This study was determined to be exempt from institutional review board review by the Northwestern University institutional review board because research involving deceased persons is not considered human subjects research, and the data are publicly available and deidentified. The study data source was deidentified retrospective case-level surveillance data from the Cook County Medical Examiner Case Archive [9]. These data were collected after death was declared; therefore, no informed consent was possible or required. To protect against possible identification of individuals in the data set, we suppress counts for cells with fewer than 5 in table presentations.

Gun-homicide decedent data were from Cook County Medical Examiner Case Archive data [9]. The archive includes data on decedent characteristics and fatal incident date and location. Population data came from American Community Survey [11].

This study included all homicides in which the fatal injury occurred in the city of Chicago with the death occurring between January 1, 2015, and December 31, 2021, where the manner of death was ruled “homicide” and the “gun-related” variable was endorsed by the Cook County Coroner’s Office. The start date is January 1, 2015, because that is the first year that electronic surveillance of these deaths had complete case ascertainment.

For our analyses, we used a continuous variable for age in years, as well as a categorical variable where we collapsed the decedent age in years into the following age groups: <15, 15-19, 20-24, 25-34, 35-44, 45-54, 55-64, and 65+ years. Age groups were based on standard age groups generated from the US Census. We also created a race-ethnicity-sex group variable by assigning decedents into 1 of the following 6 race/ethnicity and sex groups: non-Hispanic Black female, non-Hispanic White female, Hispanic female, non-Hispanic Black male, non-Hispanic White male, and Hispanic male. We did not include data on Asian, Native Hawaiian or Other Pacific Islander, or American Indian or Alaska Native racial groups due to the low numbers of decedents, which makes comparisons of distributions unreliable.

We calculate the overall Chicago annual gun-homicide rate per 100,000 persons as the annual number of gun-homicides divided by the 2020 Chicago population. For analyses, we assigned decedents to 1 of 6 race-ethnicity-sex groups (non-Hispanic Black female, non-Hispanic White female, Hispanic female, non-Hispanic Black male, non-Hispanic White male, and Hispanic male). We graphed the mean age of gun homicide decedents by the year of death and used Pearson r to evaluate the association between mean age and the year of death. We described the proportion of decedents by race-ethnicity-sex and age groups over time and tested for statistically significant change over time using comparisons of column proportions with significance set at P≤.05. The comparison of mean age by race-ethnicity-sex group was done using 1-way ANOVA with significance set at P≤.05. We also examined changes in annual gun-homicide rates per 100,000 persons by age group using the annual number of gun-homicide deaths by age group divided by the annual Chicago population for that age. Analyses were done in IBM Statistics 27. We used graphing procedures in Microsoft Excel 2016, IBM Statistics 27, and Tableau (Salesforce Inc) for data visualizations.

The annual distribution of gun-homicide decedents by race-ethnicity-sex group had been relatively stable over time, even in the context of year-to-year changes in gun-homicide rates and the 2015-2021 increase in gun-homicide rates in Chicago. There was a large but nonstatistically significant increase in the proportion of non-Hispanic Black female gun-homicide decedents from 2015 (3.6%) to 2021 (8.2%). The lack of statistical significance may be due to a relatively small (238/3723, 6.4%) number of non-Hispanic Black female individuals among gun-homicide decedents. The mean age of Chicago gun-homicide decedents increased by 3.27 years from 2015-2021. This increase in mean age coincided with a decrease in the proportion of non-Hispanic Black male gun-homicide decedents between the ages of 15-19 and 20-24 years and, conversely, an increase in the proportion of non-Hispanic Black male gun-homicide decedents between the ages of 25-34 years.

Although there is much discussion of firearm violence in Chicago in news publications [12] and, less often, research literature [13,14], there is little detailed information available on the demographics of gun-homicide decedents in the face of changes in violence other than to note broad populations at the highest risk at one point in time. This study examined the intersection of race/ethnicity, sex, and age among of Chicago gun-homicide decedents over a 7-year period in which there were year-to-year changes in overall gun-homicide rates. It provides key information on group representation in gun homicide during periods of escalation and periods of retraction. The changes in the makeup of gun-homicide decedents we detected signal a need for expanded prevention outreach and intervention engagement marketed toward non-Hispanic Black female individuals of all ages and non-Hispanic Black male individuals between the ages of 25-34 years, while at the same time, efforts are needed to expand outreach to and engagement of younger non-Hispanic Black male individuals, who consistently have the highest gun-homicide rates in Chicago.

This study has several strengths and limitations. In terms of strengths, the study incorporated very recent and reliable surveillance data from the Cook County Medical Examiner’s Office, a nonpartisan office led by an appointed chief medical examiner. However, it is well known that death certificate reporting of race and ethnicity underreports certain subgroups [15], and so, it is likely that certain race/ethnicity groups, especially Alaska Native or American Indian individuals, those with more than one race/ethnicity affiliation, and Hispanic individuals may be underreported in the Medical Examiner data used in this study. Regarding other limitations, our reliance on US Census age groups as denominators for calculating homicide rates prevents a more nuanced analysis of single-year age risks. Further, although we report on demographics of gun-homicide decedents, ideally, more information on the circumstances surrounding these gun-related homicide deaths is needed to inform intervention approaches.

The findings from this modest study provide a basis for informing firearm violence prevention and intervention strategies and can inform a research agenda going forward. Further research could address the following questions: Will this shift toward increased non-Hispanic Black female and non-Hispanic Black male gun-homicide decedents between the ages of 25-34 year endure over time? What are the drivers associated with this increase in the proportion of non-Hispanic Black female gun-homicide decedents? and Are these age and gender shifts accompanied by changes in the precipitators and features of gun-homicide events? Because gun-homicide is a leading public health problem in the United States and, even more so, locally in Chicago, understanding the stability of and changes within decedent demographic subgroups is part of a required first step to inform targeted prevention responses.