Avaliação da qualidade dos dados do Sistema de Informações sobre Nascidos Vivos (SINASC) em São Luís-MA e Ribeirão Preto-SP,

Abstract

The present study aimed to evaluate the quality of information (System coverage, completeness of the data fulfillment and reliability of the information) of SINASC in the city of Ribeirão Preto, São Paulo, through the linkage technique, comparing them to the data obtained from the project entitled Etiology factors of pre-term birth and consequences of perinatal factors on children's health , which studies birth cohorts in São Luís-Maranhão and Ribeirão Preto-São Paulo (BRISA cohort); as well as to analyze and compare the temporal trend of completeness of fulfillment of the variables of SINASC in these two Brazilian cities that make up the study of BRISA cohort, in a time series of 1996-2013. Evaluative study that analyzed the following variables: newborn sex, birth weight, Apgar indices, presence of congenital anomaly, race/color, mother's age, education, marital status, number of living children, length of gestation, type of childbirth, type of pregnancy and number of prenatal visits. The degree of concordance (reliability) of the information was evaluated by Kappa and intraclass correlation coefficients (ICC) for categorical variables and by the Bland-Altman method for numerical variables. Completeness referred to the degree of fulfillment of the field analyzed following the scoring system proposed by Romero and Cunha: excellent, when the variable presents less than 5% of incomplete fulfillment; good (5.0 to 9.9%); regular (10.0 to 19.9%); bad (20.0 to 49.9%) and very bad (50.0% or more). The trend analysis was done by estimation of polynomial regression models. In Ribeirão Preto, the estimated coverage of SINASC, in 2010, classified as good, was 88.3% (95% CI; 87.6% to 89.0%), according to the eight hospitals analyzed. In São Paulo city, the variables that showed nearly perfect or excellent agreement were: hospital of birth, newborn sex, type of pregnancy, type of childbirth, birth weight, maternal age and marital status. The variables number of prenatal visits and length of gestation obtained moderate reliability. The following variables had poor/low reliability: presence of congenital anomaly, race/color, Apgar score at 1st and 5th minutes and number of living children. On the completeness, the variables mother's age, marital status, type of pregnancy, type of childbirth, newborn sex, Apgar scores and birth weight had excellent completeness in the both cities studied. The variables congenital anomaly and race/color of the newborn obtained completeness ranging from bad to very in the capital of Maranhão, unlike the city of Ribeirão Preto, where they had a non-fulfillment percentage of less than 5%. The variable length of gestation showed variation in completeness for all time series, achieving excellence in fulfillment between the years 1999-2010. In Ribeirão Preto, this same variable had excellent completeness between 2000-2013. The chosen polynomial regression models pointed upward trend of non-fulfillment for the variable congenital anomaly in the two cities. In São Luís, the variable marital status had a declining trend for incompleteness, which did not occur in Ribeirão Preto. In São Paulo city, the variable maternal education was presented with decreasing trend of non-fulfillment, which was not evidenced in the capital of Maranhão. The other variables had decreasing trend for the incompleteness in the two cities with statistically significant values in the tested models. This research adds to the evidence of the potential of secondary data as an important source for epidemiological research. Although the results can be translated in advance, it is important to continuously invest in training in order to ensure a system capable of subsidizing policies of intervention and organization of maternal and child health.