Indian Journal of Medical Microbiology Home 

[Download PDF]
Year : 2003  |  Volume : 21  |  Issue : 3  |  Page : 202--204

Bacteriological studies of gas gangrene and related infections

A De, A Varaiya, M Mathur, A Bhesania 
 Department of Microbiology, LTM Medical College and General Hospital, Sion, Mumbai - 400 022, India

Correspondence Address:
A De
Department of Microbiology, LTM Medical College and General Hospital, Sion, Mumbai - 400 022


Gas gangrene, a life threatening condition is predominantly caused by Clostridium spp. Over a five year period (1996-2000), 580 wound swabs were processed from patients who had vehicular accidents with crush injury. The specimens were processed for anaerobic and aerobic organisms by standard laboratory techniques. Of the 580 swabs, 55(9.48%) were primary smear positive for spore bearing bacteria, of which 48 grew Clostridium spp. They were the sole anaerobic isolates in 30 patients (62.5%) whereas Clostridium spp. along with nonsporing anaerobes were isolated in 18 patients. Clostridium perfringens was the predominant isolate in 39 out of 48 cases (81.3%). However, 16 wound swabs which were primarily smear negative for spore bearers, grew Bacteriodes fragilis (11) and Fusobacterium species (5). All these 16 patients had foul smelling discharge with crepitation in the tissues. Our study highlights that apart from Clostridium spp., non-clostridial crepitant infections may present as myonecrosis simulating gas gangrene.

How to cite this article:
De A, Varaiya A, Mathur M, Bhesania A. Bacteriological studies of gas gangrene and related infections.Indian J Med Microbiol 2003;21:202-204

How to cite this URL:
De A, Varaiya A, Mathur M, Bhesania A. Bacteriological studies of gas gangrene and related infections. Indian J Med Microbiol [serial online] 2003 [cited 2017 Oct 22 ];21:202-204
Available from:

Full Text

Gas gangrene, a life threatening condition is predominantly caused by Clostridium spp. in the civilian set up. There has been rising incidence of this infection in vehicular accident cases.[1] It is typically characterized by presence of gas in the tissues and necrotising myositis. However, these symptoms do not always indicate a clostridial infection as recent reports have shown that non crepitant infection also simulate gas gangrene. A five year retrospective bacteriology study with clinical correlation of patients suspected of gas gangrene is reported.

 Materials and methods

A total of 580 wound swabs were processed from patients admitted to Lokmanya Tilak Municipal General Hospital clinically suspected of gas gangrene over a period of 5 years (1996-2000). Swabs were collected from the depth of the wound, transported in Stuart's transport medium and processed immediately in the laboratory. Gram staining of all smears were done to study the morphology of bacteria and look for presence of inflammatory cells. All specimens were processed for aerobes and anaerobes. Aerobic culture was done on blood agar (BA) and MacConkey agar (MA) and incubated at 37C overnight. The colonies were identified by standard laboratory techniques.[2]

For anaerobic culture the swabs were inoculated in RCM (Robertson's cooked meat) medium along with direct culture on neomycin blood agar (NBA). Both RCM and NBA were incubated for 48 hours in McIntosh Fildes' anaerobic jar at 37C. Smears were prepared from RCM to study morphology of organisms and subculture was done on NBA. Subculture plates were further incubated anaerobically in the anaerobic jar at 37C for 48 hours, if direct plating did not grow any organisms.[3]

Colonies were identified by Gram staining and colony morphology, along with aerotolerance on BA. Those which were aerotolerance positive and showed oval subterminal spores were further identified as Clostridium histolyticum / Clostridium tertium on the basis of lipase production and sugar fermentation tests. Those which were aerotolerance negative were further speciated on the basis of lecithinase and lipase production, hemolysis on NBA, spore enhancing test, disc sensitivity test and sugar fermentation test.[4]


Out of total 580 swabs processed, growth was seen in 529 (91.2%) - only aerobes were isolated from 373 samples (70.5%), a mixture of aerobes and anaerobes from 98 samples (18.5%), and only anaerobes were isolated from 58 samples (11%). Out of 529 samples with growth, clostridia were isolated in 48 and other aerobic and anaerobic bacteria in 481.

Primary smear was positive for spore bearing organisms in 55 patients (9.5%). Total aerobes isolated were 907, of which gram negative bacilli were 756 (83.3%) and gram positive cocci were 151 (16.7%). Total anaerobes isolated were 156, of which gram negative bacilli were recovered from 55 (35.2%) of the isolates, gram positive cocci from 53 (34%) and gram positive spore bearing bacilli from 48 samples (30.8%).

[Table] shows the total aerobes and anaerobes isolated from the swabs. Amongst aerobic gram negative bacilli, E.coli was predominant (23.3%) followed by Proteus spp. (21.9%). Amongst aerobic gram positive cocci S.aureus was predominant (11.2%). Sensitivity was maximum to amikacin (30g) 80%, followed by augmentin (30g) 70%. Pseudomonas spp. showed 78% sensitivity to piperacillin.

Clostridium spp. were recovered from 48 patients of which 39 were Clostridium perfringens (81.3%). Amongst anaerobic gram negative non sporing bacilli, Bacteroides fragilis was isolated in 25 cases (16%) out of which 11 (44%) grew Bacteroides fragilis as the sole agent. The clinical presentations in these patients were brown to black foul smelling discharge from the edge of the lesion, with discolouration of the surrounding skin. Crepitus was detected in most cases. In some, X-ray showed gas in the muscle or under the skin.

Amongst 48 cases where clostridia were the causative agents, in 45.8% cases only anaerobes were isolated and in 54.2% cases a mixed growth of aerobes and anaerobes was seen. Amongst 481 cases where non-clostridial bacteria were isolated, majority showed only aerobic growth (77.5%), mixed growth was seen in 15% and only anaerobes (other than clostridia) in 7.5%.


Though Clostridium perfringens is recognized as the leading cause of gas gangrene, it can also be caused by non clostridial bacteria.[5] In addition, anaerobic streptococcal myositis also mimics gas gangrene.[6]

Correlation between Gram staining of direct smear and culture positivity in our study was 87.3%. Baradkar et al had shown cent percent correlation between the two.[6] Out of total 156 anaerobes isolated in our study, 48 were Clostridium spp. (30.8%). Rao et al had reported 56% Clostridium spp. in their study.[7] The predominant Clostridium spp. in this study was C. perfringens (81.3%), which is also the commonest isolate among the gas gangrene cases reported from various studies.[6],[7],[8] The Clostridium spp. isolated in this study were C. histolyticum and C. novyi. These organisms were also isolated in the study by Rao et al.[7] Out of 48 patients in whom Clostridium spp. were isolated, they were the sole pathogen in 30 (62.5%) and Clostridium spp. along with non sporing anaerobes from 18 patients (37.5%). However, Clostridium bifermentans, C. sporogenes and C. septicum which were reported by other,[1],[6],[7],[8] and which are also causative agents of gas gangrene, were not encountered in this study.

Local cleaning and debridement was done in all the cases. Anti gas gangrene serum was instituted after skin testing in all the patients. However, in 18 cases, amputation had to be done as the local lesions were rapidly progressive. In our study, it was found that 69.2% of the total anaerobes were non-clostridial anaerobes. Among them, anaerobic cocci predominated (33.9%). Gram staining of direct smear from these swabs showed gram positive cocci and plenty of pus cells with the absence of gram positive spore-bearing bacilli. Clinically they were crepitant infections with evidence of gas in the tissues. Rao et al had reported 44% of non clostridial anaerobes in cases of gas gangrene and 15% of them were anaerobic cocci.[7] So, the absence of spore-bearing bacilli in the smear does not rule out the possibility of gas gangrene. From 25 cases we reported Bacteroides fragilis, of which in 11(44%) patients B.fragilis was the sole isolate and all these patients had foul smelling discharge with crepitation in tissues simulating gas gangrene. Rao et al had isolated Bacteroides spp. from 36% cases.[7] Bacteroides spp. are known to produce crepitant lesions.[9] These cases were treated conservatively as the patients were not highly toxic. They recovered with local debridement and metronidazole. In all these patients, direct smear report was negative for Clostridium spp.

Amongst the aerobes, gram negative bacilli predominated (83.3%). Only aerobes were isolated from 70.5% swabs. Aerobes, usually members of Enterobacteriaceae, S.aureus and Streptococcus spp. are identified as the causative agents of clinical gas gangrene.[1] E.coli was the commonest aerobic isolate (23.3%), which is in accordance with the study by Baradkar et al.[6] This was followed by Proteus spp. (21.9%). In the study by Rao et al the commonest aerobic isolate was Proteus spp. (41%).[7] In 30 patients Streptococcus pyogenes was isolated in the present study.

Combination of gas in tissues and necrotising myositis does not always indicate a clostridial infection. Bessman and Wagner reported 49 cases of gas gangrene with myositis and soft tissue gas of which only one was caused by clostridial organisms.[5] Differential diagnosis of soft tissue infection include in addition to Clostridium spp., aerobic infections like haemolytic staphylococcal fascitis, hemolytic streptococcal gangrene, coliforms, anaerobic streptococcal infections, Bacteroides infections mixed aerobic and anaerobic infections.[10] Our study is in accordance with other studies which have also revealed that non clostridial crepitant infections can cause gas gangrene.[1],[5],[6],[7],[8] Thus, samples from gas gangrene cases should also be processed for non sporing anaerobes, which will help in institution of specific therapy and therefore better patient management.


1Udgaonkar US, Dharmadhikari CA, Kulkarni RD, Kulkarni V, Pawar SG. Clinicobacteriological study of gas gangrene. J Indian Med Assoc 1990;88(1):8-10.
2Baron EJ, Peterson LR, Finegold SM (Eds). Conventional and rapid microbiological methods for identification of bacteria and fungi, Chapter 10. In : Bailey & Scott's Diagnostic Microbiology, 9th ed. (Mosby, St.Louis) 1994:97.
3Baron EJ, Peterson LR, Finegold SM (Eds). Processing clinical specimens for anaerobic bacteria : Isolation and identification procedures, Chapter 35. In : Bailey & Scott's Diagnostic Microbiology, 9th ed. (Mosby, St.Louis) 1994:474.
4Sutter VL, Citron DM, Finegold SM (Eds). Processing clinical specimens and isolation and identification procedures, Chapter 3. In : Wadsworth Anaerobic Bacteroiology Manual, 3rd ed. (The C.V.Mosby Company, St.Louis) 1980:28.
5Bessman AN, Wagner W. Non clostridial gas gangrene : A report of 48 cases and review of literature. JAMA 1975;233:958-963.
6Baradkar VP, Patwardhan NS, Deshmukh AB, Damle AS, Karyakarte RP. Bacteriological study of clinically suspected cases of gas gangrene. Indian J Med Microbiol 1999;17(3):133-134.
7Rao SR, Natarajan MK, Ramesh I. An eight year bacteriological study of gas gangrene in Pondicherry. Indian J Med Microbiol 1995;13(3):151-154.
8Chaudhry R, Dhawan B. Gas gangrene and related infections in a tertiary care hospital. Indian J Med Microbiol 1998;16(4):165-168.
9Altermier WA, Fullen WD. Prevention and treatment of gas gangrene. JAMA 1971;217:806-813.
10Hart GB, Lamb RC, Strauss MB. Gas gangrene : A collective review. J Trauma 1983;23:991-1000.