While advantageous by casting a wider diagnostic net, multiplex panels can be problematic if the pretest probability is low. A significant increase in reported Clostridioides difficile infections (CDI) was noted at our institution following introduction of a multiplex comprehensive GI (CGI) panel which includes an analyte for C. difficile. Due to these concerns, the C. difficile analyte result was suppressed when reporting and providers were advised to order a standalone C difficile PCR (CDPCR) test if CDI was a concern. The objective of this study was to investigate concerns of false positive C. difficile results from the CGI panel.
C. difficile diagnostic practices were prospectively evaluated from April to August 2017. Patient charts were reviewed in response to a positive C. difficile analyte on the CGI panel. CDPCR results were reviewed if ordered. If not ordered, chart review and discussion with the provider was conducted to investigate clinical suspicion for CDI. The results were analyzed to examine the performance of the C. difficile analyte on the CGI panel.
Overall, a total of 1611 CGI panels were performed with C. difficile being detected in 156 specimens. Of these positive results, a sub analysis was performed on 123 positive specimens for whom complete data was available. A CDPCR was performed in 80 (65%) of these specimens. Among those, only 44 (55%) were CDPCR positive, 22 (28%) were CDPCR negative (likely a false positive CGI result), and 14 (17%) were rejected because of specimen consistency. For the remaining 43 C. difficile positive CGI panel specimens that did not have an accompanying CDPCR, 7 were in children below 2 years of age. Direct provider discussion occurred in the remaining 36 cases. Providers declined CDPCR testing in 24 of those cases due to a lack of clinical concern.
The use of the CGI panel for C. difficile led to over diagnosis of CDI. This could have significant consequences for clinical care and the reporting of hospital acquired infections.
J. A. Ribes, None
S. Cotner, None
K. L. Wallace, None
D. Forster, None