344. Attributable Burden of Hospital Onset Clostridium difficile Infection: A Propensity Score Matched Study
Session: Poster Abstract Session: Clostridium difficile - Epidemiology, Diagnosis, Treatment, and Prevention
Friday, October 21, 2011
Room: Poster Hall B1
Background: Patients with hospital onset Clostridium difficile infections (CDIs) tend to be  severely ill and have high exposure risk. These baseline characteristics confound their clinical and economic outcomes. We sought to examine the attributable in-hospital mortality, length of stay (LOS), and cost of CDI using the propensity score matching method.

Methods: We analyzed clinical data between 2007 and 2008 from 6 US hospitals. We defined non-recurrent CDI cases among adult inpatients by a positive C. difficile toxin assay result >8 weeks following any previous positive result. The specimen collection day was >48 hours from the in-patient admission. We developed a CDI propensity model and matched CDI cases with non-CDI cases on their CDI propensity scores. We further matched patients within the same hospital, in the same principal disease group, and with similar length of lead time from admission to the onset of CDI. We used 1,000 bootstrap iterations to construct 95% confidence intervals (CI) for attributable effect of the CDI.

Results: Among 78,067 discharges during the study period, 310 CDI cases were identified. The CDI propensity model fit well (AUROC=0.80). The propensity score matching resulted in 95% matched rate (295 pairs). The pre-matching median age was 74 vs. 67 yrs, P<0.0001 for CDI vs. non-CDI cases. The post-matching median age was 75 vs. 73, P=0.86, respectively. The corresponding ICU admission was 36% vs. 16%, P<0.0001 for pre- and 37% vs. 41%, P=0.21 for post- matching cohorts. Other post-matching covariates also balanced well. Compared with the matched non-CDI cohort, CDI patients had higher mortality (11.2% vs. 6.8%, P=0.05); longer LOS (median [IQR] days:12 [8, 21] vs. 11 [8, 17], P<0.01); and higher cost (median [IQR]: $19,311 [$10,361, $37,907] vs. $14,572 [$8,906, $32,903], P<0.001). The estimated attributable effect of CDI for mortality was 4.2% (CI: -0.3%, 8.5%); for LOS, 2.2 days (CI: 0.7, 4.0); and for cost, $5,823 (CI: $1,477, $10,916).

Conclusion: The hospital onset CDI incurs additional attributable mortality, LOS, and cost burden, when compared with matched patients with similar CDI exposure risk and the lead time of hospitalization from admission, in the same primary clinical condition, and other baseline characteristics.

Subject Category: N. Hospital-acquired and surgical infections, infection control, and health outcomes including general public health and health services research

Ying P. Tabak, PhD1, RS Johannes, MD, MS2, Xiaowu Sun, PhD1, Marya Zilberberg, MD, MPH3, Linda Hyde, BS, RHIA1 and L. Clifford McDonald, MD4, (1)MedMined Services, Clinical Research, CareFusion, Marlborough, MA, (2)MedMined Services, Clinical Research, CareFusion and Harvard Medical School, Marlborough, MA, (3)Univ. of Mass. & EviMed Research Group, LLC, Amherst, MA, (4)CDC, Atlanta, GA


Y. P. Tabak, CareFusion: Employee, Salary

R. Johannes, CareFusion: Employee, Salary

X. Sun, CareFusion: Employee, Salary

M. Zilberberg, None

L. Hyde, CareFusion: Employee, Salary

L. C. McDonald, None

Findings in the abstracts are embargoed until 12:01 a.m. EST Thursday, Oct. 20 with the exception of research findings presented at IDSA press conferences.