The effect of using "race of child" instead of "race of mother" on the black-white gap in infant mortality due to birth defects.

OBJECTIVE: For at least 20 years, birth defects have been the leading cause of infant mortality in the United States. Some studies have reported higher rates for black infants than white infants of mortality due to birth defects, while other studies have reported no black-white differences. The authors analyzed the effect on these rates of a change in the way the National Center for Health Statistics (NCHS) tabulates "race" for newborns.
METHODS: The authors calculated infant mortality rates due to birth defects for 1980-1993 using two standard methods of assigning newborns to "racial" categories: a "race of child" algorithm and the "race of mother" approach currently used by NCHS.
RESULTS: From 1980 through 1993, birth defect-specific infant mortality rates (BD-IMRs) were significantly higher for black infants than white infants 12 of the 14 years by "race of mother" and only 5 of 14 years by "race of child." Calculation of BD-IMRs by "race of mother" reduced the rate for white infants and increased the rate for black infants in each of the 14 years. The choice of method for assigning newborns to "racial" categories had a progressively greater effect over time on the black-white gap in BD-IMRs.
CONCLUSIONS: Calculations of trends in "race"-specific BD-IMRs by may vary substantially by whether "race of mother" or "race of child" is used. Identifying the method of tabulation is imperative for appropriate comparisons and interpretations.