Development of prediction models for cardiovascular disease mortality risk in maintenance hemodialysis patients based on nomogram and CART algorithm

Key Points

• This research aims to develop prediction models to assess cardiovascular disease mortality risk in maintenance hemodialysis patients.
• Identified key risk factors including serum albumin, procalcitonin, and myoglobin.
• Developed models using nomogram and CART algorithm for prediction.
• Evaluated the models' performance in a cohort of maintenance hemodialysis patients.
• Both prediction models showed good performance in estimating CVD mortality risk.
• Serum albumin, procalcitonin, and myoglobin were consistently associated with mortality risk.
• Findings indicate the potential for these models to aid clinical risk assessment, pending further validation.