Diagnosis,management and outcomes of necrotising soft tissue infection within a Plastic and Reconstructive Surgery Unit

Abstract

Background and aims

Necrotising soft tissue infection (NSTI) is an extremely serious condition that relies on a high index of suspicion, prompt diagnosis and emergent radical surgical treatment. We explored the presentation, management and outcomes of NSTI within our department. We also assessed the potential benefit of using risk predictor scoring system.

Methods

Retrospective review using departmental electronic database and hospital records.

Results

Twenty-four patients were treated for NSTI within our department between 2004 and 2010. Seventeen presented in our hospital to various surgical and medical teams. All patients presented with pain, swelling, erythema and tenderness at palpation. Only 40% of necrotising fasciitis and 28.6% of Fournier’s gangrene were diagnosed as NSTI at initial assessment. Average mean interval time from admission to primary surgery was 17.7 h and 4 h from diagnosis to primary surgery. There were four mortalities. The average risk predictor Laboratory Risk Indicator for Necrotising Fasciitis score was 7.9. Significant morbidities post-operatively included bowel stoma, long-term urinary catheter and new diagnoses of carcinomas.

Conclusion

Physicians and surgeons need to be suspicious of NSTI in severe cases of soft tissue infection to prevent delay in diagnosis and life-saving treatment. Scoring system can be used judiciously as adjunct to aid diagnosis.

Keywords

Necrotising soft tissue infection necrotising fasciitis Fournier’s gangrene diagnosis management

Introduction

Necrotising soft tissue infections (NSTI) are described as the most aggressive form of skin and soft tissue infections.^1,2 It rapidly and progressively breaks down the fascia and subcutaneous layers although all layers within a soft tissue compartment can be destroyed.^2,3 Necrotising fasciitis (NF) and Fournier’s gangrene (FG) are subgroups of NSTI and are potentially limb and life threatening. They are typically caused by multiple organisms such as the Streptococci, Enterococci, Staphylococci and Clostridium groups^3,4 affecting the immunosuppresed⁵ or individuals with systemic or local risk factors.^4,6,7

The average incidences in major hospitals were reported to be between 11 and 33 per annum from 1997 to 2005.^2,3,8,9 A United Kingdom population-based surveillance confirmed 136 cases of NF caused by Streptococcus pyogenes between 2003 and 2004.¹⁰ In reports published within the past decade,^4,6,10–12 mortality from this infection was still alarmingly high between 14.4% and 34%, regrettably because early recognition of NF or FG is difficult as they present with similar benign signs as other soft tissue infections in the early stage.^8,13 Therein, risk indices have been described to help diagnosis and as prognostic value.^6,8,14,15 Principles of treatment involve resuscitation, provision of antibiotics and early control through surgical debridement followed by physiological support and further debridement if required.^2,9,11,13

In this report, we describe the presentation, management and outcomes of patients treated for NF and FG in a tertiary centre Plastic Surgery and Reconstructive department. We also explored the retrospective values of published diagnostic and prognostic tools.^14,15

Methods

Using our department’s prospective electronic database, we retrospectively identified patients who were treated for NSTI between 2004 and 2010. Casenotes were reviewed for data collection. The Laboratory Risk Indicator for Necrotising Fasciitis (LRINEC)¹⁴ incorporating C-reactive protein (mg/l), white cell count (total/mm³× 1000), haemoglobin (g/dl), serum sodium (mmol/l), serum creatinine (micromol/l) and serum glucose (mmol/l), and Fournier’s Severity Index (FSI)¹⁵ incorporating temperature (degree Celsius), heart rate (beat per minute), respiratory rate (breath per minute), serum sodium (mmol/l), serum potassium (mmol/l), serum creatinine (mg/100 ml), haemotocrit (%), white cell count (total/mm³× 1000) and serum bicarbonate (mmol/l) were collated from admitting blood samples and calculated for patients seen and had surgery in our hospital. Statistical analyses were performed using SPSS version 18.

Results

Demographics

Twenty-four patients (18 male and 6 female) were identified. Thirteen cases were NF and 11 FG. Mean age was 52.1 (range 22–71 years old). Table 1 shows the patients’ significant co-morbidities, medications and sites involved. Six (25%) of the patients did not have significant known co-morbidity at presentation. Only eight (33.3%) had a known cause of the infection.

Table 1.

Demographics and sites of infection.

Age	Sex	Diagnosis	Site	Precipitant	Co-morbidity	Significant medication
22	M	FG	PR	n/k	Alcoholism	n/k
38	M	NF	LE	n/k	Ulcerative Colitis	Prednisolone, Sulphasalazine
54	F	FG	PR	n/k	RA	Sulphasalazine
48	M	NF	Mediastinum	Burn forearm	n/k	n/k
^63	F	NF	LE	Laceration	Asthma	Prednisolone
41	M	NF	LE	Laceration	Asthma	n/k
^68	F	NF	Peri-orbital	n/k	Thrombocytopenia	Hydroxycarbamide
30	M	NF	LE	n/k	n/k	n/k
66	M	NF	LE	Scorpion bite	DM	Metformin
^66	F	NF	LE	Vein harvest	DM, IHD, post-CABG, IABP right groin	n/k
^66	M	NF	LE	n/k	Varicose vein	n/k
55	M	NF	UE	n/k	Bronchiectasis	n/k
71	M	FG	PR	n/k	IHD, TIA, PVD, Lung carcinoma	Prednisolone
37	M	FG	PR	Vasectomy	n/k	n/k
56	M	FG	PR	n/k	n/k	n/k
47	M	FG	PR	n/k	Obesity	n/k
63	M	FG	PR	n/k	n/k	n/k
66	M	FG	PR	n/k	Obesity	n/k
65	M	FG	PR	n/k	n/k	n/k
62	M	FG	PR	n/k	RA, Gammopathy, IHD, COPD	Hydroxychloroquine
47	M	FG	PR	n/k	DM, obesity	Metformin
53	F	NF	LE	Thigh abscess	DM, IHD, COPD, obesity	Metformin, Gliclazide, Prednisolone
33	M	NF	LE	n/k	Asthma	n/k
34	F	NF	LE	Drug injection	IVDU	n/k

Note: M: male; F: female; NF: necrotising fasciitis; FG: Fournier’s gangrene; n/k: none known; DM: diabetes mellitus; IHD: ischaemic heart disease; CABG: coronary artery bypass graft; IABP: intra-aortic balloon pump; TIA: transient ischaemic disease; PVD: peripheral vascular disease; RA: rheumatoid arthritis, COPD: chronic obstructive pulmonary disease; IVDU: intravenous drug user; PR: perineum; LE: lower extremity; UE: upper extremity; ^non-survivor.

Figure 1 shows patients’ pathway during treatment. Sixteen patients (9 NF and 7 FG) were seen and treated fully in our hospital. The remaining eight patients (4 NF and 4 FG) were referred from outlying hospitals. Seven had undergone debridement primary debridement in the original admitting hospitals while one NF case had primary surgery in our hospital due to lack of specialist service provision in the referring hospital.

Figure 1.

Pathway of patients management.

Patient history

All patients who were seen in our hospital presented with pain, swelling, erythema and tenderness at the affected site. There was also presence of necrotic patch (n = 5, 2 with discharging pus, 2 with spreading erythema and 1 with oedema). The initial treatment diagnoses were cellulitis (n = 3), soft tissue abscess (n = 2), osteomyelitis (n = 2), scrotal abscess (n = 2), perineal abscess (n = 2), compartment syndrome (n = 1), septic arthritis (n = 1), ischaemic leg (n = 1), deep vein thrombosis (n = 1), peri-orbital cellulitis (n = 1), oesophageal perforation (n = 1) and scrotal infection (n = 1).

Only 4 of the 10 (40%) NF cases were diagnosed or suspected as NSTI at the initial examination, the others on subsequent examination. In the FG group, only 28.6% (2 out of 7) were diagnosed during initial assessment. We were not able to accurately document the patients’ symptoms and signs from other hospitals due to documentation restriction.

Timing of primary surgery

Fifteen patients who underwent primary debridement in our hospital were further reviewed. In this subgroup, the average time intervals from presentation to primary surgery were 17.7 h (range 3–48) and 4 h (range 1–8 h) from diagnosis to primary debridement. However, this might be an overestimation as two patients had an initial incision and drainage for presumed abscess before NSTI was suspected. Another patient required hospital transfer due to the absence of specialist service in the referring hospital. In the non-survivors (n = 4), the mean time interval from presentation to primary surgery was 18.3 h (not statistically significant compared with the survivors).

Further surgery

All patients underwent at least another surgery after primary debridement, ranging from 1 to 12 operations (average 3.9). Fifteen patients had split skin grafting and two required artificial dermis. In the FG cases, two patients required an advancement flap to provide soft tissue coverage to perineum.

Post-operative care

Within the subgroup of 15 patients, 12 (80%) patients required higher level of care post-operatively, averaging 12.9 (range 1–85) days in the Intensive Care Unit or the Higher Dependency Unit. Nine required inotropes, five required renal support and four received intravenous immunoglobulins (IVIg).

Organisms and antimicrobials

All 15 patients had variable microbiology investigations during the peri-operative period. Thirteen patients had wound swabs with organism isolated in 12 (92.3%) and 12 blood cultures (50% positive organism isolation), 11 had tissue cultures (10 positive organism isolation) and 9 had tissue histology.

Fourteen patients had at least one organism isolated. One patient had a single organism, Pseudomonas isolated. Table 2 shows types of organisms involved.

Table 2.

Types of organisms isolated.

Gram-positive	n	Gram-negative	n
Group A Streptococcus	5	Escherichia coli	4
Streptococcus, other	3	Bacteroides sp	4
Staphylococcus, other	4	Pseudomonas sp	2
Staphylococcus aureus	1	Prevotella sp	2
Aerococcus sp	3	Enterobacteriaceae, other	2
Enterococcus sp	1	Anaerobes	1
Clostridium sp	1
Corynebacterium sp	1
Peptostreptococcus sp	2

Note: Supplementary ‘Results’ information. The table shows types of micro-organisms isolated from the subgroups of 15 patients.

All patients received antibiotics before surgery. Antibiotics given included intravenous Flucloxacillin, Benzypenicillin, Co-Amoxiclav, Cefuroxime, Clarithromycin, Metronidazole, Gentamicin, Co-Fluampicil and Piperacillin/Tazobactam.

Mortality

The overall mortality rate (death in hospital) was 16.7% (4/24). Mortality in patients who were treated primarily in our hospital was 4 out of 16. There was no mortality in the FG group. Two died within 5 days after primary debridement of NSTI. One died after second debridement surgery. One was managed non-operatively.

Laboratory markers and scoring indices

The average LRINEC scores were 7.9 (SD 2.4), 7.6 (SD 2.8) and 8.7 (SD 0.6), respectively, in all patients, survivors and non-survivors. The average FSI scores, in survivors and non-survivors were 8.3 (SD 4.8) and 9.3 (SD 1.2). The mortality rate for FSI score of more than 9 was 33.3% (2 in 6). There was no statistical difference between survivors and non-survivors.

We found that there were acidosis (serum bicarbonate [HCO^–₃] 19 mmol/l or less), acute renal injury (serum creatinine 210 mmol/l or more) and hypoalbuminaemia (serum albumin 28 mmol/l or less) in all the non-survivors but these were statistically not significant compared with survivors.

Morbidity

The average length of stay was 50.4 days (range 8–108) for the subgroup of 15 patients compared with 9.4 days (range 2–18) for patients transferred from other hospitals who did not require primary debridement in our hospital. Patients were followed up for an average 29.8 months (range 1–42). Significant morbidities were creation of bowel stomas (n = 5), long-term urinary catheterisation (n = 4), chronic wound infections (n = 2), recurrent anocutaneous fistula (n = 1), limb amputation (n = 1), orchidectomy (1 total and 1 hemiorchidectomy) and severe elbow scar contracture (n = 1). Three patients with FG had new diagnosis of malignancy (B-cell non-Hodgkin’s lymphoma, low-lying non-resectable anorectal adenocarcinoma and anorectal adenocarcinoma with metastasis).

Discussion

NSTI remains an extremely serious and difficult condition to treat that relies on a high index of suspicion, a prompt diagnosis and emergent radical surgical treatment. Despite the recognition of this fatal infection, overall mortality in published literature is still reported between 14.4% and 34% in the past decade.^4,10–12 From our experience, it is a multi-disciplinary disease, presenting in and requiring input from various hospital teams peri-operatively.

Eron et al.¹ classified skin and soft tissue infection into four classes of increasing severity to help treatment of soft tissue infections. These are simple cellulitis (Class 1), cellulitis in a low-risk stable patient who is febrile and ill appearing (Class 2), which then progress to toxic appearance with background co-morbidity or limb-threatening level, for example, the appearance of haemorrhagic bullae (Class 3) and finally patients with sepsis syndrome or life-threatening infection for example necrotising fasciitis (Class 4).

NSTI can affect any part of the body, namely the head and neck, trunk, perineum and the extremities with similar initial presentation with simple cellulitis.^4–8,13 We observed that NSTI can affect both ‘healthy’ and ‘at risk’ patients. Interestingly though may be of little clinical significance, similar to other report,⁷ we found presence of existing co-morbidities which had been linked as a risk factor for developing NSTI, such as colorectal carcinoma. In addition, one patient was diagnosed with B-cell lymphoma.

Wong et al.¹⁴ retrospectively quantified and validated LRINEC score to aid diagnosis of NF. Our patients’ average LRINEC score was higher than their proposed cut-off score of 6 which had a positive diagnostic predictive value of 92%. A high score should warrant higher suspicion in triaging patients for more aggressive treatment, but should not be used in isolation. In the presence of appropriate service and expertise, Majeski and Majeski⁹ utilised bedside frozen section biopsy in equivocal cases.

McHendry et al.³ prompted us to appreciate the life-saving value of early debridement in the presence of NSTI. Reports by other authors seemed to be in agreement (mean interval to surgery 45 h – mortality 29%,³ mean 540 min – mortality 14.4%,¹¹ mean 4 h – no mortality⁹), that is, earlier surgical debridement resulted in better survival.^8,9,13

Besides early surgical debridement, antibiotics and post-operative managements are important in managing NSTI. The common organisms involved in our cases are Group A Streptococcus, Staphylococcus, Escherichia coli and the Bacteroides. Our local antimicrobial for NSTI suggests provision of Clindamycin and Piperacillin/Tazobactam if a diagnosis of NSTI is likely. This should succinct as initial cover but we would strongly advise referrals to microbiology from the onset of treatment, keeping in mind other potential virulent pathogens.^4,6

Post-operatively, patients usually would need a high level of care and further exploration until wound is satisfactorily free from active infection.^2,9,11,13 Intravenous immunoglobulin was part of treatment given to three unwell patients who subsequently did not survive.

Our mortality rate was high. Three of the mortalities involved infections of the lower extremities and one was peri-orbital. First patient had NF of infected vein harvest site post-coronary bypass surgery. This patient was in extremis and a decision was made not to proceed with surgical debridement. Diagnosis was confirmed in post-mortem. Second patient sustained a leg laceration and was on long-term oral prednisolone for asthma. The third patient was taking hydroxycarbamide for essential thrombocytopaenia when suffering NSTI of the peri-orbital regions. The fourth patient had no identifiable risk factor on presentation. This patient underwent an above knee amputation to control the infection in the early period. All these patients succumbed to multi-organ failure post-operatively, a physiological effect of severe sepsis common in NSTI.

All patients with FG in our report survived the infection. Our theory is infection in this region raised suspicion and hastened the treatment of a potentially fatal infection. There was also no mortality in patients transferred from other hospitals as they were clinically stable, not requiring higher level of care and had no active infection.

We did not discuss the role of radiological investigations which should not delay surgical debridement if NSTI is suspected in the extremity. Our small number of patients may underpowered the value of scoring system but we strongly advised using this only as an adjunct. It is logistically difficult to determine the time interval of symptoms onset to operation; therefore, we used the interval time from admission to surgery as a surrogate marker. We were also limited to the retrospective nature of this study, relying on accurate medical documentation.

Conclusion

Soft tissue infection should not be regarded lightly. Conversely, to treat any cellulitis as NSTI until proven otherwise would be too extremist a practice. The key to NSTI management is early surgical debridement supported by physiological resuscitation. The biggest lesson from this study is the role of multi-disciplinary management that these patients require. A local or regional guideline of managements of soft tissue infection should be put in place to raise awareness of NSTI and ensure proper treatment in the presence of one.

Footnotes

Declaration of conflicting interests

No competing financial or commercial interest.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Eron

Lipsky

Low

. Managing skin and soft tissue infections: expert panel recommendations on key decision points. J Antimicrob Chemother 2003; 52: S1,i3–17.

Anaya

Dellinger

. Necrotizing soft-tissue infection: diagnosis and management. Clin Infect Dis 2007; 44: 705–710.

McHendry

Piotrowski

Petrinic

. Determinants of mortality for necrotizing soft-tissue infections. Ann Surg 1995; 221: 558–563.

Hsiao

Weng

Yuan

. Predictors of mortality in patients with necrotizing fasciitis. Am J Emerg Med 2008; 26: 170–175.

Whallet

Stevenson

Wilmhurst

. Necrotising fasciitis of the extremity. J Plast Reconstr Aesthet Surg 2010; 63: e469–473.

Anaya

McMahon

Nathens

. Predictors of mortality and limb loss in necrotizing soft tissue infections. Arch Surg 2005; 140: 151–157.

Moslemi

Gilani

MAS

Moslemi

. Fournier gangrene presenting in a patient with undiagnosed rectal adenocarcinoma: a case report. Cases J 2009; 2: 9136–9136.

Wong

Chang

Pasupathy

. Necrotizing fasciitis: clinical presentation, microbiology and determinants of mortality. J Bone Joint Surg Am 2003; 85-A: 1454–1460.

Majeski

. Necrotizing fasciitis: improved survival with early diagnosis by tissue biopsy and aggressive surgical treatment. South Med J 1997; 90: 1065–1068.

10.

Lamagni

Neal

Keshishian

. Severe Streptococcus pyogenes infection, United Kingdom, 2003-2004. Emerg Infect Dis 2008; 14: 202–209.

11.

Hong

Chou

Liu

. The effect of prolonged ED stay on outcome in patients with necrotizing fasciitis. Am J Emerg Med 2009; 27: 385–390.

12.

Ersay

Yilmaz

Akgun

. Factors affecting mortality of Fournier’s gangrene: review of 70 patients. ANZ J Surg 2007; 77: 43–48.

13.

Dellinger

. Severe necrotizing soft-tissue infections. Multiple disease entities requiring a common approach. JAMA 1981; 246: 1717–1721.

14.

Wong

Khin

Heng

. 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: 1535–1541.

15.

Laor

Palmer

Tolia

. Outcome prediction in patients with Fournier’s gangrene. J Urol 1995; 154: 89–92.