Linked Birth/Infant Death Data Set
The linked Birth/Infant Death data set links information from the birth certificate to information from the death certificate for each infant death in the United States. The purpose of the linkage is to provide a data set that adds the many additional variables from the birth certificate to the death certificate record to enable more detailed analyses of infant mortality patterns. The linked file is used for calculating infant mortality rates by race and ethnicity, which are more accurately measured from birth certificate data than from data on the death certificate.
Infant mortality rates are based on infant deaths per 1,000 live births. An infant deaths is defined as a death before the infant’s first birthday.
Linked birth/infant death data files are available in two different formats: period data and birth cohort data. The numerator for the period linked file consists of all infant deaths occurring in a given data year linked to their corresponding birth certificates, whether the birth occurred in that year or the previous year. The numerator for the birth cohort linked file consists of deaths to infants born in a given year. In both cases, the denominator is all births occurring in the year. For example, the 2010 period linked data set contains a numerator file that consists of all infant deaths occurring in 2010 that have been linked to their corresponding birth certificates, whether the birth occurred in 2009 or 2010. In contrast, the 2010 birth cohort linked data set contains a numerator file that consists of all infant deaths to babies born in 2010, whether the death occurred in 2010 or 2011. Although the birth cohort format has methodological advantages, it creates substantial delays in data availability because it is necessary to wait until the close of the following data year to include all infant deaths in the birth cohort. Therefore, period linked files are used to compute infant mortality rates for Healthy People 2020.
Record weights are used to compensate for the 1%–2% of infant death records that could not be linked to their corresponding birth records. In addition, not-stated birth weight is imputed if the period of gestation was known. This imputation is done to improve the accuracy of birth weight-specific infant mortality rates because the percentage of records with non-stated birth weight is generally higher for infant deaths (3.2% in 2010) than for live births (0.1% in 2010). In 2009, non-stated birth weight was imputed for 0.07% of births. In 2010, non-stated birth weight was imputed for 0.08% of births.
About 98%─99% of infant death records can be linked to their corresponding birth certificates.
A new revision of the birth certificate was introduced in 2003 and is being adopted by states on a voluntary, rolling basis. Data on mother’s educational attainment, tobacco use during pregnancy, and prenatal care based on the 2003 revision are not comparable with data based on the 1989 revision of the U.S. Standard Certificate of Live Birth.
Mathews TJ, MacDorman MF. Infant mortality statistics from the 2006 period linked birth/infant death data set. National vital statistics report; vol 58 no 17. Hyattsville, MD: NCHS; 2010. Available from: Infant Mortality Statistics From the 2006 Period Linked Birth/Infant Death Data Set.