Epidemiology of Clostridium difficile Infection in an Integrated Healthcare System Over a 10-year Period
Methods: We queried Intermountain Healthcare’s electronic data warehouse for C. diff tests collected from >185 clinics and 22 hospitals between 2003 and 2012. Lab defined cases of CDI were categorized as healthcare associated (HA) (positive test >48 hours after admission or ≤30 days after hospital discharge), CA (no hospitalization ≥90 days prior to a positive test) or recurrent (positive test 14-90 days after a prior positive test). Positive tests ≤14 days of an index positive were excluded. Enzyme immunoassay (EIA) for Toxins A/B was used from 2003–2010 and nucleic acid amplification testing (NAAT) replaced EIA in October 2010. Incidence rates for HA and CA CDI were calculated as episode/10,000 hospital discharges and episode/10,000 patient encounters, respectively.
Results: The number of C. diff tests ordered increased 81% from 2003 (9,613 tests) to 2012 (17,385 tests). During the EIA period, 7–10% of samples tested each year were positive compared to 15% with NAAT. During the entire study period 10,689 CDI episodes were identified in 8,993 patients. HA accounted for 35% (n=3,757) of the episodes, CA 48% (n=5,108), recurrent 11% (n=1,197) and 6% (n=627) were indeterminate. From 2003 to 2012, there was a decline in proportion of HA CDI (46% vs. 26%, p<0.05) with an increase in CA (41% vs. 53%, p<0.05) and recurrent (9% vs. 15%, p<0.05) CDI (Figure A). The incidence of HA CDI increased between 2003 and 2006 (22.8 vs. 28.6/10,000 hospital discharges) and subsequently decreased through 2010. By contrast, incidence of CA CDI increased 53% from 2003-2010 (Figure B). When NAAT tests were implemented, both HA and CA rates increased but this may in part be related to increased test sensitivity. The proportion of recurrent CDI (10% vs 14%) also increased with NAAT use.
Conclusion: We detected an increase in rates of CA CDI throughout the study period. HA CDI rates declined between 2006 and 2010. With use of NAAT CDI rates increased however trends are not yet clear.
B. K. Lopansri,
E. Stenehjem, None
K. Dascomb, None
J. Shumway, None
S. Giddings, None
A. Pavia, None
J. Burke, None
See more of: Poster Abstract Session