Characterization of continued antibacterial therapy after diagnosis of hospital-onset Clostridium difficile infection: implications for antimicrobial stewardship.

Study Objectives: To determine the proportion of hospitalized adults with hospital-onset Clostridium difficile infection (CDI) who continued to receive concomitant non-CDI antibacterial agents, to characterize the antibacterial therapy that these patients received before and after the diagnosis of CDI, and to compare hospital outcomes between those patients who did and those who did not have their previous antibacterial therapy discontinued after CDI diagnosis.

Design: Retrospective cohort study.

Data Source: Drug use and administrative discharge data from 42 United States academic medical centers.

Patients: A total of 5968 adult inpatients with hospital-onset CDI between January 1, 2002, and June 30, 2006.

Measurements And Main Results: We characterized patient-level antibacterial agent use before and after CDI diagnosis. Overall, 3479 patients (58.3%) continued antibacterial therapy for 2 or more days after CDI diagnosis (interhospital range 6.7-72.2%). Although the number of different antibacterial agents received in the week preceding CDI diagnosis was positively associated with continued antibacterial therapy, the relationship between continuation and severity of illness was statistically significant but nonlinear. Patients who were receiving oral vancomycin alone were less likely to have antibacterial therapy continued (28/61 patients [45.9%]) than patients receiving metronidazole alone (1154/2333 patients [49.5%]) or receiving both metronidazole and oral vancomycin (2297/3576 [64.2%]). After adjusting for confounders, patients who continued to receive antibacterial therapy had a 62.7% (95% confidence interval [CI] 48.6-78.0%, p<0.001) longer length of hospital stay after CDI diagnosis than those who did not continue therapy; the adjusted odds of mortality and odds of readmission were 1.7 (95% CI 1.4-2.1, p<0.001) and 1.2 (95% CI 1.1-1.5, p=0.025) times higher, respectively, with continued antibacterial therapy.

Conclusion: A majority of patients with CDI continued to receive antibacterial agents after their CDI diagnosis, although the interhospital range was large. Compared with patients who did not continue therapy, hospital length of study, mortality, and subsequent admissions among patients who continued their antibacterial therapy remained significantly higher after adjusting for confounders. The adverse outcomes associated with continued therapy likely reflect the severity of the underlying primary infection and/or a poorer response to CDI therapy, suggesting an opportunity for antimicrobial stewardship programs to make important contributions to patient care.