T.J. Mathews, NCHS Demographer
Questions for T.J. Mathews, Demographer and Author of “State Variations in Infant Mortality by Race and Hispanic Origin of Mother, 2013-2015”
Q: What made you decide to focus on maternal race and Hispanic origin in this state-by-state analysis of infant mortality in the United States?
TM: We decided to focus on maternal race and Hispanic origin from this most recent special data set because previous research has shown a wide variation in infant mortality rates within, and between, race and Hispanic origin groups at the national level. We were particularly interested in seeing how rates vary by race and Hispanic origin by state, which groups had the largest within-group differences by state, and how much variation there was by state among non-Hispanic whites, non-Hispanic blacks, and Hispanics.
Q: Was there a result in your study’s analysis of infant mortality that you hadn’t expected and that really surprised you?
TM: Although we expected to see significant variations throughout race and Hispanic-origin groups, we were surprised by how much rate variation there was for specific race and Hispanic-origin demographics across states. Specifically, our finding that infants of non-Hispanic white women had the widest range in rates by state—the highest was 2.8 times as high as the lowest – was especially striking. The most interesting result from our analysis is what we found for infants of non-Hispanic black women. The lowest infant mortality rate of 8.27 in Massachusetts was higher than the highest state rates for infants of non-Hispanic white women (7.04 in Arkansas) and Hispanic women (7.28 in Michigan).
Q: What differences, if any, did you see in infant mortality among race and ethnic groups?
TM: There were several differences in infant mortality among race and Hispanic origin groups. Specifically, as in previous research, non-Hispanic blacks had significantly higher infant mortality rates than non-Hispanic whites and Hispanics. Although non-Hispanic whites had lower infant mortality, the range in rates was larger than for the other groups examined in this study.
Q: What does “linked birth and infant death data” mean exactly – which is what you reference as a source for the statistics in your report?
TM: The linked birth and infant death data is a file that connects data from infant death certificates with their birth certificate data, allowing us to obtain more accurate reporting of race and Hispanic origin group identity. Further items such as birthweight, period of gestation, plurality, prenatal care, maternal education, live birth order, marital status, and maternal smoking are included from the birth certificate that allow more detailed analyses related to infant mortality in the United States. This analysis from the linked data file is the most detailed source for infant mortality data.
Q: Why did you focus your analysis on a handful of years, 2013-2015, and not a longer term trend study going back further in time?
TM: We combined the data years 2013-2015 and did not look at each year individually or across years. So, we weren’t assessing a trend at all in this report (as we have in the past), but rather focused on ensuring we had a large enough sample size to get reliable estimates by race and Hispanic origin for a majority of the states. This new analysis is really about health disparities in U.S. infant mortality by state.
Q: Do you have any cause of death data for the infants, or health insurance coverage data for the mothers and infants in this study group?
TM: For this particular project, we did not look at the causes of death for the infants or the payment source used for the birth. We do not have data that would indicate the health insurance coverage for the infant at time of death. However, the linked birth and infant death files do contain this information in the public use data sets from the information included on the birth certificates in the data files.
Q: What is the take-home message of this report?
TM: The main take home message of this report is found in the evidence of health disparities in U.S. infant mortality. The differences – the rate variation across geographic and race and ethnic demographics – is staggering. Although the overall U.S. infant mortality rate has decreased over time, these gains have not been equally distributed across race and Hispanic origin groups or across the United States. The significant differences among some states indicates that states might be able to decrease infant mortality rates by focusing on specific demographic groups.