Construction of Logistic prediction model and countermeasures for type 2 diabetic nephropathy based on clinical data / 中国医师进修杂志

ABSTRACT

Objective:To explore the construction of a Logistic prediction model and countermeasures for type 2 diabetic nephropathy based on clinical data.Methods:The patients with type 2 diabetic nephropathy admitted to Shijiazhuang Second Hospital from September 2019 to September 2021 (study group) were selected and the patients were selected according to a 1∶1 ratio using individual matching (control group), each group with 200 patients. Single and multiple factors analysis were used to analyze the factors influencing type 2 diabetic nephropathy, and Logistic regression equation models were developed to verify their predictive value.Results:Logistic regression equation model showed that the course of type 2 diabetes, glycosylated hemoglobin (HbA 1c), fasting plasma glucose (FPG), homocysteine (Hcy), urinary microalbumin, and serum creatinine (Scr) were high risk factors for type 2 diabetic nephropathy ( P<0.05). The results of Logistic regression model evaluation showed that the model was established with statistical significance, and the coefficients of the regression equations had statistically significant differences. The Hosmer-Lemeshow goodness-of-fit test showed that the model fitting effect was good. Logistic regression model was used to statistically analyzed the data set, and the receiver operating characteristic (ROC) curve of type 2 diabetic nephropathy was drawn, the area under the curve was 0.949(95% CI 0.922 - 0.968), the prediction sensitivity was 81.50%, the specificity was 95.50%, the calibration curve showed that the predicted results was in good agreement with the observed results. Conclusions:The independent predictors of type 2 diabetic nephropathy involve HbA 1c, FPG, Hcy, urinary microalbumin. The Logistic prediction model based on these predictors has reliable predictive value and can help guide clinical diagnosis and treatment.