TABLE 4

Observed and BAS-Predicted Rates of Early Breastfeeding Cessation in the Derivation and External Validation Cohorts

BASCessation Rate by 7–10 d in the Derivation Cohort9Cessation Rate by 14 d in the Validation Cohort
No. of PatientsBAS-Predicted, %Observed, %No. of PatientsaBAS-Predicted, %Observed, %
101733.21.7462.40.0
92885.15.6954.13.1
82448.17.4957.32.1
718312.716.47711.26.5
610119.217.85417.311.1
54928.126.53823.913.2
41839.033.31132.69.1
31351.238.5842.40
2563.280.00
1173.8100.00
<100
All patientsb107510.510.542410.85.2
  • a Seventy-five observations with a missing value for 1 or more clinical predictors comprising the BAS were discarded from the main analysis.

  • b Calibration-in-the-large estimate is relevant only for external validation because there is a correspondence between the average predicted probability and observed frequency of the event of early breastfeeding cessation in the derivation cohort.22 Calibration-in-the-large estimate was 5.6% (95% CI: 3.5 to 7.8) in the external validation cohort.