Background
Metronidazole may still be an appropriate therapeutic option for mild Clostridium difficile infection (CDI) in select patients, but data are limited to guide clinicians in identifying these patients.
Methods
Our 2-stage study included a national cohort of Veterans with a first episode of mild CDI (2010–2014). First, among those treated with metronidazole, we identified predictors of success, defined as absence of all-cause mortality or recurrence 30 days posttreatment, using multivariable unconditional logistic regression. Second, among a subgroup of patients with characteristics predictive of success identified in the first stage, we compared clinical outcomes among those treated with metronidazole compared with vancomycin, using Cox proportional hazards models for time to 30-day all-cause mortality, CDI recurrence, and failure.
Results
Among 3656 patients treated with metronidazole, we identified 3282 patients with success and 374 patients without success (failure). Younger age was the only independent predictor of success. Age ≤65 years was associated with an odds of success 1.63 times higher (95% confidence interval [CI], 1.29–2.06) than age >65 years. Among 115 propensity score–matched pairs ≤65 years of age, no significant differences were observed between metronidazole and vancomycin (reference) for all-cause mortality (hazard ratio [HR], 0.29 [95% CI, .06–1.38]), CDI recurrence (HR, 0.62 [95% CI, .26–1.49]), or failure (HR, 0.50 [95% CI, .23–1.07]).
Conclusions
Among patients ≤65 years of age with initial mild CDI, clinical outcomes were similar with metronidazole and vancomycin. These data suggest that metronidazole may be considered for the treatment of initial mild CDI among patients 65 years of age or younger.