Clinical Edge

Summaries of Must-Read Clinical Literature, Guidelines, and FDA Actions

Machine-learning model predicts NASH based on common clinical and lab values

Key clinical point: A machine-learning model based on standard clinical and laboratory values could serve as a screening tool to reveal undiagnosed cases of nonalcoholic steatohepatitis (NASH).

Major finding: One version of the tool, including 14 variables, had good performance, with an area under the curve of 0.76.

Study details: Exploratory analysis including 704 patients with NASH or non-NASH nonalcoholic fatty liver disease (NAFLD), followed by validation in 1,106 NASH patients in an electronic health records database.

Disclosures: Dr. Schattenberg provided disclosures related to AbbVie, Novartis, MSD, Pfizer, Boehringer Ingelheim, BMS, Intercept Pharmaceuticals, Genfit, Gilead, and Echosens. Several study coauthors reported employment with Novartis and stock ownership.

Citation:

Huang J et al. The Liver Meeting 2019, Abstract 0190.