Abstract: In the context of escalating prevalence of chronic kidney disease (CKD) globally, the mortality rate among patients with end-stage renal disease (ESRD) has remained a concern point in the field of nephrology. Maintenance hemodialysis (MHD) is the predominant renal replacement therapy for ESRD. However, mortality risk in MHD patients is still higher with significant inter-patient variability. Accurate prediction of mortality risk in MHD patients is critical for clinical decision-making, resource allocation, and patient management. This review aims to discuss and outline the mortality risk prediction models developed for hemodialysis patients, both domestically and internationally. By comparing the modeling methods, predictive capabilities, generalizability and clinical utility of the models, this review provides a reference for constructing an appropriate predictive model.

Key words: Hemodialysis, Prediction model, Machine learning