Background: Birthweight and gestational age are key indicators for perinatal health obtained through the birth certificate. Knowing the validity of birth certificate data is crucial when identifying needs and evaluating birth outcomes. In Spain, vital information is reported by parents and is not checked for consistency with any hospital document. Therefore, to perform a validation study and describe the variables associated with misreporting is essential to improve the quality of birth certificates.
Methods: A database was created using birth information from hospital medical records that were individually linked with the Spanish National Institute of Statistics (INE) birth certificate files. Measures of reliability and validity were used to compare the data from the two sources. Logistic regression models were adjusted to model the odds of being misreported in birthweight and gestational age, controlling for relevant variables.
Results: INE overestimated the prevalence of birthweight and gestational age. The degree of agreement between the two sources was good for most gestational age groups (Kappa = 0.74), very good for the very preterm (Kappa = 0.85) and very good also for all categories of birthweight (Kappa = 0.88). Misreporting was significantly higher among immigrants, unmarried mothers and girls. Being a preterm birth increased the odds of being declared with errors in gestational age; having low birthweight and missing information on gestational age were associated with misreporting birthweight.
Conclusions: The reliability of INE information could be greatly improved if hospitals included birthweight and gestational age on the document provided to parents for registering the birth.
© 2012 Blackwell Publishing Ltd.