The spectrum of presentation of Clostridium difficile infections (CDIs) ranges from mild, requiring oral antibiotics only, to fulminant, requiring intensive care admission and emergent total abdominal colectomy. Unfortunately, the clinical prediction rules for identifying which patients are at increased risk for poor outcomes are suboptimal; many have not been validated or have low diagnostic accuracy.1 Furthermore, none of the available prediction rules include variables with a mechanistic basis. By contrast, in this issue of JAMA Surgery, Kulaylat et al2 developed a model based on findings that were translated from a mouse model. They hypothesized that admission eosinopenia, defined as a peripheral eosinophil level of 0.0 cells/μL (to convert to ×109 per liter, multiply by 0.001), is associated with worsened outcomes. Using a training and validation cohort, they determined that eosinopenia is an independent predictor of inpatient mortality (odds ratio, 2.01; 95% CI, 1.08-3.73). Additionally, it is associated with an increased need for admission to a monitored setting, vasopressors, and emergent total colectomy.