ASN's Mission

ASN leads the fight to prevent, treat, and cure kidney diseases throughout the world by educating health professionals and scientists, advancing research and innovation, communicating new knowledge, and advocating for the highest quality care for patients.

learn more

Contact ASN

1401 H St, NW, Ste 900, Washington, DC 20005

email@asn-online.org

202-640-4660

The Latest on Twitter

Kidney Week

Abstract: TH-PO1022

Decision Tree Prediction Models for the Development of ESRD in Immunoglobulin A Nephropathy

Session Information

Category: Glomerular Diseases

  • 1203 Glomerular Diseases: Clinical, Outcomes, and Trials

Authors

  • Han, Xin, West China Hospital of Sichuan University, Chengdu, Sichuan, China
  • Tang, Yi, West China Hospital of Sichuan University, Chengdu, Sichuan, China
  • Qin, Wei, West China Hospital of Sichuan University, Chengdu, Sichuan, China
Background

IgA nephropathy (IgAN) is the most common glomerulonephritis worldwide and up to 40% will develop end-stage renal disease (ESRD) within 20 years.

Methods

The aim of this study was to develop a predictive model to predict the risk of progressing to ESRD by using decision tree algorithm. Data were evaluated from biopsy-proven IgAN patients in West China hospital, Sichuan University in China between 2009 to 2017. 2 final models were selected by Gini index and the area under receiver-operating characteristic (ROC) curve.

Results


Clinical model was developed by only clinical factors, as pathological model was conducted by both clinical and pathological factors. In clinical model (Figure 1), recursive partitioning indicated that the best single predictor of renal deterioration was proteinuria, followed by severe urine acid for patients with severe proteinuria (> 0.98 g/d). Systolic blood pressure (SBP) and estimated glomerular filtration rate (eGFR) were placed in the third tier of the decision tree model. T and S score were put at high levels in pathological model (Figure 2) and followed with eGFR at third level. Proteinuria and SBP were placed at forth level. Nephrotic syndrome was presented at last. The accuracy of clinical and pathological model were 0.85 and 0.86, respectively. The ROC of clinical model was 0.83, compared with pathological model with 0.85.

Conclusion

Decision model can practically predict the risk of incidence of ESRD in IgAN patients. Model with pathological factors can be more accurate.

Figure 1. Clinical model of incidence of ESRD in IgAN (Line to left: yes; right: no)

Figure 2. Pathological model of incidence of ESRD in IgAN (Line to left: yes; right: no)