Household trials with very small samples predict responses to nutrition counseling intervention.

Household trials were conducted to test the acceptability and feasibility of the recommendations to be delivered to the mothers in the context of a randomized intervention implemented in Pelotas, Brazil. A first home visit was paid to assess child health and feeding problems. In a second visit, the mother was encouraged to select one or two recommendations to try out over five days. The last visit was used to assess the mothers' experiences in attempting to implement the recommendations. Nonexclusive breastfeeding, use of the bottle, monotonous diet, and low energy density of foods were the most common problems. The most frequently selected recommendations were those aiming to increase the energy density of foods. Mothers generally reported positive responses to the recommendations. The household trials highlighted the acceptability and feasibility of the planned recommendations and correctly predicted the changes that were successfully implemented by the mothers in the large intervention study.