An assessment of the data quality for live births and infant deaths for the 1998-birth cohort, city of São Paulo, using information from SINASC and SIM

In this article we use data from the 1998 birth-cohort from the City of São Paulo for live births and data on deaths from the same birth cohort that took place before the first year of life. Information on live birth came from the Information System of Live Births (SINASC) and on deaths came from the Information System on Mortality (SIM). Our aim was to evaluate two aspects of data quality, which are the percentage of missing information and consistency of the data based on empirical regularities observed in other populations or with the criterion of plausibility. Preference for digits, sex ratios at birth and at death and the Apgar score distribution were examples of quality indicators evaluated in this population. We concluded that information is much more frequently recorded for live births than for infant deaths (except for race/color of the infant) and that the internal consistency of the information can be considered satisfactory for the study population.