Development and validation of a risk prediction model for end-stage renal disease in patients with type 2 diabetes

The aim of this study is to develop a prediction model for ESRD in patients with type 2 diabetes. A retrospective cohort study was conducted, consisting of 24,104 Chinese patients with type 2 diabetes. We adopted the procedures proposed by the Framingham Heart Study to develop a prediction model for ESRD. Participants were randomly assigned to the derivation and validation sets at a 2:1 ratio. The Cox proportional hazard regression model was used for model development. A total of 813 and 402 subjects (5.06% and 5.00%, respectively) developed ESRD in the derivation and validation sets over a mean follow-up period of 8.3 years. The risk-scoring systems included age, gender, age of diabetes onset, combined statuses of blood pressure and anti-hypertensive medication use, creatinine, variation in HbA1c, variation in systolic blood pressure, diabetes retinopathy, albuminuria, anti-diabetes medications, and combined statuses of hyperlipidemia and anti-hyperlipidemia medication use. The area under curves of 3-year, 5-year, and 8-year ESRD risks were 0.90, 0.86, and 0.81 in the derivation set, respectively. This risk score model can be used as screening for early prevention. The risk prediction for 3-year, 5-year, and 8-year period demonstrated good predictive accuracy and discriminatory ability.