Main Menu Browse by Day or Program Author Keywords

101 A Comparison of Unnecessary Antibiotic Use Associated with a Two-step Algorithm for Detection of Clostridium difficile Infection (CDI) Using Either Cell Culture Cytotoxicy Neutralization Assay (CCNA) or PCR Toxin Gene Detection (PCR) Confirmatory Tests

Friday, March 19, 2010
Grand Hall (Hyatt Regency Atlanta)
Abigail A. Lenhart, MD , Johns Hopkins Hospital, Baltimore, MD
Brandon Trollinger, Pharm, D , Johns Hopkins Hospital, Baltimore, MD
Kentra Tillman , Johns Hopkins Hospital, Baltimore, MD
Edina Avdic, Pharm, D, MBA , Johns Hopkins Hospital, Baltimore, MD
Karen C. Carroll, MD , Johns Hopkins Hospital, Baltimore, MD
Lisa L. Maragakis, MD, MS , Johns Hopkins Hospital, Baltimore, MD
Sara E. Cosgrove, MD, MS , Johns Hopkins Hospital, Baltimore, MD
Background: A two-step algorithm with an initial glutamate dehydrogenase (GDH) antigen screening test followed by a CCNA has been demonstrated to be sensitive and specific relative to toxin enzyme immunoassays in the detection of CDI.  CCNA has a slow result time, often 24-48 hours, which leads physicians to initiate the use of antibiotics prior to receipt of the confirmatory test results.  Our institution recently implemented use of rapid PCR toxin gene detection method in place of the CCNA to improve time to detection of CDI in the laboratory.

Objective: To evaluate unnecessary antibiotics used to treat suspected cases of CDI before and after the initiation of a more rapid PCR toxin gene detection method (PCR) in place of the CCNA.

Methods: 100 GDH positive patients were identified both in the time period when the CCNA test was used for confirmation (1/09-3/09) and in the period when PCR test was used for confirmation (6/09-8/09). Data on demographics; co-morbidities; receipt of antibiotics, including metronidazole (MTZ) and oral vancomycin (VANC); and symptoms were collected.  Patients were excluded if they were CDI positive on their respective confirmatory test or if they had non-CDI indications for MTZ. The number of hours the remaining patients were treated with MTZ or oral VANC as well as total antibiotic hours while in hospital were assessed.  Comparison of mean duration of antibiotics was by t-test.

Results: From the first period, 59/100 patients were GDH screen positive and CCNA negative and 36/59 (61%) received either MTZ or oral VANC.  7/59 (12%) were on MTZ for another indication and were excluded from the analysis.  The mean total antibiotic time for this first cohort of 52 patients was 47.13 hours. Mean MTZ time was 30.47 hours.  From the second period, 42/100 patients were GDH positive and PCR negative and 18/42 (43%) received either MTZ or oral VANC.  2/42 (5%) were on MTZ for another indication and were excluded from the analysis.   The mean total antibiotic time for this second cohort of 40 patients was 11.35 hours. Mean MTZ time was 10.07 hours.   The reduction in total antibiotic time with the PCR toxin gene detection method was statistically significant (p=0.03) as was the reduction in MTZ time (p=0.03).

Conclusions: The two-step algorithm for detection of CDI has become standard practice in many institutions; however, because of the time required to obtain CCNA results, it appears to lead to significant excess antimicrobial use. Substitution of PCR as the confirmatory test can decrease this excess antimicrobial use in patients suspected to have CDI by providing more rapid results.  As both MTZ and oral VANC can trigger CDI, reducing exposure to these agents in patients who are GDH positive is desirable.