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

Table 9 Diagnostic efficiency of the two models when the predicted probabilities is either >0.59 or <0.26 (the cut-off point = 0.4)

Predicted membership of the models	IgAN	Non-IgAN	Sensitivity	Specificity
Logistic regression model PRE-1
IgAN	54	7	85.7%	88.5%
Non-IgAN	9	54	85.7%	88.5%
Discriminant analysis model PRE-2
IgAN	51	6	85.0%	90.0%
Non-IgAN	9	54

Abbreviations: IgAN, immunoglobulin A nephropathy; non-IgAN, non-immunoglobulin A nephropathy; PRE, Predicted probability.

ISSN: 1472-6947