August 2000, Volume 50, Issue 8

Original Article

Diagnosis of Clostridium Difficile Antibiotic associated Diarrhoea Culture versus Toxin Assay

Qamar Sultana  ( Department of Pathology, Postgraduate Medical Institute, Lahore. )
Naseer Ahmed Chaudhry  ( Department of Pathology, Postgraduate Medical Institute, Lahore. )
Muhammad Munir  ( Department of Pathology, Postgraduate Medical Institute, Lahore. )
Muhammad Saeed Anwar  ( Department of Pathology, Postgraduate Medical Institute, Lahore. )
Muhammad Tayyab  ( Department of Pathology, Postgraduate Medical Institute, Lahore. )


Objective: To compare the results of Clostridium Difficile (CD) on culture with detection of C. difficile toxin by Enzyme Immunoassay (ETA) in the stool specimens of hospitalized patients with antibiotic associated diarrhea (AAD).
Patients and Methods: The study included 80 adult patients with AAD and 20 adult patients with non-AAD. Stool specimens of all these subjects were inoculated on cycloserine cefoxitin fructose agar and incubated anaerobically to isolate C. difficile. At the same time, all the stool specimens were tested for C. difficile toxin by EIA technique using cytoclone A and B kit manufactured by Cambridge Biotech Corporation, Worecester, Massachusette.
Results: Out of  80 adult patients with AAD, thirty were females and fifty males. C. difficile was isolated on culture from stool specimen of 16 patients, while twenty-three stool specimens were positive for C. difficile toxin. From 20 control subjects, C. difficile was isolated from stool specimen of only one subject. No stool specimen from the controls was positive for toxin.
Conclusion: Diagnosis of CDAAD by culture is difficult and time consuming because of strict anaerobic nature of organism. Moreover, mere isolation of C. difficile on culture is not sufficient to establish the pathogenic role of these isolates. C. difficile toxin detection by ETA technique is a highly sensitive and specific method for diagnosis of CDAAD. Using this method, results are available in three hours time. Therefore, ETA is recommended for rapid diagnosis of CDAAD (JPMA 50:246, 2000).


Clostridium difficile is widely distributed in nature and is particularly prevalent in the hospitals1-4. It has been implicated as an etiologic agent in 90-100% cases of pseudomembranous colitis and 15-25% cases of antibiotic associated diarrhoea5,6. Antibiotic associated diarrhoea due to C. difficile is a toxin mediated disease7-11.
Isolation of C. difficile from stool specimen on culture and/or detection of C. difficile toxin has been used for laboratory diagnosis of C. difficile antibiotic associated diarrhoeal disease6,12-14 The isolation of C. difficile from stool provides only presumptive diagnosis of C. difficile associated enteric disease, because some strains of C. difficile are non-toxigenic12,14,15 while the presence of toxin indicates that diarrhoea is really due to a pathogenic strain8,10,16,17.
Tissue culture assay, counter current immunoelectrophoresis, enzyme linked immunosorbent assay and polymerase chain reaction have been used for detection of C. difficile toxin7,16,18,19. Out of these, enzyme linked immunosorbent assay has been found to be rapid and cost effective method15,6,20-22.
Antibiotic associated diarrhoea is common in Pakistan due to indiscriminate use of antibiotics. Present study was done to compare the results of C. difficile by culture with detection of C. difficile toxin by enzyme immunoassay in the stool specimens of hospitalized patients with antibiotic associated diarrhoea.

Patients and Methods

The present study was carried out on one hundred adult patients irrespective of sex, admitted in surgical and allied wards of Sir Ganga Ram Hospital/Services Hospital, Lahore. Out of these, 80 patients were suffering from antibiotic associated diarrhoea. They were taking antibiotics for a least 4-5 days prior to the onset of diarrhoea. From the same wards, 20 patients having non-antibiotic associated diarrhoea were studied as control subjects. The patients with diarrhoea were passing three or more stools with decreased consistency per 24 hours. The patients were included in the study irrespective of passage of mucus/blood in stool and other gastrointestinal complaints. Moreover, the patients were not subjected to sigmoidoscopic examination.
The diarrheal stool specimens of the patients were processed for isolation of C. difficile on culture and detection of C. difficile toxin with enzyme immunoassay method. For C. difficile culture a small amount of stool was added to Robertson cooked meat medium. The stool specimens were also taken to laboratory as such. In the laboratory gross examination was done and then half portion of these specimens was kept at -20°C till further use for toxin detection23.
Culture for Clostridium Difficile
For isolation of C. difficile, the stool specimens were inoculated onto cycloserine cefoxitin fructose agar (CCFA) containing defibrinated horse blood and egg yolk. In order to eliminate overgrowth of commensal organisms present in the stool specimens, one ml of liquid or one gram of solid stool specimens was homogenized in one ml of methylated spirit and then plated out on CCFA plates within 30 minutes24. The plates were incubated anaerobically at 37°C for 48 hours. Robertson cooked meat enrichment broth was sub-cultured onto CCFA plates. These plates were also incubated at 37°C under anaerobic environment for 48 hours. Anaerobic conditions were produced by Oxoid anaerobic system using BR 58 (Oxiod) gas-generating kits.
Preliminary identification of Clostridium difficile was done by colony morphology, gram staining, lecithinase/lipase activity, aerotolerance test and flourescence under long wave length (365 mm) ultraviolet light. Definitive identification of the isolates was done by various biochemical tests23,25 such as urease, indole, esculin hydrolysis, gelatin digestion, motility and sugar fermentation tests.
Detection of Clostridium difficile Toxin in Stool Specimens
All the stool specimens were tested for the presence of toxin A and toxin B against C. difficile by enzyme linked immunosorbent assay (ELISA) using Cytoclone A and B EIA kit (Cambridge Biotech Corporation, Worcester, Massachusetts) following the instructions of the Manufacturer.
This kit contained rnicrotiter plate, with wells being precoated with toxin A and B specific monoclonal antibodies. If either toxin was present in the stool samples, toxin monoclonal antibody and biotinylated goat polyclonal conjugate (specific for both toxins) complexes were formed which remained in the micro wells after washing. A strepta-vidin-horse radish peroxidase conjugate was then added into the wells. This got attached to biotinylated conjugate toxin complex if present. After final working step a sub-strate (urea peroxide) and chromogen (tetramethylbenzidine) mixture was added to the wells. Any bound conjugate converted the substrate chromogen to blue color. Addition of acid (stop solution) converted blue to yellow colour. The intensity of the yellowish color was proportional to the amount of toxin(s) bound in the wells. For statistical analysis of results, chi square (X2) test with yates correction was applied.


Out of 80 adult patients suffering from antibiotic associated diarrhoea, 30 (37.5%) were females and 50 (62.5%) males with male to female ratio of 1.66:1. Of the 20 subjects with non-antibiotic associated diarrhoea, 11 were females and 9 males. The age range of the patients was 18-65 years (mean of 35 years) and controls 23-60 years (mean age 43 years).
Out of 80 patients with antibiotic associated diarrhoea, the stool specimens of 4 were culture positive, toxin negative; twelve were culture positive, toxin positive while 11 were culture negative, toxin positive. Against that in control subjects only one was culture positive, while toxin was not detected in any stool specimen (Table). Based on positivity on culture and/or presence of toxin evidence of C. difficile infection was observed in twenty-seven (33.75%) patients and one control subjects (5%). Statistical analysis showed that infection was significantly higher in AAD patients (p<0.05) as compared to non-AAD controls.


Antibiotic associated diarrhoea/colitis is caused by Clostridium difficile and is a toxin mediated disease7-9,26,27. Use of broad spectrum antibiotics alter the colonic flora allowing the C. difficile to flourish in the human colon6,10,15,28,29. When the organism is established in the colon, pathogenic strains of C. difficile produce antibiotic associated diarrhoea through the action of two toxins, toxin A and toxin B9,26,30,31.
The isolation of Clostridium difficile from the stool is used for presumptive diagnosis of C. difficile associated enteric disease12,14,15,17. However, it is a less efficient method of establishing laboratory diagnosis because some strains of C. difficile are non-toxigenic5,15,19. The only way to establish the diagnosis is to detect toxin production by the isolates (toxigenic culture)13,16,33. However, this is a time consuming process14,15,33. The other method used for diagnosis of this disease is direct detection of C. difficile toxin(s) in the stool specimen of these patients13,16,18,20.
Successful recovery of Clostridium difficile by culture is challenging13. As it requires the proper selection of media for primary isolation, strict anaerobic environment for incubation and specific confirmatory methods24,25 are needed. As far as toxin detection is concerned, one ff the earlier methods available to detect the toxin B and toxin A was tissue culture cytotoxicity assay7,18,34. Limitations of the cytotoxicity assay are that the tissue culture facilities are expensive and expert personnel with specialized laboratory equipment are required5,21,33. Keeping in mind these difficulties, efforts were made to device some faster methods for the detection of Clostridium difficile toxin A, toxin B or both. Some of the methods which were tried, are latex agglutination test, gas liquid chromatography, countercurrent imm unoelectrophores is and enzyme linked immunosorbent assay35-40.
Enzyme linked immunosorbent assay (ELISA) has proved to be a good prospect32,38,40. Monoclonal antibodies of high specificity and sensitivity to Clostridium difficile toxins were developed and coated on microtitre plates41. ELISA takes about four hours which is a considerable saving of time when compared with the time required for cytotoxicity assay and/or toxigenic cultures15,16,20,22,23.
Proper laboratory diagnosis of Clostridium difficile antibiotic associated diarrhoea includes both culture and toxin assay for C. difficile42-44. In the present study, out of  80 stool specimens of antibiotic associated diarrhoea patients.27 cases showed evidence of C. difficile infection. Out of these, four cases were culture positive, toxin negative. Eleven cases were culture negative, toxin positive while 12 cases were culture positive as well as toxin positive. The finding that stools of antibiotic associated diarrhoea patients can be culture negative, toxin positive, have also been observed by many other workers18,27,45-47.
In the present study, out of 27 cases showing evidence of Clostridium difficile infection, four antibiotic associated diarrhoea patients were culture positive, toxin negative. This observation is in agreement with that of other workers42,48-50. Stools of 6-21% antibiotic associated diarrhea patients in these studies were found to be positive for C. difficile culture but negative for its toxin.
In the present study, C. difficile isolates were not tested for toxin production in vitro. Therefore, it is difficult to say whether these isolates represent colonization by nontoxigenic strains or by toxigenic strains. According to Butterworth et al (1998), C. difficile isolates from 73% of culture positive, toxin negative cases, showed toxin production in vitro. Keeping this finding in view, culture positive toxin negative patients in the present study may have infection due to toxigenic C. difficile and not colonization by non-toxigen ic strains.
Out of 20 control subjects in the present study, only one case was positive for Clostridium difficile on culture, while the toxin was not detected in stool specimen from any subject. This finding is in agreement with findings of other workers17,50,52. Their studies show that 5-11.5% hospitalized patients having non-antibiotic associated diarrhoea were positive on culture and negative for C. difficile toxin.
Cytotoxin tissue culture assay is widely accepted “gold standard’ test for detection of Clostridium difficile toxin20,47,53-55. However, it requires at least 24 hours to obtain results by this method39,43. Therefore, rapid methods are required to diagnose this disease. Out of these rapid methods, enzyme immunoassay and latex agglutination tests are good alternatives13,21,38,39,53,55 Latex agglutination test can be performed in three minutes. However, it has low sensitivity as well as spcificity13,21,34,56. Moreover, this method based on detection of glutamate dehydrogenase antigen which is not exclusive to toxigenic strains of C. difficile, may give false positive results15,21,38.
ELISA can be performed in three hours. Thus, it is a rapid method of detection of Clostridium difficile toxin in stool specimens. Moreover, it has been proved to be a sensitive as well as specific test15,20,34,47,57. Therefore, enzyme immunoassay test is recommended for rapid diagnosis of C. difficile antibiotic associated diarrhoea disease. This will help in timely treatment of patients with this disease and consequently reducing its morbidity and mortality.


1.Hafiz S, Oakley CL. Clostridium difticile: Isolation and characteristics. J.Med. Microbiol., 1976;9:129-37.
2.Fekety R, Kim KH, Batts DH, et al. Studies on the epidemiology of antibiotic Clostridinin difficile colitis. Am. J. Clin. Nutr 1980;33:2527-32.
3.Larson HE, Barclay FE, Honour P, et al. Epidemiology of Clostridiun difficile in infants. J. Infect. Dis., 1982;146:727-33.
4.Brazier JS. The epidemiology and typing of Clostridium difficile. J Antimicrob. Chemother.. 1 998;41 (Suppl):47-57.
5.Bartlett JO. Management of Clostridium difficile infection and other antibiotic associated diarrhoea. Eur. J. Gastrocntcrol. Hepatol., 1996;8:1054-61.
6.Kelly CP, Lamont JT. Clostridium difficile infection. Annu. Rev. Med.,1998;49:375-90.
7.Larson HE, Parry JV, Price AB, et al. Undescribed toxin in Pseudomembranous colitis. Br. Med.J., 1977;1:1246-48.
8.Bartlett JO, Chang TW, Gurwith M, et al. Antibiotic associated pseudomembranous colitis due to toxin-producing clostridia. N. EngI. J. Med., 1978;298:53 1-34.
9.Torres J, Jennische E, Lange S, et al. Enterotoxins from Clostridiuni difficile: diarrhoeogenic potency and morphological effects in the rat intestine. Gut,1990;31:781-85.
10.Pothoulakis C. Pathogenesis of Clostridium difficile associated diarrhoea.Eur. J. Gastroenterol. Hepatol., I996;8: 1041-47.
11.Smith JA, Cooke DL, Hyde S, et al. Clostridium difficile toxin, A binding tohuman intestinal epithelial cells. J. Med. Microbiol. (A), 1997;46:953-58.
12.Bartlett AG, Taylor NS, Chang TW, et al. Clinical and laboratory observations in Clostridium difficile colitis. Am. J. Clin. Nutr., 1980;33:2521-26.
13.Staneck JL, Weckbach LS, Allem SD, et al. Multicenter evaluation of four methods for Clostridium difficile detection: Immunocard C. difficile, cytotoxin assay, culture and latex agglutination. J. Clin. Microbiol., 1996:34:2718-21.
14.Jacob J, Rudensky B, Dresner J, et al. Comparison of four laboratory tests for diagnosis of Clostridium difficile associated diarrhoea. Eur. J. Clin. Micobiol. Infect. Dis., 1996;15:561-66.
15.Bartlett JO. Management of Clostridluin difficile infection and other antibiotic associated diarrhoea. Eur. J. Gastroenterol. Hepatol., 1996;8:1054-61.
16.Doem GV, Coughlin RT, Wo L. Laboratory diagnosis of Clostridium difficile associated gastrointestinal disease: Comparison of a monoclonal antibody enzyme immunoassay for toxins A and B with a monclonal antibody enzyme immunoassay for toxin A only and two cytotoxicity assay. J. Clin. Microbiol., 1992;30:2024-46.
17.Shim JK, Johnson 5, Samore Mil, et al, Primary symptoinless colonisation by Clostridiuin difficile and decreased risk of subsequent diarrhoea. Lancet, 1998;35 :633-66.
18.Chang TW, Launuann M, Bartlett JG. Cytotoxicity assay in antibiotic associated colitis. 3. Infect. Dis., 1979;140:765-70.
19.Kato N, Or CY, Kato H, et al. Detection of toxigenic Clostrtdium difficile in stool specimens by the polyinerase chain reaction. 3. Infect. Dis.,1993;l67:455-98.
20.Dipcrsio JR, Varga FJ, Conwell DL, et al. Development of rapid enzyme immunoassay for Cloatridium difficile toxin A and its use in the diagnosis of Clostridium difticile associated disease. J. Clin. Microbiol., 1991;29:2723-30.
21.Kelly CP, Pothoulakis C. Lamont JT. Clostridium difficile colitis. N. Eng. 3. Med., 994;330:1754-55.
22.Wei SC, Wong 3M, Hsueh PR, et al. Diagnostic role of endoscopy stool culture and toxin A in Clostridium difficile associated disease. J. Formos. Med. Assoc., 1997:96:879-83.
23.Brown R, Collee JG, Poxton IR. Bacteroides Fuaobacterium and other Grain-negative anaerobic rods: anaerobic cocci, identification of anaerobes. In: Collee JG. Fraser AG. Marmion BP, Simmons A (eds). Mackic and McCartney Practical Medical Microbiology. 14th Ed. Edinburgh: Churchill Livingatone. 1996;501-19.
24.Poxton lR, Brown R, Fraser AG, et al. Enteropathogenic clostridia and clostridium botulinum. In: Cotlee 3G. Dugnid JR Fraser AG, Marmion BP (eds) Mackie and MeCartney Practical Medical Microbiology. I 3th Ed. Edinburgh: Churchill Liviiigstone, I 989;594.
25.Koneman EW. Allen SD. Janda WM, et al. Colour AlIas and Textbook of Diagnostic Microbiology. 5th Ed. Philadelphia: Lippineott.. 1997:510-600.
26.Bartlett JG. Clolstridium difficile infection: Pathopitysiology and diagnosis. Semin Gastrointest. Dia., 1997:8:12-21.
27.Borriello SP. Pathogenesis of Clostridium diffleile in the elderly. 3. Infect., 1991;22:310-12.
28.Bartlett 3G. Laboratory diagnosis of antibiotic associated colitis. Lab. Med., 1981: 2347-51.
29.Borriello SP. Pathogeneais of Clostridisim difficile infection of the gut. J. Med. Microbiol., 1990:33:207-1 5.
30.Sullivan NM, Pellett 5, Wilkins TD. Purification and eharacterisation of toxin A and B ofClostridium difficile. Infect Immun., 1982;53:1032-40.
31.Borriello SP, Barclay FE, Reed PJ, et al. Analysis of latex agglutination test for Clostridium difficile toxin A (d-i) and differentiation between C. difficile toxins A and B latex reactive protein. 3. Clin. Pathol., 1987:40:573-78.
32.Nash SV, Botirgeault R, Sands M. Colonie disease associated with a positive assay for Clostridium difficile toxin: A retrospective study. 3. Clin. Gastroenterol., 1997;25:476-79.
33.Jarvis W, Nunez-Montiel 0, Thompson F, et al. Comparison of baetenal isolation cytotoxieity assay and eounterimm unoelectrophoresis for the detection of Clostridium difficile and its toxin. 3. Infect. Dis., 1983;147:778.
34.Fille M, Larcher C, Dierich MP, et al. Evaluation of four methods for detection of Clostridium difficile toxin: Cytotoxin assay, culture, latex agglutination, and a new rapid immunoassay (C. diffleile toxin A test). Z. Gastroenterol., 1998:36:143-49.
35.Rothman SW. Technique for measuring 50% end points in cytotoxicity assay for Clostridiuin dilficile toxins. 3. Clin. Pathol., 1986:39:672-76.
36.Ryan RW, Kwasnik 1, Tilton RC. Rapid detection of Clostridiuin difiieile toxin in human feces. J. Clin. Mierobiol., l980;12:776-79.
37.Pepersack F, Labbe M, NonhotI C, et al. Use of gas-liquid chromatography as a screening test for toxigenie Clostridium difficile in diarrheal stools. 3. Clin. Pathol., 1983:36:1233-36.
38.Lyerly DM, Banoso LA, Wilkins TD. Identification of the latex test-reactive protein of Clostridium diffleile as glutamate dehydrogenase. 3. Clin. Mierobiol., 1991:29:2639-42.
39.Yolken RH, Whitcomb LS, Marien G, et al. Enzyme immunoassay for the detection of Clostridium difficile antigen. 3. Infect. Dis., 1981:144:378.
40.Laughon BE, Viaeidi RP, Gdovin SL, et al. Enzyme immunoassay for detection of Clostridium diffleile toxin A and B in fecal specimens. 3. Infect.Dis., 1984:149:781-88.
41.Krishnan C. Detection of Clostridium diffleile toxins by enzyme iminunoassay. 3. Hyg. Camb., I 986;96: 5-12.
42.Yiscidi R, Willey 5, Bartlett 3G. Isolation rates and toxigenie potential of Clostridium diffleile isolates from various patient population. Gastroenterol.,1981:81:5-9.
43.Gerding DN. Disease associated with Clostridium diffleile infection. Ann. Intern. Med., 1989;110:255-57,
44.Butterworth SA, Koppert E, Clarke A, et al. Recent trends in diagnosis and treatment of Clostridium difficile in a tertiary care facility. Am. 3. Surg.,1998:175:403-406.
45.Karlstrom 0, Fryklund B, Tullus K, et al. A prospective nationwide study of Clostridiuni difficiele Associated diarrhoea in Sweden. Clin. Infect. Dis.,1998:26: 14 1-45.
46.Gebhard RZ, Gerding DN, Olson MM, et al. Clinical and endoseopic findings in patients early in the course of Clostridium diffleile associated pseudomembranous colitis. Am. 3. Medicine, 1985;91:4548.
47.Manabe YY, Yinetz 3M, Moore RD, et al. Clostridium diffleile colitis: An efficient clinical approach to diagnosis. Ann. Intern. Med., 1995:123:835-40.
48.Issack MI, Elliott TSJ. Clostridium difficile carriage after infection. Lancet,1990;336:610-11.
49.Johnson 5, Clabots CR, Linn FV. et al. Nosocomial Clostridium difficile colonization and disease. Lancet, 1990:336:97-100.
50.Pavia AT. Clostridium diffieile carriage and diarrhoea. ACP 3. Club, 1991 :24.
51.Fluit ADC, Maurice JHM, Wolfhagen-Verdonk PHT, et al. Non-toxigenic strains of Clostridium difficile lack the genes for both toxin A and toxin B. 3. Clin. Mierobiol., 1991:29:2666-67.
52.Clabots CR, Johnson S. Olson MM, et al. Acquisition of Clostridium difficile by hospitalized patients: Evidence for colonized new adnaisaions as a source of infection. 3. Infect. Dis., 1992;166:561-67.
53.Schleupner MA, Gamer DC, Sosnowski KM. et al. Concurrence of Clostridium difficile toxin A enzyme-linked immunosorbent assay, feeal lactoferrin assay and clinical criteria with C. diffleile eytotoxin titer in two patients cohorts. 3. Clin. Mierobiol., 1 995;33:1755-59.
54.AIfa MJ, Du T, Beda G. Survey of incidence of Clostridium diffleile infection in Canadian hospitals and diagnostic approaches. 3. Clin. Mierobiol.,1998:36:2076-80.
55.Riederer KM, Lawson P, Held MS. et al. Diagnosis of Clostridium diffleile associated diarrhea: Comparison of three rapid methods employing different markers for detection. Can. 3. Microbial., 1995:41:88-91.
56.Jung K, Aronsson B. Rapid diagnosis of Clostridium diffmcile associated diarrhea using a latex agglutination test. APMIS, 1990;98:652-54.
57. Delmee M. Clostridium diffleile infection in health care workers. Lancet, l989;I 1:1095.

Journal of the Pakistan Medical Association has agreed to receive and publish manuscripts in accordance with the principles of the following committees: