Skip to main content

Table 8 Comparison of the diagnostic efficiency of the two models for predicting IgAN and non-IgAN

From: A novel differential diagnostic model based on multiple biological parameters for immunoglobulin A nephropathy

Model Cut-off point of of predicted probability Predicted membership Biopsy diagnosis Sensitivity Specificity Accuracy
   IgAN Non-IgAN    
Logistic regression model PRE-1 0.50 IgAN 64 12 68.8% 86.2% 77.5%
  Non-IgAN 29 75    
  0.40 IgAN 77 18 82.8% 79.3% 81.1%
   Non-IgAN 16 69    
Discriminant analysis model PRE-2 0.50 IgAN 64 13 68.8% 85.1% 77.0%
   Non-IgAN 29 74    
  0.40 IgAN 77 20 82.8% 77.0% 79.9%
   Non-IgAN 16 67    
  1. Abbreviations: IgAN, immunoglobulin A nephropathy; Non-IgAN, non-immunoglobulin A nephropathy; PRE, predicted probability.