Abstract
OBJECTIVE: Evaluation of predictive value of liver function tests (LFTs) for the detection of liver-related disease in primary care. DESIGN: A prospective observational study. SETTING: 11 UK primary care practices. PARTICIPANTS: Patients (n=1290) with an abnormal eight-panel LFT (but no previously diagnosed liver disease). MAIN OUTCOME MEASURES: Patients were investigated by recording clinical features, and repeating LFTs, specific tests for individual liver diseases, and abdominal ultrasound scan. Patients were characterised as having: hepatocellular disease; biliary disease; tumours of the hepato-biliary system and none of the above. The relationship between LFT results and disease categories was evaluated by stepwise regression and logistic discrimination, with adjustment for demographic and clinical factors. True and False Positives generated by all possible LFT combinations were compared with a view towards optimising the choice of analytes in the routine LFT panel. RESULTS: Regression methods showed that alanine aminotransferase (ALT) was associated with hepatocellular disease (32 patients), while alkaline phosphatase (ALP) was associated with biliary disease (12 patients) and tumours of the hepatobiliary system (9 patients). A restricted panel of ALT and ALP was an efficient choice of analytes, comparing favourably with the complete panel of eight analytes, provided that 48 False Positives can be tolerated to obtain one additional True Positive. Repeating a complete panel in response to an abnormal reading is not the optimal strategy. CONCLUSIONS: The LFT panel can be restricted to ALT and ALP when the purpose of testing is to exclude liver disease in primary care.