A study in JAMA Internal Medicine evaluated nine risk scores for predicting cirrhosis and hepatocellular carcinoma (HCC) in 853,131 veterans with imaging-confirmed noncirrhotic steatotic liver disease. The SAFE (steatosis-associated fibrosis estimator) score provided the largest net benefit for detecting 10-year cirrhosis risk, with a score of 29.5 corresponding to a 2.5% risk and a net benefit of 1.9 additional cirrhosis cases per 100 at-risk patients. The tool includes seven readily available measures: age, BMI, diabetes status, AST, ALT, globulins, and platelets.
All prediction tools showed minimal utility for HCC screening. A SAFE score of 43.8 corresponded to a 10-year HCC risk of 0.25%, yielding a net benefit of 1.6 true positives per 1,000 screened — meaning 625 patients would need routine HCC screening to identify one additional case. The authors concluded that existing risk scores do not adequately predict HCC to inform screening in noncirrhotic SLD.
The researchers suggested SAFE could help counsel patients and decide when to repeat cirrhosis assessments. An accompanying editorial noted that primary care screening pathways for steatotic liver disease are still in their early stages compared with mature frameworks for hypertension or chronic kidney disease. The study's limitations include missing data for six risk scores and a predominantly male veteran population, which may limit generalizability. The current standard, FIB-4, was among the tested tools and was described as not very sensitive or specific.