Clinical Update
E. coli, antibiotics and hemolytic-uremic syndrome in children

CMAJ 2000;163(4):438

Background

In up to 15% of North American children infected with Escherichia coli O157:H7, hemolytic-uremic syndrome (HUS) develops because of systemic absorption of Shiga toxins produced by the organism.¹ Although antibiotics have been shown in vitro to enhance release of these toxins from injured bacteria,² their effect on the development of HUS in people infected with E. coli O157:H7 is unknown.

Question

In children with stool cultures positive for E. coli O157:H7, is antibiotic therapy associated with an increased risk of HUS, independent of the severity of the initial diarrheal illness?

Design

The authors of this prospective observational study¹ assembled a cohort of 71 children in whom stool cultures obtained within 7 days of the onset of acute diarrheal illness were positive for E. coli O157:H7. Subjects were less than 10 years old and were identified through a cooperative network of 47 laboratories in the United States. Demographic data and clinical features of the illness (vomiting, bloody diarrhea and fever) were obtained through a questionnaire completed by each child's caregiver. Details concerning antibiotic therapy were validated by the treating physician or through examination of medical records.

Complete blood counts and blood urea nitrogen and serum creatinine levels were measured daily for 14 days from the onset of their symptoms, or until the development of HUS (defined as a hematocrit of less than 30% with evidence on peripheral blood smear of red blood cell destruction, a platelet count of less than 150 × 10⁹/L and a serum creatinine level exceeding the upper limit of normal).

Multivariate logistic regression was used to determine the relative risk of HUS associated with the administration of antibiotic therapy, adjusted for markers of disease severity.

Results

The study's criteria for HUS were met in 10 (14%) of the 71 subjects. HUS developed in 5 (56%) of 9 children who received antibiotic therapy and in 5 (8%) of 62 children who did not (p < 0.001). Children who received antibiotic therapy were comparable to those who did not with respect to age, sex and the baseline clinical and laboratory features of their illness.

Multivariate analysis showed that the risk of HUS was associated significantly with what were considered to be 2 surrogate markers of disease severity: initial peripheral white blood cell count (p = 0.02) and time elapsed from the day of symptom onset to the day on which stool cultures were obtained (p = 0.008). Risk was directly proportional to the white blood cell count and inversely proportional to the interval between onset of illness and stool cultures. It was inferred from the latter finding that patients with more severe illness were evaluated earlier than those with less severe illness. The relative risk of HUS among the children who were given antibiotic therapy, when adjusted for these 2 variables, was 17.3 (95% confidence interval 2.2–137, p = 0.007).

Commentary

Although this study is subject to the selection and confounding biases inherent in observational research, it offers compelling evidence of a link between antibiotic therapy and the development of HUS in diarrheal illness caused by E. coli O157:H7. The strength of the association and the biologically plausible effect of antibiotics on the amount of Shiga toxin available for absorption from the intestine support the inference of causality.

Implications for practice

Although HUS develops in patients infected with E. coli O157:H7 with or without antibiotic treatment, it occurs much more frequently when antibiotics are given. The findings of this study strongly suggest that these drugs should be withheld in children with acute diarrheal illness until stool cultures confirm growth of an organism for which antibiotic therapy is indicated (e.g., Campylobacter pylori). — Donald Farquhar

The Clinical Update section is edited by Dr. Donald Farquhar, head of the Division of Internal Medicine, Queen's University, Kingston, Ont. The updates are written by members of the division.

1.	Wong CS, Jelacic S, Habeeb RL, Watkins SL, Tarr PI. The risk of hemolytic-uremic syndrome after antibiotic treatment of Escherichia coli O157:H7 infections. N Engl J Med 2000;342:1930-6. [MEDLINE]
2.	Walterspiel J, Ashkenazi S, Morrow A, Cleary T. Effect of subinhibitory concentrations of antibiotics on extracellular Shiga-like toxin I. Infection 1992;20:25-9. [MEDLINE]

© 2000 Canadian Medical Association or its licensors