Red cell distribution width is a predictor of all-cause mortality in hospitalized dogs

Objective: To determine whether RBC distribution width (RDW) is associated with an increased odds of mortality in hospitalized dogs and cats. Design: Retrospective, single-center study; data collected from 2007 to 2017. Setting: University teaching hospital. Animals: Six thousand six hundred and sixty-one animals (5183 dogs and 1478 cats). Interventions: None. Measurements and Main Results: Medical records were identified from animals presented to the emergency service and admitted to the ICU with a CBC and serum biochemistry performed on admission. Patients were stratified into quintiles based upon presenting RDW, and logistic regression modeling was performed to evaluate the relationship between RDW and in-hospital mortality. Canine patients with a presenting RDW in the upper fourth and fifth quintiles had an increased odds of all-cause in-hospital mortality (p < 0.0001). Specifically, dogs in the upper fifth quintile had a 2.1-fold greater odds of death compared to dogs in the first quintile, and dogs in the upper fourth quintile had a 1.9-fold greater odds of death compared to dogs in the first quintile. This relationship remained significant after adjusting for age, sex, key laboratory values excluding HCT, medical versus surgical disease, and diagnosis category. This relationship was no longer significant with the inclusion of HCT. No significant association was identified between presenting RDW and in-hospital mortality in cats. Conclusions: Hospitalized dogs with higher RDW on presentation to the emergency service have greater odds of all-cause in-hospital mortality compared to dogs with a lower RDW. A similar association between RDW and mortality was not found in cats. Further studies are warranted to assess the usefulness of this biomarker for specific diseases in dogs and to better understand the mechanisms driving the association between increased RDW and mortality in canine patients.