OBJECTIVE:To assess the consistency of the results obtained in the study Assessment of the impact of prenatal nutritional assistance on pregnancy outcome by testing the reliability of data and analyzing follow-up losses. METHODS: This is a prospective intervention study in a cohort of pregnant women. Pre-intervention group (GI) was formed by 225 postpartum women (1999-2001) and the intervention group (GII), by 208 pregnant women (2005-2006) and their respective newborn infants assisted in a public maternity hospital of Rio de Janeiro, Brazil. In order to standardize data collection, field researchers were trained, recycled and supervised, and an instruction guide was developed. A pilot study was done to test the tools and assessment of data consistency was performed weekly. Evaluation of reliability used Kappa (k) statistics and intraclass correlation coefficient (ICC). RESULTS: The follow-up losses of GII were 8.4%. There was a good standardization of procedures to obtain information in both groups, with an ICC>0.82 and k>0.65 for GI and ICC>0.94 and k>0.71 for GII. There was a high concordance between the pre-pregnancy weight informed and the weight measured in the first trimester (ICC=0.96). CONCLUSIONS: The strategies adopted in the present study contributed to the low percentage of follow-up losses and to guarantee the quality of data, allowing reliability of results.
intervention studies; cohort studies; prenatal care
