Derek Gazis

Derek Gazis

Auto-generated profile, based on publicly available information

Derek Gazis

Target RWE, Durham, North Carolina, USA

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
11 June 2025 in American Journal of Gastroenterology
Verified
Journal Article
Closed access
ABA. Sidney Barritt
A. Sidney Barritt
ETElliot B. Tapper
Elliot B. Tapper
PNPhilip Noel Newsome
Philip Noel Newsome
DGDerek Gazis
Derek Gazis
HMHeather Morris
Heather Morris
ASArun Sanyal
Arun Sanyal
INTRODUCTION: Risk of cardiovascular (CV) events is estimated by the Framingham Risk Score (FRS), Pooled Cohort Equation (PCE), and the American Heart Association Predicting Risk of cardiovascular disease (CVD) EVENTs (PREVENT) equation. The applicability of these risk tools in patients with metabolic dysfunction-associated steatotic liver disease (MASLD) is uncertain. This study sought to determine the accuracy of the FRS, PCE, and PREVENT in a real-world cohort of patients with MASLD. METHODS: This analysis included US adults 30 years and older with MASLD in the TARGET-NASH study. Five to ten-year CV risk was estimated using original and recalibrated versions of the FRS and PCE, and original PREVENT equation. Discrimination among prediction models was assessed using the Harrell concordance statistic, and calibration was evaluated using a modified Hosmer Lemeshow test comparing observed and predicted CV events. Logistic regression was used to assess the contribution of liver disease to observed CV events. RESULTS: Overall, 1,090 patients were included. There was an increase in observed CV events from metabolic dysfunction-associated steatotic livers to cirrhosis. FRS demonstrated a weak ability to identify patients who went on to experience a CV event (C-statistic 0.58, 95% confidence interval 0.52–0.63) as did the American Heart Association PREVENT model (C-statistic 0.60, 95% confidence interval 0.54–0.65). Both the FRS and PCE demonstrated a significant lack of calibration ( P < 0.01), with overestimation in the highest deciles and underestimation in the lowest deciles of predicted risk. DISCUSSION: Commonly used tools to identify CV risk performed poorly in a cohort of patients with MASLD. Because CV related death is the greatest source of mortality among patients with MASLD, better risk assessment tools are required.