OBJECTIVES: Current practice for diagnosing an excessively drinking patient suspected of having advanced fibrosis in the Danish primary care sector is not well understood as there are currently no guidelines. Most GPs however, likely order a battery of liver function tests applied in parallel and refer a patient with any positive result to ultrasonography in secondary care. As recently demonstrated by a Danish clinical trial, advanced fibrosis can be effectively ruled out in primary care patients by sequentially applying non-invasive markers, with follow-up testing by liver-stiffness measurement in secondary care. Our objective was to find out which of our included alternatives could maximise the number of advanced fibrosis diagnoses without resulting in excessive unnecessary referrals to secondary care.

METHODS: We used decision trees to compare the number of correct diagnoses and unnecessary referrals to secondary care examination with four competing strategies: 1) seven liver function tests applied in parallel with referral to ultrasonography for positives, 2) serial testing with the Enhanced Liver Fibrosis (ELF) test and liver stiffness measurement (LSM) for ELF positives, 3) serial testing with the Forns index followed by ELF for index positives and referral to LSM for ELF positives, or 4) liver biopsy.

RESULTS: At a prevalence of advanced fibrosis of 6% among Danish excessive drinkers the correct-diagnoses-to-unnecessary-referrals ratio per 1000 tested was 0.24 for strategy 1, 10.1 for strategy 2, 65.8 for strategy 3, and 0.063 for strategy 4.

CONCLUSIONS: With an objective of maximising the correct number of advanced fibrosis diagnoses, and minimising unnecessary referrals to the secondary care sector, triaging patients first with the non-commercial Forns index, followed by ELF and LSM in serial for positive patients was optimal.