(Figure 4A)
The calibration of the nomogram was measured by the bootstrap (1000 resample) method. Analysis of the results showed that the predicted probability obtained from the bootstrap correction was not significantly different with the actual probabilities of preterm birth (p = 0.015 and p=0.013, respectively), implying that the nomograms predicting PTB < 37 weeks and PTB < 34 weeks were well-calibrated.