What makes mothers change their minds? Unexpected results around accuracy of breastfeeding data on the Massachusetts birth certificate

Background: Many states record infant feeding data on the birth certificate, from which breastfeeding initiation rates can be calculated. However, accuracy of birth certificate data is unclear. In Massachusetts, the birth certificate reflects the mother's answer to the question, “Are you breastfeeding or do you intend to breastfeed?”, asked prenatally on the obstetric unit. Objective: To determine whether birth certificate data accurately reflected actual feeding method. Methods: Between 2004-5, we randomly selected 306 medical records at Lowell General Hospital (LGH) and 232 medical records at Baby-Friendly Boston Medical Center (BMC). Individual birth certificate answers were compared with each infant's postpartum feeding record. Results: At LGH, 94.8% of birth certificate responses concurred with the feeding record. (7 infants described as breastfed by the birth certificate received formula; 9 described as formula fed were breastfed). At BMC, 79.8% of birth certificate responses concurred with the feeding record. (7 infants described as breastfed on the birth certificate were fed formula; 40 infants described as formula fed were breastfed). Chi square analysis found no significant difference between women who answered correctly or incorrectly. Conclusions: Although breastfeeding is a desirable health behavior, mothers did not overstate intent prenatally. Accuracy of breastfeeding data on the birth certificate varied dependent on hospital. The common assertion that women make infant feeding decisions in pregnancy may be flawed. In the Baby-Friendly hospital, many women who stated intent to formula feed apparently decided to breastfeed in the hospital postpartum. Baby-Friendly policies may affect a woman's feeding decision postpartum.