Interventions to improve breastfeeding behaviors : detailed summaries of 51 studies

Of the 51 projects reviewed, 24 were in Latin America and the Caribbean, 9 in Asia, 4 in sub-Saharan Africa, 3 in the Near East, 1 in the Newly Independent States, and 10 in Europe and the United States. Three categories of study designs are identified: design A -- studies comparing intervention and control or comparison groups before and after the intervention (29% of the studies); design B -- studies comparing intervention and control or comparison groups (no baseline) (38% of the studies); and design C -- studies comparing the same group(s) before and after the intervention (no control group) (33%). Within these three categories, studies are arranged alphabetically by country, with developing countries followed by developed countries. Of the three research designs, type A -- the pre- post design with a control group -- is the most rigorous. Study types B and C, which lack, respectively, baseline data and a control group, are seriously flawed. Only seven studies -- all type A -- were true experiments, with randomization of treatment among subjects. Tables at the beginning of each section indicate which of the following nine interventions were used to influence breastfeeding behaviors: prenatal education; hospital policies/actions; health worker training; peer counseling; mass media; women"s groups; home visits; community education; postpartum clinics. Studies that attempted to affect more than behavior are identified. Study findings are discussed in a separate publication (PN-ACH-559).