variables associated with different rates or causes of child death. Variables typically examined include geographic location, age, sex, socioeconomic and ethnocultural characteristics, and community characteristics such as density, average income or income inequality, and rates of violence.6
Reflecting social, economic, physical, and other differences, states and regions show considerable variation in child mortality by cause. One stark contrast involves infant mortality. In 1999, the District of Colombia had the highest infant mortality rate (15.0 per 1,000 live births), followed by South Carolina (10.2 per 1,000 live births). Maine and Utah had the lowest rate in 1999 at 4.8 deaths per 1,000 live births (NCHS, 2001e).
In 1999, for those aged 0 to 19, Wyoming led the nation in motor vehicle fatality rates (23.5 per 100,000), followed by Mississippi (20.9 per 100,000). The lowest fatality rates were for Hawaii (3.6 per 100,000) and Rhode Island (3.8 per 100,000) (NCHS, 2001e). For motor vehicle fatalities involving all ages, factors contributing to differences in rates appear to include population density, proportions of light and heavy trucks in use, alcohol use, and delayed medical care (see, e.g., Muelleman and Mueller, 1996). Two single-state studies, one in Colorado (Hwang et al., 1997) and one in Alabama (King et al. 1994), reported higher death rates from motor vehicle crashes and unintentional injuries for children in rural areas. Another study reported that rural children ages 1 to 19 had a 44 percent higher death rate from injuries than their urban counterparts in 1992, with the greatest differences found in the 15 to 19 age group (Ricketts, 2000). (Reported differences in urban and rural death rates may vary depending on how rural and urban are defined [Farmer et al., 1993]).
Juvenile homicide rates also differ substantially among states. Maryland led the nation in 1999 with a homicide rate of 7.8 per 100,000, followed by Illinois at 7.25 per 100,000. Hawaii and Utah had the lowest
Reporting categories for published information from various data sets that include mortality are not completely standardized. For example, the federal government’s primary mortality report provides information by Hispanic origin and by race (white and black non-Hispanics (NCHS, 2001a). In contrast, published information from the NCHS linked data set of births and infant deaths subdivides infants by Hispanic origin (black and white) and by race (white, black, American Indian, and Asian or Pacific Islander (NCHS, 2000b). Because death information is linked to information collected at birth, the latter data set also includes more detailed individual data such as the mother’s age, educational attainment, marital status, place of birth (U.S. or foreign), and smoking during pregnancy. Information about the infant includes birth order, birth weight, period of gestation, and trimester when prenatal care began.