339 Comparison of length of stay in two treatment severity criteria for Clostridium difficile infections

Monday, October 22, 2012
Westin Diplomat Resort
Kyle A. Sobecki, BS, MS, PharmD, Patrick J. Gallegos, PharmD, BCPS and Bhavin K. Mistry, PharmD, BCPS
Akron General Medical Center, Akron, OH

Purpose: IDSA C. difficile (CDI) guidelines stratify severity based on expert opinion. IDSA cites Zar et al. as evidence of vancomycin superiority over metronidazole in severe disease. The goal of this study was to determine which criteria for stratifying CDI severity is a better predictor of health care utilization.

Methods: This retrospective review included adults ≥18 with a positive test for CDI. Patients were excluded if they received >4 doses of metronidazole or vancomycin and switched therapy, received combination metronidazole and vancomycin, or were prescribed other CDI medications. The primary outcome was the difference in length of stay (LOS) between the IDSA group compared to the Zar group in patients treated with metronidazole. Secondary outcomes examined vancomycin-treated patients, readmission rates, economic impact, and proportion prescribed appropriate therapy. Subgroup analysis excluded critical care patients.

Results: Mild patients treated with metronidazole in the IDSA group had a LOS of 4.7 days, while severe patients had a LOS of 4.0 days (p=0.41). Mild patients in the Zar group had a LOS of 4.4 days, while severe patients had a LOS of 4.1 days (p=0.74). The absolute difference between groups was not significant. There were no diffferences between patients treated with vancomycin. There were 9 readmissions; 6 were treated with metronidazole. LOS in non-critical floors were similar. Patients treated with metronidazole had a LOS of 4.1 days and cost $5,801, while patients treated with vancomycin had a LOS of 5.9 days and cost $8,348. Appropriate therapy occurred in <60% of patients. There were 21 total deaths, 16 of which were treated with metronidazole. There was a higher death rate in patients stratified as severe and undertreated with metronidazole (p=0.11).

Conclusion: There were no statistical differences between criteria groups.  There was a higher death rate in severe patients treated with metronidazole. Less than 60% of patients received apropriate therapy.