Model Predicts ESLD Among Patients with HIV

A predictive model could be used to help risk of end-stage liver disease (ESLD) in people living with HIV (PWH), according to recent research.

In order to develop a risk prediction model for ESLD in PWH, the researchers used data from 13,787 PWH who received care in 12 US and Canadian cohorts of the North American AIDS Cohort Collaboration on Research and Design from 2000-2016 who had FIB-4>1.45.

Overall, 390 ESLD events occurred over a mean 5.4 years. Of those, 52% had hepatitis C, 15% had hepatitis B, and 31% had a history of alcohol use. The model developed by the researchers included age; sex; race/ethnicity; as well as values reflecting hepatic impairment, including serum albumin, AST, total bilirubin, platelets, and lipid metabolism, including triglycerides, HDL, and total cholesterol.

Their testing subset included 3173 PWH and yielded 112 ESLD events over a mean 4.3 years.

“Our model performed well in this testing set (C-statistic 0.81, 95% CI 0.76, 0.86)”, the researchers wrote.

“This model developed specifically for PWH includes readily accessible clinical parameters, appears to work well in a diverse population and may provide a convenient tool for ESLD prediction,” they concluded.

—Michael Potts

Reference:

Kim HN, Nance RM, Crane H, et al. Prediction model for end-stage liver disease among people with HIV in the NA-accord. Paper presented at Virtual CROI 2021. March 6-10, 2021. https://www.vcroi2021.org/sessions/19764903/subsession/25642261/PREDICTION-MODEL-FOR-END-STAGE-LIVER-DISEASE-AMONG-PEOPLE-WITH-HIV-IN-THE-NA-ACCORD