Failure of single-toxin assays to detect clostridium difficile infection in pediatric inflammatory bowel disease.

Abstract

The aims of this retrospective study were 1) to determine the ability of single-toxin assays for Clostridium difficile to detect infection among pediatric patients with inflammatory bowel disease (IBD) and 2) to determine the toxin assays routinely used by pediatric tertiary care hospitals in the United States. Stool specimens from patients with IBD (submitted from January, 1996, to August, 1999) were evaluated for the presence of C. difficile toxin A and toxin B. Toxin profile (toxin A alone, toxin B alone, toxin A and B together) was compared in positive specimens. A phone interview was conducted with representatives from laboratories in 22 pediatric hospitals to investigate which toxin assays were routinely used. A total of 697 specimens were submitted from 284 IBD patients. In all, 81 IBD patients (28.5%) had at least one documented infection. Toxin A assay failed to identify 41.5% of C. difficile infections. Toxin B assay failed to detect 34.9% of C. difficile infections. Toxin profile changed in 55% of patients with multiple infections. Of the hospitals surveyed, 59% did not test for both toxins. Single-toxin assays for C. difficile fail to detect a significant percentage of infections. The toxins identified during one infection are not predictive of the toxins identified in subsequent infections. Despite this, many pediatric hospitals do not routinely use both toxin assays to diagnose C. difficile infection. When infection is suspected, assays for C. difficile toxin A and toxin B should be requested.