PURPOSE: To evaluate the incidence of Clostridium difficile infection, as diagnosed by polymerase chain reaction (PCR) in a cohort of hospitalized intensive care unit (ICU) and non-ICU patients to determine ordering trends and utility of testing for infection.
METHODS: Data was collected retrospectively on all patients that underwent Clostridium difficile testing at a single, tertiary care hospital from January 1, 2008 to May 3, 2011. Demographic data was queried from the electronic medical record to include age, sex, race, co-morbidities, ordering specialty, length of stay, and APR mortality. Data was analyzed using a chi-square test with significance defined as p<0.05.
RESULTS: A total of 18,877 Clostridium difficile PCR tests were performed over the study period. 1269 tests were ordered in the ICU, and 17,608 tests were ordered on non-ICU patients. Of the tests ordered, 162 and 2726 tests were positive in the ICU and non-ICU groups. Infection rates were 5.94% and 6.85% respectively for both ICU and non-ICU patients. We did not find a statistically different rate of infection between the two groups, but there was a trend towards increased infection rates in the non-ICU group (p=0.0861).
CONCLUSIONS: The incidence of Clostridium difficile was low in our cohort of patients, and the data demonstrates that there was a high rate of testing with a low rate of infectivity in both ICU and non-ICU patients. We identified a yearly incremental increase in the rate of testing and the rate of positive tests in both groups, but non-ICU testing far outnumbered ICU testing.
CLINICAL IMPLICATIONS: Clostridium difficile infection is associated with high morbidity and mortality. Testing is often performed prior to the initiation of antibiotics. There appears to be an increase in the rate of infection over time in our data set, which may be explained by the increased rate of testing; however, the incidence of Clostridium difficile infection in both non-ICU and ICU patients is low. We conclude that this test may be over-utilized within our institution.
DISCLOSURE: The following authors have nothing to disclose: Brian Barnett, Marcia Henderson, Carl Boethel
No Product/Research Disclosure Information