Intended for healthcare professionals


Increased risk of diarrhoea caused by Clostridium difficile in elderly patients receiving cefotaxime

BMJ 1995; 311 doi: (Published 18 November 1995) Cite this as: BMJ 1995;311:1345
  1. M Impallomeni, consultanta,
  2. N P Galletly, senior house officera,
  3. S J Wort, senior house officera,
  4. J M Starr, senior registrara,
  5. T R Rogers, professorb
  1. aDivision of Geriatric Medicine, Department of Medicine, Royal Postgraduate Medical School, Hammersmith Hospital, London W12 0HS
  2. bDepartment of Infectious Diseases and Bacteriology, Royal Postgraduate Medical School
  • Accepted 8 September 1995

Clostridium difficile is a Gram positive anaerobic spore forming bacillus whose pathogenicity is related to exotoxin production in the large bowel. This may result in disease ranging from trivial diarrhoea to life threatening pseudomembranous colitis. Between 1982 and 1993 there was a 15-fold national increase in reported C difficile infections that was most marked in patients aged over 65.1 C difficile diarrhoea is almost exclusively acquired in hospital and strongly associated with the use of broad spectrum antibiotics.1 In 1993 the British Thoracic Society recommended cefotaxime and cefuroxime as first line antibiotics for treating severe community acquired pneumonia of unknown cause in adults.2 As a result the use of cefotaxime increased 20-fold in our unit. After November 1993 we also saw an unexpected increase in the incidence of C difficile diarrhoea, and we therefore sought to determine whether the two events were related.

Subjects, methods, and results

The geriatric unit at Hammersmith Hospital has 46 beds. We reviewed the clinical notes of all our patients with C difficile diarrhoea from April 1993 to November 1994. C difficile diarrhoea was defined as the passing of unformed stools in which C difficile toxin A was detected using a commercial enzyme immunoassay (Meridian Diagnostics Inc). In the case of a relapse only the first episode was counted. Affected patients were isolated or nursed together in a bay of the ward. “Notional courses” were used to estimate the total number of courses of each antibiotic (a seven day course of the most commonly prescribed dose regimen).3

From 1 April 1993 to 30 November 1994 1037 patients aged over 65 (median 83.8 years) were admitted; 43 (15 men) developed C difficile diarrhoea after antibiotic treatment. The average length of stay for these patients was 62 days, compared with 21 days for the whole group. Relapse of C difficile diarrhoea occurred in 11 of the 43 patients, and 18 (42%) died during their hospital admission; overall mortality in the unit was 25%. Two of the 43 patients were readmitted; data from the second admission were excluded from the results.

The monthly incidence of new cases of C difficile diarrhoea seemed to be strongly related to monthly expenditure on cefotaxime (figure). Expenditure on other antibiotics did not have such clear temporal relation. Moreover, the highest relative risk for developing diarrhoea among patients receiving an antibiotic compared with those not receiving it occurred with cefotaxime (7.2, 95% confidence interval 3.9 to 13.2), followed by cefuroxime (5.2 (2.9 to 9.45)), and erythromycin (2.8 (1.5 to 5.2)). No significant increased risk occurred with other antibiotics. Although many patients received combination therapy the data were not available to study the potential interaction between different agents.


A sudden increase in the incidence of C difficile diarrhoea followed a 20-fold increase in the use of cefotaxime. Infection control measures did not succeed in preventing new cases, which only decreased when the use of cefotaxime was stopped (figure). Nearly one in five patients who received cefotaxime developed C difficile diarrhoea.

Risk factors for the development of C difficile diarrhoea include increasing age; hospitalisation; malignancy; renal impairment; use of antibiotics, nasogastric feeding, laxatives, H2 antagonists; and general disability. Many of these existed in our patients, but we could not firmly assess their risk ratios in our retrospective analysis.

C difficile diarrhoea has been reported following the administration of cefotaxime,4 but we are not aware of any report of such a rapid increase in cases related to its introduction. Another recent study did, however, show a similar relation with another broad spectrum antibiotic, clindamycin.5 We suggest cefotaxime should be used in elderly patients only if there is no suitable alternative.


Relation between use of cefotaxime (as measured by expenditure) and new cases of C difficile diarrhoea, April 1993 to November 1994


We thank Mrs Beryl Langfield, pharmacy department, for providing total and monthly expenditure on antibiotics in the geriatric unit.


View Abstract