Skip to main content

Table 2 Performance of the PREP-S (risk at various time points) and PREP-L (overall risk by discharge) models in the derivation cohort and the external validation datasets for prediction of complications in women with early onset pre-eclampsia

From: Prediction of complications in early-onset pre-eclampsia (PREP): development and external multinational validation of prognostic models

Model performance

Development cohort

External cohorts

PREP

PIERS

PETRA

PREP-S model

N = 946

N = 339

 

C-statistic (95% CI)

  

–

 At 48 hours

0.84 (0.81–0.87)b

0.75 (0.69–0.81)a

 

 At 1 week

0.79 (0.76–0.81)b

0.72 (0.68–0.76)a

 

 Overall

0.75 (0.73–0.78)b,c

0.71 (0.67–0.75)a

 

Calibration slope (95% CI)

  

–

 At 48 hours

1

0.80 (0.62–0.99)

 

 At 1 week

1

0.75 (0.61–0.89)

 

 Overall

1

0.67 (0.56–0.79)

 

PREP-L model

N = 946

N = 437

N = 211

C-statistic (95% CI)

0.82 (0.80, 0.84)b,d

0.81 (0.77–0.85)a

0.75 (0.64–0.86)a

Calibration slope (95% CI)

1

0.93 (0.72–1.13)

0.90 (0.48–1.32)

  1. CI confidence interval; N number of women analysed; PREP PRediction of complications in Early-onset Pre-eclampsia study; PIERS Pre-eclampsia Integrated Estimate of RiSk study; PETRA Pre-Eclampsia TRial Amsterdam study
  2. aApparent C-statistic
  3. bOptimism adjusted C-statistic
  4. cOverall apparent C-statistic 0.77 (95% CI 0.75, 0.79)
  5. dApparent C-statistic 0.84 (95% CI, 0.82–0.87)