Necrotising fasciitis (NF) is a fulminant soft tissue infection that spreads rapidly along the fascial planes, and when unabated can lead to septic shock and death.¹ It is a rare soft tissue infection with reported incidence of 0.3–3/100 000 persons per year with a mortality rate of up to 32% in Asia and Europe.^2,3 Necrotising fasciitis can affect any part of the body; lower limbs (32%), upper limbs (24%), perineum (16%), trunk (16%), and head and neck (10%).¹ The commonly reported risk factors associated with NF are diabetes, alcoholism, chronic kidney disease, and liver cirrhosis.²

The clinical presentation of NF can overlap with clinical features of other soft tissue infection such as cellulitis. Wong et al.⁴ reported a misdiagnosis of 73% of NF cases as cellulitis and subsequently developed Laboratory Risk Indicator for Necrotising Fasciitis (LRINEC) score to aid in distinguishing NF from other soft tissue infections. The score is calculated from six parameters: C-reactive protein, total white cell count, haemoglobin, sodium, creatinine, and glucose.⁴

There are four types of NF classified according to the microorganisms often isolated in culture of NF infections. Type I is a polymicrobial infection including, Group A Streptococcus (GAS), Bacteroides fragilis, Staphylococcus aureus, Clostridium perfringens and Enterobacterales. Type II infections are monomicrobial mainly the Group A Streptococcus or methicillin-resistant Staphylococcus aureus (MRSA). Type III is the marine Vibrio vulnificus species that inoculate injuries exposed to sea water.¹ Type IV is an uncommon type of NF caused by fungal infection.^5,6

Once a diagnosis or suspicion of NF is made, broad-spectrum antimicrobial treatment is commenced and surgical debridement is performed. Early surgical debridement and an appropriate antimicrobial therapy is crucial to improve survival outcomes.⁷ Antimicrobial resistance has emerged to be a threat to public health. In 2019, Antimicrobial Resistance Collaborators estimated 4.95 million (3.62–6.57) deaths associated with bacterial antimicrobial resistance. Sub-Saharan regions have been found to have the highest mortality associated with antimicrobial resistance at 27.3 per 100 000 (20.9–35.3). The common six pathogens associated with antimicrobial resistance deaths were: Pseudomonas aeruginoasa, Escherichia coli, MRSA, Klebsiella pneumoniae, Streptococcus pneumoniae and Acinetobacter baumannii. Staphylococcus aureus alone was responsible for 100 000 deaths associated with antimicrobial resistance in 2019.⁸ There has been emerging reports of multidrug resistance Acinetobacter baumannii skin and soft tissue infections across different parts of the world. Of concern, is that some of the multidrug resistance Acinetobacter baumannii skin and soft tissue infections were reported in previously healthy patients.^9,10,11

Necrotising fasciitis is a fulminant soft tissue infection that has not been well studied in South Africa. There were only a handful of case reports published on NF in a South African setting. These cases of NF were associated with treatment; immune suppressive therapy, radiation therapy, and diclofenac intramuscular injection.^12,13,14 It is imperative to grow a body of knowledge on this subject. The cornerstone of treatment in NF is surgical debridement and antibiotics. Amoxicillin and/or Clavulanic acid and clindamycin are the antibiotic of choice in our institution; however, this choice was not informed by prior clinical research. Antimicrobial resistance profile in NF infections had not been studied in our setting. The primary objective was to describe the microbiology profile of NF and to describe the antimicrobial susceptibility thereof. The secondary objectives were to describe the clinical and laboratory findings of NF, to determine the comorbidities and risk factors associated with NF, and to describe the LRINEC score in the patients with NF.

The results of this study will help inform antimicrobial protocols in the Antimicrobial Stewardship Committees in the institution and in other institutions like ours. There is no doubt that studies on NF are very much needed in South Africa to fill up the existing gap in the literature.

A total of 65 patients were admitted with NF to Edenvale Hospital between January 2014 and December 2020. A total of 12 patients were excluded from the study; 4 patients had no culture sent for microscopy, 3 patients’ cultures had no growth, and 5 patients had incomplete operation notes. A total of 53 patients were included in the study. The median age of the patients was 45.5 (38.5–56) years and majority of the patients were males (35 [66%]). Many of the patients had no comorbidities reported (25 [47.17%]). Necrotising fasciitis affected mainly the lower limbs (17 [32.08%]) with a low risk LRINEC score of 5 or less (13 [59.09%]) (Table 1). The LRINEC score analysis was underpowered as only 22 patients had all the parameters for LRINEC score calculation.

Microbiology analysis revealed predominately type II (monomicrobial) category (31 [58.49%]) (Table 1). There was a total of 88 first isolates including probable contaminants, which were included in the final analysis. A total of 48 (55%) were Gram-positive bacteria, 39 isolates (44%) were Gram-negative bacteria, and 1 isolate (1%) was Candida species. The number of anaerobic microorganisms isolated in the analysis period was 5 (6%). Staphylococcus aureus represented the largest proportion of Gram-positive bacteria cultured (18 [38%]), and 17 (94%) was susceptible to cloxacillin (Figure 1). Methicillin resistance was found in 1 case of (6%) Staphylococcus aureus (Table 2). There were only 2 (4%) cases of Streptococcus pyogenes in the group of Gram-positive bacteria, which were also susceptible to clindamycin.

Escherichia coli was the main species isolated in the Gram-negative bacteria (14 [36%]) with 11 (92%) isolates within the Enterobacterales susceptible to amoxicillin and/or clavulanic acid (Figure 2). Multi-drug resistance (MDR) was detected in Klebsiella pneumoniae (three isolates) and Enterobacter species (one isolate) being resistant up to the third-generation cephalosporins, and carbapenem-resistant Enterobacterales (CRE) was detected in two isolates (Table 2). The majority of first isolates (60 [68%]) including Staphylococcus aureus and the anaerobes would have been covered by the amoxicillin and/or clavulanic acid and clindamycin used.

The non-fermenter Gram-negative bacteria (NF-GNB) were minimal (7 [8%]) from the total first isolates with 5 (6%) Pseudomonas aeruginosa isolates being highly susceptible to most routinely tested antipseudomonal antibiotics (ceftazidime and piperacillin-tazobactam [80% for both] and 100% for meropenem and imipenem). The other isolated NF-GNB (Acinetobacter baumannii [n = 1] and Pseudomonas species [n = 1]) demonstrated lower susceptibilities with the aminoglycosides (gentamycin and tobramycin) and the antipseudomonal carbapenems (imipenem and meropenem) being susceptible.

There was only one yeast isolate (Candida albicans) identified during the study period. Although there were no susceptibility results reported, there has not been reports of fluconazole resistant Candida albicans reported in our institution.

Bivariate analysis between demographics, comorbidity, LRINEC score, and microbiology type of NF showed no statistical significance. However, there was an association between the anatomic site affected by NF and the microbiology type (p = 0.03) (Table 3).

Necrotising fasciitis is a lethal soft tissue infection that spreads rapidly along the fascial planes and carries a significant mortality rate if not treated promptly.¹ The demographics of the patients in our study were similar to the global trend previously reported in the literature with mostly male patients, except that our patients were slightly younger than in previous studies, which reported an average age of more than 50 years.^15,16,17 South Africa is known to suffer from the infamous ‘Quadruple burden of disease’, which includes trauma of which majority of trauma victims are young patients.¹⁸

Less than 50% of the study population had no comorbidities and only human immunodeficiency virus (HIV) and diabetes mellitus were reported. Other complex comorbidities associated with NF such as liver disease and kidney disease were not reported. Diabetes mellitus is the commonly reported comorbidity associated with NF across the world.^2,15,19 Al-Qurayshi et al.¹⁵ reported that half the patients in a cross-sectional study of 4178 in the United States with NF had diabetes. Diabetic patients have increased susceptibility to infection. In our context, HIV was the most commonly reported comorbidity (30.1%), less so in other African countries such as for example, Kenya (6%).²⁰ Apart from the immunosuppression caused by HIV infection, associated organ dysfunction and opportunistic infections may further weaken the already fragile immune system. The HIV population is also more prone to infections that would normally be innocuous to a fully functional immune system.²¹

Type II (monomicrobial) was the common type of NF. Staphylococcus aureus represented the largest proportion of the Gram-positive bacteria with a very low rate of methicillin resistance. Escherichia coli constituted a great proportion of the Gram-negative microorganism. Although there was a concerning resistance of Klebsiella pneumoniae and Enterobacter species in the Gram-negative group fortunately, Klebsiella pneumoniae and Enterobacter species were not common isolates in the NF specimens. In an England cohort study of 11 042 by Bodansky et al.,¹⁹ a predominance of Gram-positive isolates (mostly were Staphylococcus aureus) followed by Gram-negative isolates which were mostly Escherichia coli and Klebsiella pneumoniae being reported. These findings are similar to our study. Contrary to a Scandinavian multicentre cohort of 409 patients and Malaysian cohort of 469 patients which reported large numbers of Group A Streptococcus in the group of Gram-positive bacteria. Group A Streptococcus is a virulent microorganism and was associated with septic shock and kidney failure; however, it was not associated with higher mortality.^17,22 There were only two cases of Group A Streptococcus in our study which were 100% sensitive to penicillin and clindamycin. The choice of antimicrobial therapy in the Scandinavian and Malaysian studies were penicillin (Ampicillin and/or sulbactam in Malaysia) and clindamycin.^22,23 The findings of this study support that amoxicillin and/or clavulanic acid is an appropriate antimicrobial choice for empiric therapy to cover both the commonly isolated Gram-positive and Gram-negative bacteria in our institution. Although Staphylococcus aureus was not tested for susceptibility to amoxicillin and/or clavulanic acid, Clinical & Laboratory Standard Institute (CLSI) guidelines considers methicillin (oxacillin) susceptible staphylococci to be susceptible to beta lactam combination agents such as amoxicillin and/or clavulanate, ampicillin and/or sulbactam, and piperacillin and/or tazobactam.²⁴ Although Group A Streptococcus was not a common isolate in our institution, it is imperative to add clindamycin to the empiric therapy in order to neutralise the virulent exotoxins released by Group A Streptococ cus. Adjunctive clindamycin has been proven to reduce mortality and morbidity in Group A Streptococcus infections.²⁵ Empiric clindamycin can be discontinued as soon as culture results have excluded Group A Streptococcus to prevent side effects such as Clostridioides difficile infection.

Because of the retrospective nature of study, other predisposing factors could not be explored such as smoking, drug injection, and trauma. As a result of incomplete records, we were not able to determine the risk factors associated with NF. The LRINEC score calculation was also underpowered as there was only about 40% of the patients who had all the parameters required for the score calculation. Patients with complex comorbidities were not studied in this study as they were referred to a tertiary centre. It remains unclear whether they would have a different microbiology profile of NF infections. It is recommended that prospective multicentre studies are conducted to understand NF in details in the South African context.

In our study we found relatively younger patients, predominantly males with a low LRINEC score with NF affecting mainly the lower limbs. The NF was mostly type II (monomicrobial). Staphylococcus aureus and Escherichia coli were the most common pathogens cultured in the Gram-positive and Gram-negative groups, respectively. There was an overall low rate of antimicrobial resistance to amoxicillin and/or clavulanic acid, and clindamycin, thus the combination can be used as an empiric therapy.