Study objective: To determine the diagnostic usefulness of percutaneous liver biopsy in evaluating patients with chronically elevated liver-associated enzymes.
Design: Comparison of diagnosis made before biopsy by one physician on the basis of a noninvasive work-up (history, physical examination, laboratory values, and imaging studies) and final diagnosis made after biopsy by a second physician formulated after review of all available noninvasive information and study of the biopsy specimen.
Setting: Referral-based gastroenterology clinic at a U.S. Navy medical center.
Patients: Sequential sample of 107 patients with elevated liver-associated enzymes for a minimum of 3 months. Ninety patients were eligible for study.
Interventions: The final diagnosis made by the second physician blinded to the first clinician's diagnosis served as the criterion standard.
Measurements and main results: Four diagnostic groups were selected for analysis: Alcoholic liver disease, fatty liver, chronic necroinflammatory diseases, and miscellaneous. The positive predictive value of the prebiopsy diagnosis ranged from 88% (CI, 75% to 100%) for alcoholic liver disease to 56% (CI, 37% to 75%) for fatty liver. Higher elevations of transaminase values (greater than three times the upper limit of normal) correlated positively with increased prebiopsy diagnostic accuracy. Fatty liver was present in 19% of the cohort. Liver diseases requiring specific therapy other than alcohol abstinence were overlooked and diagnosed only after review of the biopsy in five cases. Conversely, four cases of liver disease, thought to require specific therapy on the basis of noninvasive work-up, were ruled out by biopsy.
Conclusion: The cause of chronic liver disease is best elucidated when the noninvasive work-up is complemented by review of a biopsy specimen.