On the Optimization of Clostridioides difficile Diagnostics Through RT-PCR Cycle Threshold Defined Zones of Disease Probability
Date
2022-02-02
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Clostridioides difficile is an opportunistic pathogen with a large burden of disease and no gold standard test. Quantitative polymerase chain reaction (qPCR) offers excellent sensitivity but overcalls clinical C. difficile infections (CDI) due to the prevalence of colonization. The hypothesis of this thesis is that the CDI qPCR results can be titrated to determine clinical CDI more accurately and aid in predicting disease severity. A cross-sectional study was conducted on suspected CDI patients evaluating if qPCR cycle threshold (Ct) can be correlated to probability of CDI. Latent class analysis (LCA) was employed with observed variables including four commercial qPCR tests, toxin detection by enzyme immunoassay, toxigenic culture, fecal calprotectin, and clinical diagnosis. Three defined zones as a function of qPCR cycle threshold (Ct) were identified: CDI likely (>90% probability), CDI equivocal (<90% and >10%), CDI unlikely (<10%). A model comprising toxigenic culture, clinical diagnosis, and toxin EIA demonstrated the best fitness. The following Ct cut-offs for 4 commercial test platforms delineated CDI probability zones: GeneXpert®: 24.00, 33.61; Simplexa®: 28.97, 36.85; Elite MGB®: 30.18, 37.43; and BD Max™: 27.60, 34.26. A prospective cohort study was conducted to investigate if these zones can be further correlated to indicators of severe CDI. Primary diagnosis, demographic data and indicators of disease severity were captured: white blood cell, creatinine, albumin, C-reactive protein, and hospital length of stay. A sub analysis was conducted evaluating a subset of the patient population attempting to isolate patients whose clinical variables were most influenced by CDI. No significant correlations were found between the clinical variables investigated and Ct values or Ct zones. This work establishes a method of using deployed diagnostics to allow clinicians to reduce overdiagnosis of CDI. Decreasing false positives could have broad impacts, increase targeted treatments, and decrease antibiotics. The average cost attributed to CDI for one patient is estimated at $11,917. LCA models predict that qPCR confirmation overdiagnoses patients in Calgary by at least 20.9%. If CDI confirmation were reduced by 20.9% this could equate to massive savings; Foothills Medical Center alone could save over $929,000 annually with no additional investment in laboratory infrastructure.
Description
Keywords
Clostridioides difficile, Diagnostics, Latent Class Analysis, Polymerase Chain Reaction
Citation
Doolan, C. P. (2022). On the Optimization of Clostridioides difficile Diagnostics Through RT-PCR Cycle Threshold Defined Zones of Disease Probability (Master's thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca.