In people suffering from NAFLD, the steatosis-associated fibrosis estimator score helps in the estimation of liver fibrosis and may also be beneficial to detect low-risk NAFLD in primary care.

The study aimed to evaluate liver fibrosis with the aid of steatosis-associated fibrosis estimator (SAFE) score in patients with non-alcoholic fatty liver disease (NAFLD).

Extraction of biopsy-proven NAFLD patient's data from the 'Non-alcoholic steatohepatitis Clinical Research Network' observational study (n = 676) was done. To differentiate ≥F2 from F0/1, prediction models were constructed utilizing logistic regression and machine-learning methods. Using magnetic resonance elastography data, testing of models was done in 280 subjects in a trial ('FLINT') and 130 local NAFLD people.

The final model was implemented on 11,953 candidates in the National Health and Nutrition Examination Survey III (NHANES) to associate with long-term mortality. Selection of a multivariable logistic regression model as SAFE score was done. It encompassed globulins (total serum protein minus albumin), alanine aminotransferases, aspartate, platelets, diabetes, body mass index, and age.

The model ceded areas under receiver operating characteristic curves ≥0.80 in differentiating F0/1 from ≥F2 in testing datasets, uniformly greater than those of NAFLD Fibrosis and fibrosis-4 (FIB-4) scores. As found, the negative predictive values to rule out ≥F2 at SAFE of 0 came out to be 92% and 88% in the two testing sets.

In NHANES III set, survival up to 25 years of people with SAFE <0 was similar to people without steatosis. For people with SAFE>100, elevating SAFE scores associated with reduced survival with an adjusted hazard ratio of 1.54.

The SAFE score uses extensively available variables that help in the estimation of liver fibrosis in NAFLD patients. This score may be utilized in primary care to identify low-risk NAFLD.