Skip to main content

Table 3 Performance measures of biomarkers in identifying preeclampsia

From: The prognostic utility of soluble fms-like tyrosine kinase-1 (sFlt-1) and placental growth factor (PIGF) biomarkers for predicting preeclampsia: a secondary analysis of data from the INSPIRE trial

Model

Regression model output

Model performance corrected for optimism

Coefficient

Intercept

Calibration slope

CITL

C-index

R2

BIC

AUC

Model 1: sFlt-1 onlya

7.37

-29.92

0.98

0

0.94

0.55

144

0.94

Model 2: PIGF onlya

-4.77

8.08

0.99

0

0.87

0.38

184

0.89

Model 3: sFlt-1/PIGF ratioa

-9.47

4.19

0.99

0

0.94

0.57

139

0.94

Model 4: sFlt-1/PIGF cut-off at 38b

4.37

-5.81

0.84

0

0.88

0.46

165

0.89

  1. sFlt-1 Soluble fms-like tyrosine kinase – 1, PIGF Placental growth factor, CITL Calibration-in-the-large, BIC Bayesian information criteria (the lowest BIC values indicating better model performance), C-index Concordance index, AUC Area under the curve
  2. aPredictor variables are continuous values included in the model equation of biomarker/s as every unit increases in log-scale. All model results are optimism corrected and adjusted for the trial arm
  3. bNo log transformation applied for the cut off predictor variable