Building a baby-friendly community: Development, strategy, implementation and evaluation of a BFCI program.

Participants were recruited by convenience sampling, so they may have been more motivated and more likely to breastfeed than the general population. Many participants were university educated, primiparous, or had higher household incomes, which limits generalisability to lower socio-economic groups. Attrition was substantial: 1348 staff consented at baseline, 1339 completed post-workshop, and only 290 responded at staff follow-up; 1540 women consented at baseline, 1224 completed post-workshop self-efficacy, and 696 completed 1-month self-efficacy follow-up. Follow-up stopped at 1 month, so the study cannot show whether breastfeeding continued after mothers returned to work or after the early postpartum period. The study was quasi-experimental, so changes may reflect self-selection, normal postpartum learning, or other supports. The mall staff training arm was not followed long enough to link staff knowledge gains to mothers' breastfeeding outcomes.

Funded by the Health and Medical Research Fund (HKU-BF). The authors declare no known competing financial interests or personal relationships that could have influenced the work.

Do not read the follow-up breastfeeding rates as proof the BFCI program works. The study used convenience samples, had heavy attrition, and only followed women for one month. It also could not link mall staff training to mothers' outcomes.