2007. To Treat or Not To Treat: Does a More Sensitive and Specific Testing Methodology Make the Treatment Decision More Clear?
Session: Poster Abstract Session: Diagnostics: Bacteria and Mycobacteria
Saturday, October 6, 2018
Room: S Poster Hall
  • Poster ID Week 10-2018.pdf (367.8 kB)
  • Background:

    Clostridium difficileInfection (CDI) is a leading cause of infectious diarrhea in healthcare settings in the United States. Accurate testing methodology provides guidance to clinicians as to when to treat. Our study was designed to determine if more sensitive testing methodology implemented in 2013reduced unnecessary treatment of hospital associated diarrhea (HAD).


    In 2012, patients with HAD were tested with the less sensitive testing method of C. difficile Toxin Assay by EIA. In 2013, a three step algorithm incorporated CDI glutamate dehydrogenase antigen (GDH) in combination with an enzyme linked immunoassay for Toxin A and B was introduced. Those samples with discrepant results(positive on only one of the two) were considered indeterminate and subjected to the nucleic acid amplification test (NAAT) for CDI genes. In a retrospective chart review of HAD, we assessed the decision to treat based on the lab results available at the time in the pre-algorithmic and post-algorithmic periods. Multiple demographic factors and comorbid conditions were analyzed to provide clues to why the patient may have had continued treatment despite negative assays.


    The rateof treated patients despite negative CDI testing in the pre-algorithm period was 59% (118/444) and 41% (82/249) in the post-algorithm period (p=0.0765). A multiple logistic regression analysis was done for all tested factors. The factors that led to treatment despite negative testing in both time periods included: organ transplantation (p=0.0003), other immunosuppressive conditions (p=0.0447), prior hx of CDI (p=0.0021), longer length of stay (p=0.0105), and hx of hypertension (p=0.0173).


    While there was a downward trend toward holding CDI treatment in those with negative CDI testing as the more sensitive and specific algorithm was introduced it did not reach statistical significance. The higher risk patients were statistically more likely to be treated even if the testing was negative. Further efforts should be made to educate clinicians as to the accuracy of the testing methods so that appropriate antibiotic de-escalation can be achieved even in high risk patients with diarrhea.

    Bakri Kulla, MD, ID, Eastern Virginia Medical, Norfolk, VA, Patrick Haggerty, MD, ID, Eastern Virginia Medical School, Norfolk, VA, Shivanjali Shankaran, MD, Infectious Diseases, Eastern Virginia Medical School, Norfolk, VA, Hadiza Galadima, PhD, Sentara Healthcare Analytics and Delivery Science Institute (HADSI), Eastern Virginia Medical School, Norfolk, VA and Matthew Peworchik, NA, Eastern Virginia Medical School, Norfolk, VA


    B. Kulla, None

    P. Haggerty, None

    S. Shankaran, None

    H. Galadima, None

    M. Peworchik, None

    Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 3rd with the exception of research findings presented at the IDWeek press conferences.