An unusual cause of false-positive results with the BD GeneOhm Cdiff assay.

The introduction of commercially available in vitro diagnostic (IVD) PCR assays into the clinical microbiology laboratory has facilitated enhancements to patient care, improving diagnostic sensitivity and specificity (1–3). Recently, PCR assays for the detection of the Clostridium difficile toxin B gene have become available. A number of these commercial assays require little or no special expertise for performing or interpreting the results of the real-time PCRs. However, laboratories adopting these assays need to be aware of potential problems that may occur with molecularly based IVD assays. To validate an inexpensive, in-house stool extraction method for use with the BD GeneOhm Cdiff assay (BD Diagnostics, Quebec, Canada), we compared the PCR results of our in-house extraction with those of the BD GeneOhm Cdiff kit’s extraction. During this validation, we found one specimen that was positive using the BD kit’s extraction but tested negative with our in-house extraction. It was noted that the initially positive sample was tested in the A9 I-CORE module of the Cepheid SmartCycler (Cepheid, Sunnyvale, CA) block. Multiple repeat testing with the original stool specimen using both the BD and in-house extraction reagents were all negative in non-A9 I-CORE modules. Further investigation showed that 52 different patient specimens tested in the A9 I-CORE module were positive between 1 October 2011 and 27 February 2012 in 113 runs (positivity rate of 46%). This was in contrast to the positivity rate for all the other I-CORE modules in SmartCycler block A, which had an average positivity of 20% (range, 10.1% to 34.2%). There were no warning/error messages with any of the runs, and there were no other false-positive results with I-CORE modules in block B or in any other position apart from A9 in block A. After we were granted access from the manufacturer to review the instrument raw data, all amplification curve data were assessed. A total of 37/52 (71.2%) specimens reported to be positive when tested in the A9 I-CORE module had abnormal curves that were not consistent with the result of a positive real-time PCR (Fig. 1). Thus, a total of 37 patient specimens had been falsely reported as positive. Replacement of the A9 ICORE module did not correct the problem. The problem was resolved with replacement of the entire SmartCycler block. The fluorescence threshold and algorithm used for assessing the amplification curves are set to maximize the assay’s sensitivity, but this may have contributed to the false-positive results. The second derivative maximum, calculated from the derivative of the maximum change in signal (i.e., slope) corresponding with the