Economic Impact of Clostridium difficile Infection in a Multihospital Cohort of Academic Health Centers

Abstract

To assess the economic impact of Clostridium difficile infection (CDI) in a large multihospital cohort. Retrospective case-control study. Administrative claims data from 45 academic medical centers. A total of 10,857 patients who developed health care-associated CDI and were discharged between April 1, 2002, and March 31, 2007 (cases); each case patient was matched by hospital, age, quarter and year of hospital discharge, and diagnosis related group to at least one control patient who did not develop health care-associated CDI (19,214 controls). Patients with health care-associated CDI were identified by using a previously validated method combining the International Classification of Diseases, Ninth Revision, Clinical Modification code for CDI with specific CDI drug therapy (oral or intravenous metronidazole, or oral vancomycin). Costs were determined from charges by using standardized cost:charges ratios and were adjusted for age, All Patient Refined-Diagnosis Related Group (APR-DRG) severity of illness level, race, and sex with use of multivariable linear regression. The adjusted mean cost for cases was significantly higher than that for controls ($55,769 vs $28,609), and adjusted mean length of stay was twice as long (21.1 vs 10.0 days). The interaction between CDI and APR-DRG severity of illness level was significant; the effect of CDI on costs and length of stay decreased as severity of illness increased. This large CDI economic evaluation confirms that health care-associated cases of CDI are associated with significantly higher mean cost and longer length of stay than those of matched controls, with the greatest effect on costs at the lowest level of severity of illness.