[The influence of information from the Brazilian National Health Survey on the current estimate and trend in exclusive breastfeeding in Brazil].

The study aimed to discuss the consistency of estimates of maternal breastfeeding in the Brazilian National Health Survey (PNS) and the implications for the analysis of trends in these indicators from 1986 to 2013. The main limitation for the current estimate (2013) of breastfeeding indicators based on the PNS and their trend (1986-2013) is the absence of the interview date in the microdata, thus hindering a precise calculation of age. Previous studies have shown different scenarios for exclusive breastfeeding based on the PNS: 36.6% (stabilization of prevalence in comparison to 2006 [Brazilian National Survey of Demography and Health - PNDS]) and 20.3% (downward prevalence). Simulations of the interview date using the date of data collection in the PNS (August 2013 to February 2014) indicated a variation of prevalence 25.3% to 42.8% (exclusive breastfeeding); 54% to 61.5% (breastfeeding); and 32.7% to 40.5% (breastfeeding, second), respectively. The discrepancy in the scenario for breastfeeding status calls attention to the unfeasibility of the exclusive breastfeeding estimate for 2013, pending clarification of the interview date.