Brief communication: Clinical implications of short-term variability in liver function test results

Background: Clinicians sometimes order liver tests as part of a screening examination or general work-up. Current guidelines do not recommend routine retesting of abnormal results in asymptomatic patients. Objective: To characterize the magnitude of intraindividual variability of liver test results and determine the proportion of adults with persistently elevated levels after 1 positive test. Design: Reliability study. Setting: The NHANES (National Health and Nutrition Examination Survey) III First and Second Examinations (1988 to 1994). Participants: 1864 men and women age 18 years or older living in the United States. Measurements: Repeated measurements (mean, 17.5 days apart) of aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase, γ-glutamyltransferase, and bilirubin. Results: Using NHANES III cutoffs for normal levels, 38% of adults with initially elevated bilirubin levels had normal levels at the second examination. These proportions were 36%, 31%, 17%, and 12% for AST, ALT, alkaline phosphatase, and γ- glutamyltranferase, respectively. More than 95% of initially normal results remained normal. The results did not differ by alcohol consumption; hepatitis A, B, or C serologic status; recent infection; body mass index; or sociodemographic characteristics. Intraindividual variability was significantly higher for bilirubin (coefficient of variation, 23.4%) and ALT (coefficient of variation, 20.4%) than for AST (coefficient of variation, 13.9%), γ-glutamyltransferase (coefficient of variation, 13.8%), and alkaline phosphatase (coefficient of variation, 6.7%). Limitations: Only 2 measurements were available. Complete liver disease history was lacking. Conclusion: If retested, more than 30% of adults with elevated AST, ALT, or bilirubin levels would be reclassified as normal. Clinicians should be aware of the high intraindividual variability in common liver tests, and practice guidelines should explicitly recommend retesting of asymptomatic individuals with abnormal liver test results.