Deaths from homicide are a major public health problem in the United States and throughout the world. Further, the United States has experienced homicide rates that are 6.9 times higher than those in other high-income countries, and the vast majority are attributable to firearms.1 For many decades there have been striking regional variations among various cities throughout the United States in deaths from homicide, and size is a major determinant.2 In addition, there has been considerable publicity about high rates of homicide in some major cities like Chicago and low rates in others like New York City.3
We noted that New York City (specifically, each of the 4 boroughs of the Bronx, Brooklyn, Manhattan, and Queens) and Baltimore City are defined as peers by the US Centers for Disease Control and Prevention according to 19 population-based characteristics (population size, percent foreign born, median household income, population growth, percent high school graduates, receipt of government financial assistance, population density, single-parent households, Gini index of income inequality, population mobility, median home value, overall poverty, percent children, housing stress, elderly poverty, percent elderly, percent owner-occupied housing units, unemployment, and sex ratio).4 In addition, New York City and Baltimore City have been identified as components of the “northeastern megalopolis,”5 a highly and densely populated area of interconnected communities constituting an organic cultural region with a distinct history and identity, occupying a roughly similar physical environment, linked through a major transportation infrastructure and forming a functional urban network via goods and service flows, ultimately establishing a usable geography that is suitable for large-scale regional planning.6 This provided a unique opportunity to explore secular trends in homicide rates between New York City and Baltimore City.
Methods
For overall and race-specific (blacks and African Americans [blacks] and whites) homicide mortality from birth to 85+ years of age, we ascertained underlying cause of death from the Compressed Mortality File (1979-1998 and 1999-2016)7, 8 as presented on the United States Centers for Disease Control and Prevention Wide-ranging ONline Data for Epidemiologic Research (WONDER) public Internet website. These data, which include age-adjusted rates and 95% confidence intervals, are based on death certificates. Death certificates are issued for decedents who are legal residents of the United States and are estimated to capture 99% of all such deaths.9Moreover, the accuracy of assault mortality data is very high.10 We used the International Classification of Disease (ICD) IX Codes E 960-E969 (Homicide and injury purposely inflicted by other persons) from 1979 to 1998, and ICD Codes X85-Y09 (Assault) and Y 87.1 (Sequelae of Assault) from 1999 to 2016.8 All rates included at least 20 deaths, the criterion for reliability established by the National Center for Health Statistics.7, 8
Results
Figure 1 compares murder rates between Baltimore City and New York City from 1979 to 2016. While rates were similar and increased during most of the 1980s, New York City rates declined beginning in the 1990s while those in Baltimore City have remained high.
To read this article in its entirety please visit our website.
-Robert S. Levine, MD, Rebecca P. Schneid, Roger J. Zoorob, MD, MPH, Charles H. Hennekens, MD, DrPH
This article originally appeared in the January issue of The American Journal of Medicine.
