Sensitivity and specificity of fasting ammonia and serum bile acids in the diagnosis of portosystemic shunts in dogs and cats

Background Portosystemic shunt (PSS) is the most common cause of hepatic encephalopathy in dogs and cats. Fasting ammonia and serum bile acids (SBA) are used to diagnose PSS, but their true sensitivity and specificity have not been fully evaluated, especially in cats. Objectives The purpose of this study was to determine the diagnostic accuracy of fasting ammonia and SBA concentrations in the diagnosis of PSS in dogs and cats and to compare diagnostic accuracy between species. Methods A retrospective analysis of data from 373 dogs and 85 cats presented to the clinic from 1996 to 2006 was carried out. Based on clinical, laboratory, and imaging findings, animals were grouped as having PSS, parenchymal hepatic disease, or extrahepatic disease. The sensitivity and specificity of ammonia and SBA concentrations for the diagnosis of PSS were calculated and receiver-operating characteristic analysis was used to optimize cut-offs. Results Using the upper limit of laboratory reference intervals (ammonia, 59 mu mol/L; SBA, 20 mu mol/L), the sensitivity and specificity of ammonia was 85% and 86% in dogs, and 83% and 76% in cats, respectively. The sensitivity and specificity of SBA was 93% and 67% in dogs, and 100% and 71% in cats, respectively. Using optimal cut-off points for ammonia (dogs, 57 mu mol/L; cats, 94 mu mol/L) the sensitivity and specificity was 91% and 84% in dogs and 83% and 86% in cats, respectively. Using optimal cut-off points for SBA (dogs, 58 mu mol/L; cats, 34 mu mol/L) the sensitivity and specificity was 78% and 87% in dogs and 100% and 84% in cats. Conclusion Increased fasting ammonia and SBA concentrations are accurate indicators of PSS. An improvement in diagnostic accuracy can be achieved by using defined optimal cut-off points for the selective diagnosis of PSS.