Methodology of maternal and child health populational surveys: a statewide cross-sectional time series carried out in Ceará, Brazil, from 1987 to 2017, with pooled data analysis for child stunting

Abstract

Background: Cross-sectional studies are fundamental studies in the practice of epidemiological science. This article aims to present in detail the methodology for conducting a series of cross-sectional studies, as well as the analysis of data through pooled data. Methods: The series of studies are population cross-sectional studies, with statewide coverage, searching for representative sample of reproductive aged women and pre-school children in Ceará, Brazil. The sampling plan followed simple random, stratified, systematic and by conglomerates, in sequence. About 300 variables were collected. For each of the individual studies, multivariate data analysis was used to verify associations between dependent variables. For all the studies together, techniques used were trend chi-squared and pooled data analysis using linear mixed modeling procedures. Results: There were 6 studies in sequence, for 30 years. Among other findings, the variables income, maternal education and breastfeeding time proved to be associated with the reduction of malnutrition in children considering all the period (p values 0.013, 0.033 and 0.037, respectively). Conclusions: Cross-sectional studies can be replicated at regular time series following the methodology exposed in this, even for locations with limited resources, ensuring adequate management of decisions of using federal funding aimed at achieving targeted programs to maximize the results obtained with the public resource available.