BACKGROUND: The glutaraldehyde test (GT) is used as a point-of-care test to detect inflammatory disease in cattle. OBJECTIVES: Describe the relationship between GT and fibrinogen as well as GT and total protein (TP) and determine the ability of the GT to predict death in cattle with and without sepsis. ANIMALS: A total of 367 hospitalized, adult cattle. METHODS: Clinical and laboratory findings, diagnoses, treatments, and outcome were recorded. Spearman correlation was used to assess the relationship of the GT with fibrinogen and GT with TP in septic and non-septic cattle. Likelihood ratios (LHRs) were calculated, and a receiver operating characteristic (ROC) curve was constructed to identify the most appropriate threshold predictive of death. Logistic regression models were used to assess the predictive ability of the GT at this threshold. RESULTS: The negative correlation between GT and fibrinogen was stronger in non-septic cattle (rho = -0.7
p <
0.002) compared with septic cattle (rho = -0.59
p <
0.002). Similarly, GT and TP showed stronger correlation in non-septic (rho = -0.52) than in septic cattle (rho = -0.34
p <
0.002, respectively). Non-septic cattle with a GT of ≤ 3 min were 2.7 times (95% confidence interval [CI], 1.79-4.02) more likely to die during hospitalization than cattle with a GT >
3 min. No significant associations were identified between GT results and death in septic cattle. CONCLUSIONS: The GT can be recommended as a valuable tool for predicting death in non-septic cattle but should not be used in acutely septic cattle.