Cardiovascular Disease Risk Prediction in Patients With Metabolic Dysfunction‐Associated Steatohepatitis

BACKGROUND AND AIMS: Metabolic dysfunction-associated steatohepatitis (MASH) is associated with an increased risk of cardiovascular disease (CVD) morbidity and mortality. This study aimed to develop the first prediction models for CVD risk in a cohort of patients with MASH. METHODS: This was a retrospective cohort study using data from the UK Clinical Practice Research Datalink (CPRD) database. Accelerated failure time (AFT) models were used to predict CVD risk independently for males and females with MASH. Covariables from the QRisk3 algorithm were included: age, deprivation, body mass index, cholesterol ratio, systolic blood pressure, ethnicity, smoking status, CVD family history, diabetes, treated hypertension, rheumatoid arthritis, atrial fibrillation, chronic kidney disease (CKD), migraine, corticosteroids, anti-psychotic medication, serious mental illness and erectile dysfunction. Measures of calibration and discrimination were determined. Observed and predicted risks were used to compare the AFT models with the QRisk3 algorithm. RESULTS: Utilising a cohort of 10 461 patients with MASH (5364 female and 5097 male) models to predict time to CVD were developed with moderate predictive power (C-statistics 0.7-0.72) identifying age, cholesterol ratio, type 2 diabetes and CKD as risk factors that decrease time to CVD. Comparing the observed and predicted CV risks indicated the AFT models more accurately predicted CVD risk than the QRisk3 algorithm in patients with MASH. CONCLUSIONS: We describe a first-in-kind predictive model to assess the risk of CVD in patients with MASH, which has the potential to more accurately inform the treatment and management of this population.