Comparison of disease severity scoring systems for dairy cattle with acute coliform mastitis

Objective-To compare use of 4 disease severity scoring systems to predict bacteremia (yes vs no) and outcome (survived vs died or culled) in dairy cows with acute coliform mastitis (ACM). Design-Retrospective cohort study. Animals-99 dairy cows with ACM. Procedures-Cows were classified as having mild, moderate, or severe disease with a scoring system based on systemic disease signs alone (systemic severity score [SSS] system), a system based on local disease signs alone (local severity score [LSS] system), and 2 previously described systems based on a combination of local and systemic signs (local-systemic score 1 [LS1] and local-systemic score 2 [LS2] systems). Test performance was calculated to determine whether a severe disease classification could be used to predict bacteremia or outcome. Results-21%, 53%, 63%, and 38% of cows were classified as having severe disease with the SSS, LSS, LS1, and LS2 systems, respectively. For both bacteremia and outcome, sensitivity was highest for the LS1 system, but specificity and accuracy were highest for the SSS system. Examination of a scatterplot of true-positive rate versus false-positive rate for each of the scoring systems indicated that the SSS and LS2 systems were similar in their ability to correctly identify cows with bacteremia or an adverse outcome. Conclusions and Clinical Relevance-Results suggest that the SSS scoring system was better for identifying cows with bacteremia or an adverse outcome than was the LSS system and that the LS1 and LS2 systems were intermediate in their discriminatory abilities.