Construction and validation of a nomogram to predict 1-year mortality risk in patients with HIV/AIDS undergoing maintenance hemodialysis.

Abstract

This single-center retrospective study aimed to explore the 1-year mortality risk factors in 166 patients with human immunodeficiency virus (HIV)/acquired immune deficiency syndrome (AIDS) undergoing maintenance hemodialysis (MH) between 6 June 2017 and 6 June 2023, and construct a 1-year mortality prediction model. The patients were classified into survival and mortality groups based on the 1-year follow-up results, and into training and validation sets at a ratio of 1:1 (53 mortalities and 53 survivors in the training set and 48 mortalities and 58 survivors in the validation set). Stepwise logistic regression was used to construct a 1-year mortality prediction model and to visualize it as a nomogram. Receiver operating characteristic (ROC) analysis, calibration curves, and decision curves were used for nomogram evaluation in the training set and validation in the validation set. Age (≥52 years) (OR (95% CI): 2.05 (3.191-18.892), p < .001), neutrophil to albumin ratio (NAR) (≥0.135) (OR (95% CI): 4.753 (2.011-11.234), p < .001), and HIV-RNA (≥24,650) (OR (95% CI): 13.786 (5.493-34.598), p < .001), represents three of five independent risk factors of 1-year mortality in HIV/AIDS undergoing MH. The AUC of the nomogram for the training and validation sets were 0.908 (95% CI: 0.853-0.963) and 0.939 (95% CI: 0.896-0.983), respectively. The 1-year mortality prediction showed good separation capacity, calibration capacity, and clinical net benefit, which may benefit the management of patients with HIV/AIDS undergoing MH.