A Test of Criteria for Determining "Optimal" PPOR Model Reference Groups: Measuring Excess Infant Deaths in the Border Region
Background: Infant mortality is one of the strongest indicators of social and economic development in all countries. Understanding infant mortality is critical for understanding the health of nations. Strategies for lowering infant death rates are needed. The Perinatal Periods of Risk Model (PPOR) is an approach that allows communities to determine when infant death and excess death is occurring by categorizing death into four perinatal categories. Excess infant deaths are determined by defining a reference group with the lowest mortality rate, and then comparing the reference group rate to other subgroups. The typical reference group has been defined as white, non-Hispanic females, 20+ years of age, with 13+ years of education. Whether this reference group in fact captures the lowest possible infant death rates is rarely examined. This study examined how changing the criteria for education and ethnicity at the national, external, and internal level altered the estimated infant mortality rate and estimation of excess infant deaths. Aims: This study (1) examined the number and rate of infant deaths by perinatal period of risk (PPOR) category for new reference groups defined by maternal education; and (2) tested differences in excess infant deaths calculated using infant death rates for typically and newly defined reference groups. Methods: This was a secondary data analysis of the dataset previously collected by the Center for Disease Control and Prevention (CDC). Alternate national, external, and internal reference groups for estimating infant mortality and excess deaths were created Each of the three levels included three reference groups defined according to mother’s years of education. Results: The data set included 120,777 cases of infant death. In the non- U.S. border states, infant mortality was significantly higher for all of the newly defined groups as compared to the national typically defined group. Significant differences were also found between the external (U.S. border states) typically defined reference group and the newly defined reference group with less than 9 years of education (chi-square=21.815, p<0.001); and between the internal (U.S. border counties) typically defined reference group and the newly defined reference group with education between 9 and 13 years (chi-square=8.921, p<0.012). Excess infant deaths existed in the almost all newly defined categories. At the national level, education had a marked impact on infant deaths among Hispanic women. Hispanic women with less than nine years education experienced an excess death rate of 10 per 10,000 live births, while Hispanic women with 9-13 years of education experienced an excess death rate of 8 per 10,000 live births. At the internal level, Hispanics residing on the border with 13+ years of education experienced no excess death, while Hispanics residing on the border with 9-13 years of education experienced an excess death rate of 7 per 10,000 live births. Conclusions and Recommendations: Excess death rates in all categories were inversely proportional with education. However, excess death rate was more influenced by a mother’s education level than by a mother’s ethnicity (newly defined reference groups with 13+ years of education did not have meaningful differences in excess death). The completion of phase two PPOR analyses needs to be completed to determine which risk and preventative factors in the maternal health/prematurity category would have the largest impact on improving infant mortality.^
Ramos, Alexis Nicole, "A Test of Criteria for Determining "Optimal" PPOR Model Reference Groups: Measuring Excess Infant Deaths in the Border Region" (2018). ETD Collection for University of Texas, El Paso. AAI10817258.