¿Qué tan bien recuerdan las madres? Una validación de datos de desempeños al nacer para Uruguay

To assess maternal and child well-being and health goals derived from national and international agreements, many household surveys collect information on reproductive, maternal and child health. However, scarce representative validation exercises are comparing these data to administrative records information at the individual level. In this study, we contrast household surveys against data from clinical histories and vital statistics information. Also, we explore whether the significance and magnitude of the relationships between the variables change depending on the data source used. Based on the merging of vital statistics microdata with the official Uruguayan Nutrition, Child Development and Health Survey (ENDIS), we analyze the accuracy in mothers recall of weeks of gestation and prenatal visits. In general terms, our exercise validates the use of survey data, even though the identification of premature births and early prenatal visits have significant differences. We find that the accuracy of the recall by mothers is high and the discrepancies are associated with lower maternal education levels, the recall period (child´s age), and the birth order. However, the incidence of prematurity and early uptake is underestimated by 20%. Although the (respective) association of gestational weeks and prenatal controls with birth weight shows similar signs in both data sources, the levels of significance and effects magnitude appear to be greater in the administrative records.