Jung Oak Kang1, Bo-Moon Shin2, Dongsoo Han3, Tae Yeal Choi1
Departments of 1Laboratory Medicine and 3Internal Medicine, Hanyang University College of Medicine, 2Department of Laboratory Medicine, Sanggye Paik Hospital, Inje University College of Medicine, Seoul, Korea
Background: Since the emergence of variant Clostridium difficile strains that fail to produce detectable toxin A, diagnostic kits targeted to detect toxin A only showed a considerable rate of false negative results. The aim of this study was to evaluate a toxins A and B (toxins A/B) detection kit recently marketed in Korea, and to compare toxin positive rates before and after introduction of the new kit.
Methods: The results of 5,783 toxin A assays performed during the 7-year period from 2001 through 2007 were analyzed and compared them to the toxins A/B assay data of 519 samples obtained from January to June 2008 in a university hospital. An enzyme-linked fluorescent immunoassay for toxins A/B (VIDAS C. difficile Toxin A & B, bioMerieux SA, France: VIDAS CDAB) and PCR for toxin genes A/B were performed directly in 102 stool samples from hospitalized patients.
Results: The positive rates of toxin A assays tended downward annually from 2001 to 2007 (16.3%, 17.8%, 13.9%, 11.4%, 13.8%, 8.2%, and 5.8%, respectively), but increased to 12.1% in 2008 after changing to the toxin A/B detection kit. The concordant rate of the VIDAS CDAB kit with the PCR method was 82.4%. Compared to the PCR method, the sensitivity and specificity of the toxin A/B kit were 60.7% and 90.5% respectively.
Conclusion: Testing kits for C. difficile toxin A only could result in a misdiagnosis more frequently than the testing kit for toxins A/B. The sensitivity of the newly launched toxin A/B detection kit from bioMerieux SA needs to be improved, but it showed a good specificity. (Korean J Clin Microbiol 2008;11:107-111)