Widening Disparities in Infant, Neonatal, and Postneonatal Mortality among Major US Metropolitan Cities, 1985-2002

Infant mortality in US cities has been consistently higher than the national average. We examined differentials in infant, neonatal, and postneonatal mortality rates across 64 major US metropolitan cities between 1985 and 2002. Using the 1985-2002 national linked birth/infant death cohort files, we analyzed city differentials in infant mortality rates and risks using Cox regression and several inequality indices. Considerable disparities in infant mortality exist across major US cities, with the rate in 1999-2002 varying from a low of 4.2 per 1000 live births in San Francisco to a high of 15.5 for Birmingham, AL. When stratified by race, the rate varied from a low of 2.9 for white infants in San Francisco to 18.8 for black infants in Pittsburgh. The overall infant mortality range during 1985-1988 was 8-20.5. While the infant mortality rate for most cities declined by >=30%, it did not change significantly between 1985 and 2002 for Birmingham, St Louis, Milwaukee, Memphis, Wichita, Cincinnati, Cleveland, and Toledo. Overall disparities in infant, neonatal, and postneonatal mortality widened between 1985 and 2002, although the increase in postneonatal mortality disparities was more marked. Adjustment for maternal age, race/ethnicity, marital status, parity, prenatal care, education, smoking, and birthweight reduced city differentials. However, Birmingham, Cincinnati, Nashville, Memphis, Tulsa, Wichita, Milwaukee, and Omaha maintained 75-92% higher risks of infant mortality in 1999-2002 than San Francisco. Besides individual- level predictors, community-level factors, such as regional distribution of neonatal health services, poverty, unemployment, and racial segregation, may explain disparities in infant mortality.