Home    中文  
 
  • Search
  • lucene Search
  • Citation
  • Fig/Tab
  • Adv Search
Just Accepted  |  Current Issue  |  Archive  |  Featured Articles  |  Most Read  |  Most Download  |  Most Cited

Chinese Journal of Kidney Disease Investigation(Electronic Edition) ›› 2021, Vol. 10 ›› Issue (05): 246-251. doi: 10.3877/cma.j.issn.2095-3216.2021.05.002

• Original Article • Previous Articles     Next Articles

Construction of a model for the onset-risk of diabetic kidney disease based on the classification tree method

Hongjuan Ji1, Juan Lin2,()   

  1. 1. Department of General Practice; Nanjing Hospital of Traditional Chinese Medicine, Nanjing 210006, Jiangsu Province, China
    2. Department of Pharmacy; Nanjing Hospital of Traditional Chinese Medicine, Nanjing 210006, Jiangsu Province, China
  • Received:2021-03-11 Online:2021-10-20 Published:2021-11-23
  • Contact: Juan Lin

Abstract:

Objective

To apply the classification tree method for constructing a model for the onset-risk of diabetic kidney disease (DKD), and to evaluate its value in application.

Methods

285 diabetic patients diagnosed in the Nanjing Hospital of Traditional Chinese Medicine from February 2017 to August 2020 were selected, including 76 patients with DKD (DKD group) and 209 patients with simple diabetes (control group). The correlation between the clinical data and DKD was analyzed for the two groups. Single-factor and multi-factor logisitic regression method was used to analyze the relevant influencing factors of DKD. The chi-squared automatic interaction detector (CHAID) classification tree algorithm was used to establish a predictive risk model for the onset of DKD. The value for application of the model was evaluated with the income graph, index graph, and misclassification probability.

Results

The classification tree model for the onset-risk of DKD included 3 layers with a total of 8 nodes. Four explanatory variables were screened out, including diabetes course, body mass index, systolic blood pressure, and glycosylated hemoglobin (HbA1c), among which HbAlc was the most important predictor. The misclassification probability of the model was 0.149, indicating that the fitting effect of the model was good.

Conclusion

The classification tree model could not only effectively predict the onset-risk of DKD, but also analyze the interaction between the variables. Controlling the weight, blood sugar, and blood pressure of patients with long-term diabetes may help prevent and control the onset-risk of DKD.

Key words: Diabetic kidney disease, Risk factors, Classification tree, Prediction model

京ICP 备07035254号-35
Copyright © Chinese Journal of Kidney Disease Investigation(Electronic Edition), All Rights Reserved.
Tel: 010-66937011 E-mail: zhsbyj@126.com
Powered by Beijing Magtech Co. Ltd