In Sheth and colleagues' review of NASH [1], the authors advocate extensive evaluation of patients with unexplained elevations in liver enzyme levels. After a history and, presumably, a physical examination, specific recommendations include viral hepatitis serologic testing; iron studies; measurement of ceruloplasmin levels, -antitrypsin levels and phenotype, and anti-mitochondrial and antinuclear antibodies; liver ultrasonography; and Doppler assessment of hepatoportal vasculature. The authors advise that "if none of the above tests are definitive, then a liver biopsy is recommended to confirm or rule out NASH ... ."

I believe that their diagnostic approach is excessive. Do the authors suggest that all elderly patients with elevated liver enzyme levels be screened for Wilson disease? Should men with elevated aminotransferase levels but without cholestasis have antimitochondrial antibody titers checked routinely? In addition, imaging studies of the liver are rarely useful in patients with a hepatocellular injury pattern. Finally, why consider liver biopsy, given its risks, to diagnose NASH when the authors later point out that "no proven therapy for NASH exists"?

Physicians treating patients with abnormal liver blood test results should individualize the evaluation. Moreover, tests should be recommended when the results would have meaningful management implications for patients. As in all clinical decision making, we should avoid using diagnostic "grapeshot" and pursue what we need to know rather than what we want to know.