy and impaired phagocytic activity of the reticuloendothelial system [10,12,14] confirms previous reports that cases of K. pneumoniae necrotizing fasciitis have shown a significant association with diabetes mellitus, present in 62.5–100% of such patients [17,18,27]. In this study, we also observed that hepatic dysfunction was associated with mortality in patients with A. hydrophila infection and that diabetes mellitus was associated with mortality in patients with K. pneumoniae infection. Although patients with K. pneumoniae necrotizing fasciitis have been reported to have a higher incidence of concomitant distant abscess, we did not find any distant abscesses in our cases [15,16,17,18,19,20].

In our previous studies, the clinical course of necrotizing fasciitis caused by Gram-negative microorganisms was more rapid and fulminant than that of Gram-positive infection, and the patients with diabetes mellitus alone had significant associations with Staphylococcus aureus necrotizing fasciitis [4,29]. The mortality rate of necrotizing fasciitis caused by S. aureus in diabetic patients was 18.7% (6/32) [29]. The 7 cases of K. pneumoniae necrotizing fasciitis were associated with diabetes mellitus, resulting in a mortality rate of 42.8% in this study. The clinical signs and symptoms of necrotizing fasciitis caused by K. pneumoniae and S. aureus in diabetic patients 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. Due to the high mortality rate and fulminant clinical course, K. pneumoniae infection as well as S. aureus infection should be initially suspected in diabetic patients with necrotizing fasciitis.

In this study, there were several limitations. First, we did not detect the capsular serotype K1/K2. Genotype K1 strains are significantly virulent and predominant in the East; however, we need to identify the roles of these genotypes in necrotizing fasciitis. The second limitation was that we could not reduce the mortality associated with monomicrobial A. hydrophila and K. pneumoniae necrotizing fasciitis, and we did not identify the causes of death even though we performed early diagnosis and surgical intervention.

Conclusion

Monomicrobial necrotizing fasciitis caused by A. hydrophila and K. pneumoniae had similar clinical courses and high mortality rates. The initial clinical course of A. hydrophila infection was characterized by more rapidly progressive disease than that of K. pneumoniae infection. Patients with hepatic dysfunction and necrotizing fasciitis should be suspected of having A. hydrophila infection, and diabetic patients with necrotizing fasciitis should be suspected of having K. pneumoniae infection initially if they have no history of contact with seawater or raw seafood.