OBJECTIVES: Liver cirrhosis is associated with osteoporosis leading to an increased risk of fractures. We aimed to establish whether a risk stratification strategy using a fracture risk calculation tool (FRAX) to determine which patients should receive a dual-energy X-ray absorptiometry (DXA) scan is effective in reducing scan rates without compromising sensitivity for detecting osteoporosis. METHODS: A retrospective analysis of 252 patients with liver cirrhosis attending hepatoma surveillance clinics. Receiver operating characteristic analysis was performed to assess sensitivity and specificity at 10-year fracture risk thresholds of 5, 10 and 15%. RESULTS: DXA scans were performed among 252 patients. Mean age was 61.6±10.2 years, of which 53.2% were male. Cirrhosis aetiology was largely a result of alcohol excess (n=33.3%), chronic hepatitis C virus infection (n=20.2%) and nonalcoholic fatty liver disease (n=15.9%). The majority of patients were in good prognostic groups (87.4% Child-Pugh A, 11.3% Child-Pugh B, 1.3% Child-Pugh C). Osteoporosis was present in 19.0% of those who underwent DXA scanning. The optimum 10-year fracture risk threshold was found to be 10% using the FRAX tool. This retained a high sensitivity of 95.8%, specificity 64.7%, and negative predictive value 98.5%. Introduction of a 10% FRAX threshold would result in a reduction of the DXA scanning rate to 46.8% of the current rate. CONCLUSION: A risk stratification strategy for DXA scanning using a fracture risk assessment tool (FRAX) and a 10-year fracture risk threshold of 10% leads to a significant reduction in scan rates without compromising osteoporosis detection rates.
OBJECTIVES:Liver cirrhosis is associated with osteoporosis leading to an increased risk of fractures. We aimed to establish whether a risk stratification strategy using a fracture risk calculation tool (FRAX) to determine which patients should receive a dual-energy X-ray absorptiometry (DXA) scan is effective in reducing scan rates without compromising sensitivity for detecting osteoporosis. METHODS: A retrospective analysis of 252 patients with liver cirrhosis attending hepatoma surveillance clinics. Receiver operating characteristic analysis was performed to assess sensitivity and specificity at 10-year fracture risk thresholds of 5, 10 and 15%. RESULTS: DXA scans were performed among 252 patients. Mean age was 61.6±10.2 years, of which 53.2% were male. Cirrhosis aetiology was largely a result of alcohol excess (n=33.3%), chronic hepatitis C virus infection (n=20.2%) and nonalcoholic fatty liver disease (n=15.9%). The majority of patients were in good prognostic groups (87.4% Child-Pugh A, 11.3% Child-Pugh B, 1.3% Child-Pugh C). Osteoporosis was present in 19.0% of those who underwent DXA scanning. The optimum 10-year fracture risk threshold was found to be 10% using the FRAX tool. This retained a high sensitivity of 95.8%, specificity 64.7%, and negative predictive value 98.5%. Introduction of a 10% FRAX threshold would result in a reduction of the DXA scanning rate to 46.8% of the current rate. CONCLUSION: A risk stratification strategy for DXA scanning using a fracture risk assessment tool (FRAX) and a 10-year fracture risk threshold of 10% leads to a significant reduction in scan rates without compromising osteoporosis detection rates.