This study was designed to test whether ultrasound can be used to diagnose nonalcoholic fatty liver disease (NAFLD) utilizing a prototype. We collected 115 ultrasounds. A prototype was chosen that represented NAFLD; 5 features of NAFLD prototype were described. Ultrasounds were read blinded to diagnosis as matching prototype or not. A 20-minute teaching session was made to a group of 15 providers. Ten ultrasounds were presented for comparison to prototype with intraobserver reliability measured. Of 20 patients shown by liver biopsy to have NAFLD, 16 were successfully predicted by comparison to the prototype (sensitivity 80%). In 94 of 95 cases, ultrasound predicted those without NAFLD (specificity 99%). The positive predictive value was 94% and negative predictive value 96%. Training results showed substantial agreement with a kappa score of 0.76 with 95% of cases identified correctly. In conclusion, physicians can apply a bedside ultrasound to identify NAFLD when compared to prototype.
All Science Journal Classification (ASJC) codes