Aim:

Validation of the FLI (Fatty Liver Index) diagnostic test by means of a cross-sectional design.

Methods:

Patients with a prior diagnosis of obesity and/or diabetes in whom an ultrasound would be indicated to rule out steatosis were included. The FLI and the Gold Standard test (ultrasound) were performed. Sample size: 135 individuals were included. ROC curve, area under the curve and the FLI threshold for classification as steatosis were all calculated. Sensitivity, specificity and the positive and negative predictive values for FLI were estimated. The SPSS programme was used for the analysis. All patients were given a study information sheet and informed consent was requested.

Results:

Prevalence of steatosis of 60.7%, with mild and moderate steatosis predominating. There was a statistically significant relationship between steatosis and triglycerides, but not for BMI (body mass index), GGT (gamma-glutamyl transferase) and abdominal perimeter. The FLI ROC curve was very close to the midline, and the area under the curve was 0.666. This reveals a low predictive capacity for FLI. Considering a threshold of 76 for the FLI, the sensitivity, specificity, positive and negative predictive values (PPV and NPV) were 75.6%, 50.94%, 70.45% and 57.45%, respectively. The Positive and Negative Likelihood Ratios were 1.53 and 0.49, respectively. This reveals that FLI cannot be deemed a good test to diagnose steatosis.

Conclusions:

The Fatty Liver Index test does not adequately predict patients with diabetes and/or obesity who would have associated steatosis. Therefore, the use of FLI to diagnose steatosis or to replace ultrasound cannot be recommended in general.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )