The Ninth Revision of the International Classification of Disease (ICD) was used for defining homicide and its subclassifications. The homicide and its subclassification ICD codes (E960-E989) were used to abstract the homicide data from the death certificate files of the State of New Jersey Department of Health for the years 1989 through 1997. Only the Ninth Revision of ICD was applied during the period of this study; therefore, categorization change in ICD coding is not an issue. This principle of selection is consistent with the current standard, as proposed by the 1997 recommendations of the Injury Control Community. This study included a total of 1,317 homicides of which 494 homicide victims occurred among adolescents (age 15-19 years) and 823 homicide victims occurred among young adults (age 20-24 years) in New Jersey’s 21 counties for the eight-year period of this study (1989-1997). The urbanization index, density of population, and annual per-capita income data for New Jersey’s 21 counties were obtained from the U.S. Bureau of Census (1990 Census). The urbanization index was defined as percent of the population living in urban areas in each of New Jersey’s 21 counties. The density of population was defined as the number of persons residing per square mile of land in each of New Jersey’s 21 counties. The level of education of the victim and the means (weapons) used to commit homicide were obtained from death certificate files. The county-level educational achievement was measured as the percentage of the population with at least a high-school education.

Descriptive statistics include age, race, and sex-specific annual homicide incidence rates, the proportion of homicides committed by weapon type or means, and the correlation analysis.

The race-, sex-, and age-specific homicide rate analyses are limited to white and African-American persons because age-specific Hispanic, Asian, and American Indian populations by New Jersey’s 21 counties were too small to produce meaningful analyses. The ratio of homicide incidence rates was presented by race and gender. canadian antibiotics

The association between county-level, age-specific homicide incidence rates and county-level urbanization index, density of population, level of education, and annual per-capita income were assessed using the Spearman rank correlation. The SAS statistical programs were used for these analyses.