2831 Altering Perceptions of In-Hospital Stool Testing for Clostridioides difficile Infection Reduces Inappropriate Prescribing of Antibiotics and Length of Hospital Stay

Abstract

INTRODUCTION: Clostridium Difficile Infection (CDI) occurs predominantly in the intestinal tract due to alteration of normal gut flora often due to antibiotics. Testing for CDI is common on the inpatient setting. However, we recognized that over 95% of ordered stool tests are negative for CDI which indirectly was associated with an increased length of hospital stay, the unnecessary use of antibiotics and placing patients on isolation precautions. A hospital initiative led by our team that included provider education and a special electronic order set increased the accuracy of testing for CDI at our institution. METHODS: A one-year retrospective study was done evaluating the number of symptomatic and asymptomatic patients that were tested by the toxin A/B EIA and PCR assays. Similar data was obtained from the following year after the initiation of our hospital initiative. The hospital initiative included education for all hospital prescribers as well as the creation of a special electronic order set including alerts and special restrictions implementing the Infectious Disease Society of American (IDSA) guidelines. CDI cases were reported and compared between the two groups. RESULTS: During the first study period, there were a total of 877 CDI Antigen & Toxin tests ordered for inpatients. This resulted in 712 or 81% negative results for both antigen & toxin. Out of 877 tests, 120 were found to have antigen positivity but negative for CD toxin (indeterminate results). These 120 patients were further analyzed by polymerase chain reaction (PCR). Only 45 out of 877 were truly positive for CDI antigen & toxin (5% of total ordered tests). During the second study period, post intervention a total of 828 total CDI tests were ordered. This result showed 678 CDI tests or 82% ordered tests were negative for CDI antigen & toxin. Out of 828, 101 tests were antigen positive but negative for toxin which were further analyzed by PCR. Only 49 resulted positive for CDI antigen & toxin (6% of total tests ordered with positive for CDI). CONCLUSION: Our hospital initiative that included guideline-based education for ordering physicians as well as implanting an electronic order set enforcing testing criteria was effective at identifying more true positive results of CDI. We can infer that the underlying cause for the majority of symptomatic patients resulted from an etiology other than CDI. Further practitioner education with stricter measures must be enforced to reduce excessive testing.