Early breastfeeding cessation: validation of a prognostic breastfeeding score

Aim: To validate a simple breastfeeding score to identify mothers who stop breastfeeding within 4 months after birth.
Methods: Two independent cohorts of Danish mothers in 1999 and 2004 with 4 months of follow-up on breastfeeding duration were used. The breastfeeding score was developed from 471 mothers' responses to a questionnaire in 1999 and based on duration of schooling, previous breastfeeding experience, self-efficacy, and mother's confidence in ability to produce milk. The 2004 cohort consisting of 723 mothers was used to validate the score.
Results: A breastfeeding score of 7 or higher classified 45% of the mothers in the 2004 cohort as being at risk of breastfeeding cessation. With this cut-point the sensitivity was 70% and the specificity 71%. Among primipara the cut-point gave a sensitivity of 76% and a specificity of 54% and classified 60% to be in the risk group. Among multipara the corresponding figures were 66%, 81% and 34%, respectively. The area under the ROC curve was 0.78.
Conclusion: The breastfeeding score based on a simple scoring system derived from four risk factors was capable of predicting the breastfeeding duration in an independent sample. It may help health professionals to identify mothers at risk of breastfeeding cessation before 4 months.