Prospective assessment of two-stage testing for Clostridium difficile

Commonly used immunoassays have limitations as stand-alone tests for the diagnosis of Clostridium difficile infection (CDI). In particular, the specificity of these assays means that these tests generate a relatively large number of false-positive results. We introduced a two-stage regimen for CDI as routine. Unformed stool samples received in our laboratory were initially tested with a Meridian Premier enzyme immunoassay (EIA) and positive samples were retested with reference testing methods (toxigenic culture and cell cytotoxicity assay). Clinicians received diagnostically useful information on the day that the sample arrived in the laboratory, with definitive negative and provisional positive results made available. We reviewed the first 3643 unformed stool specimens of which 158/3643 (4.3%) were provisionally positive by EIA. Of the 158 samples that were EIA positive, 119 were confirmed as being positive by at least one of the reference methods, giving a positive predictive value in this population of 75% (95% confidence interval: 67.6-81.7%). Comparison of the optical density values of the EIA lying between true and false-positive results suggests that the introduction of a second cut-off value would improve diagnostics. A test with two cut-offs would give the following results: 'positive', 'negative' and 'indeterminate result, please perform confirmatory test'. This algorithm was a simple and cost-effective method to immediately improve diagnostics, but there is an urgent need for further research in laboratory diagnosis for CDI.