|| Checking for direct PDF access through Ovid
We describe the proportion of health care facility-onset Clostridium difficile infection (HO-CDI) National Healthcare Safety Network laboratory-identified events at our facility that were deemed nontrue HO-CDIs. Reasons included testing in a patient without significant diarrhea or with recent laxative use, or delayed testing. Standardized infection ratios using only true HO-CDI in the numerator were improved compared with publically reported standardized infection ratios. A prioritization matrix identifies which clinical services could benefit most from directed diagnostic stewardship interventions.