Evaluating hospital breastfeeding programs: Comparing policies, practices and mother's experiences

Effective outcome evaluation requires that program components are clearly identified. Two hospital-based breastfeeding (BF) programs were assessed for program implementation (based on Baby Friendly's ten steps) using interviews and surveys from administrators, practicing nurses and post-partum mothers.

Two geographically proximal hospitals serving similar populations, both with well-defined BF programs, (one designated Baby Friendly (BFH); one community hospital (CH)) were scored by five nursing/medical administrators (interviews using previously published instrument). Nurses scored nine components (BFH= 61; CH=43)anonymous survey; adapted from administrator interview) and postpartum mothers scored seven (BFH=428; CH=414) (PRAMS survey questions). For comparisons, scores were standardized; ratings above 90% were considered fully implemented (FuImpl); partial implementation >75% (PtImpl). Between group differences of <20% indicated rating concurrence. Six components were rated by all three groups (help initiating, showing how, only breastmilk, rooming in, no pacifier and fostering support).

BFH administrators' rated five of these six as FuImpl and one as PtImpl. Nurses concurred with all administrator ratings. Among the five FuImpl components, mothers' ratings concurred with three (help initiating, showing how and no pacifier).

CH administrators' scores indicated four FuImpl and one PtImpl component. Nurses' ratings differed, identifying three FuImpl and one PtImpl. They concurred on FuImpl of two components. Of these, mothers' ratings concurred with one (no pacifier).

While these hospitals had some similarities based on program descriptions, they differed substantively on extent of implementation. This finding reinforces the importance of documenting implementation in order to correctly interpret and attribute outcome findings to a particular programmatic intervention.