Trends in Hospital Care for Intentional Assault Gunshot Wounds Among Residents of Cook County, Illinois, 2018–2020

Abstract

Objectives. To examine gun violence with respect to hospital visits for treatment of intentional assault gunshot wounds (IGWs).

Methods. IGW-coded visits among residents of Cook County, Illinois, were matched to census zip code tabulation areas (ZCTAs) to map changes in IGW visit frequencies between 2018 and 2020. Patient characteristics were compared across years, and Poisson regression models for the likelihood of an inpatient admission or in-hospital death were estimated.

Results. Over the study period, Cook County residents made 7122 IGW-coded hospital visits to 89 Illinois hospitals, resulting in $342 million in charges and 24 894 hospital days. The number of visits almost doubled between 2018 and 2020, from 1553 to 3031; 6 ZCTAs had increases of more than 60 visits. Approximately one third of patients with a visit were admitted, and 6.5% died.

Conclusions. Hospital statistics do not include the full toll of nonfatal gun injuries or the costs of related community-level trauma. The health care system remains crucial in implementing epidemiological approaches to violence prevention. Addressing the national spike in shootings will require large investments in community economic development and a professional public safety workforce. (Am J Public Health. 2022;112(5):795–802. https://doi.org/10.2105/AJPH.2022.306747)

The 2020 increase in gun violence in Cook County, Illinois, mirrors similar trends in many large urban areas.¹ Although the per capita gun homicide rate in Chicago, Illinois, was not among the highest (and was lower than that of St. Louis, MO, or Detroit, MI), it had the largest increase in the number of gun homicides of any US city in 2020.¹ Shootings have increased in suburban Cook County as well. In this study, we examined the alarming increase in gun violence in Cook County in terms of hospital visits for treatment of intentional assault gunshot wounds (IGWs).

Most studies of trends in gun violence rely on police crime records and homicide rates in particular. Gun homicides appear to account for about 20% of identified shooting victims, with a much larger number of nonfatal gun injuries.² Because hospital emergency departments are a critical component of evolving violence prevention programs³ and are legally obligated to report gun injuries to the police, we sought to study patterns of emergency department and inpatient hospital use for IGWs.

We used administrative hospital data to study escalating gun violence in Cook County between 2018 and 2020. A number of previous studies have shown that hospital IGW data significantly underrepresent the incidence of nonfatal gunshot wounds.^4,5 For example, hospital data do not include homicide victims who die before reaching the hospital or those who refuse medical treatment. Moreover, these data often suffer from inconsistent International Classification of Diseases (ICD) coding. It was nevertheless of interest to study IGW visits as part of an overview of the 2020 spike in gun violence in Chicago and surrounding suburban areas.

METHODS

All records from emergency department visits and inpatient hospitalizations for residents with zip codes in Cook County that were coded for IGWs were obtained from the Illinois Hospital Association’s Comparative Health Care and Hospital Data Reporting Services database. Cook County includes 5.27 million residents, of whom 2.74 million live in the city of Chicago. We analyzed data for all IGW-coded visits by Cook County residents at Illinois nonfederal hospitals from January 2018 to December 2020, excluding 0.5% of visits with missing zip code data. To provide a perspective on the extent to which IGW-coded hospital data reflect larger trends in crime data, we compared annual IGW hospital visits with publicly available Chicago Police Department (CPD) data on nonfatal gunshot injuries.⁶

Hospital administrative data included patient sex, age, Chicago versus suburban Cook County residence, insurance status, race and ethnicity, and whether the visit occurred on a weekend. We used 2019 American Community Survey zip code tabulation area (ZCTA) data to classify patients according to their area’s level of poverty (< 5%, 5%–9.99%, 10%–19.99%, ≥ 20% of households categorized as poor). Changes between 2018 and 2020 in the number of IGW hospital visits in each ZCTA were mapped with Esri ArcGIS Desktop 10.8.1 software (Redlands, CA). Changes in IGW visit frequencies in ZCTA “hot spots” in 2020 were categorized as reductions, no changes, less than 5 additional visits, 5 to 19 additional visits, 20 to 59 additional visits, and 60 or more additional 2020 visits.

We used the χ² test to determine the significance of differences in patient characteristics by year. Poisson regression models with robust variance estimates were used to analyze the likelihood of inpatient admissions and hospital deaths after control for patient characteristics with standard errors adjusted for clustering of visits within hospitals.⁷ Statistical analyses were conducted with Stata version 16 (College Station, TX).

RESULTS

Over the 36-month study period, Cook County residents made 7122 IGW-coded hospital visits to 89 Illinois hospitals. Almost half (48%) of these institutions had 5 or fewer IGW visits, with almost 4000 visits occurring at just 3 institutions. Each year approximately 15% of IGW visits were made by patients from suburban zip codes. Figure 1 displays the monthly totals for all IGW visits as well as for inpatient admissions and deaths. Gun violence in 2018–2019 peaked in the summer months, and the number of gunshot wound visits increased from 136 in July 2018 to 337 in July 2019. In 2020 there was a similar summer spike, with 336 visits in July (directly after the lifting of the COVID-19 pandemic stay at home order). However, unlike the previous years, 2020 IGW visits then continued to increase into the fall, with a new peak of 432 in October. Total annual IGW emergency department and inpatient visits almost doubled from 1553 in 2018 to 3031 in 2020.

FIGURE 1—

Monthly Trends in Intentional Assault Gunshot Wound Visits, Inpatient Admissions, and Hospital Deaths by Cook County Residents at 89 Illinois Hospitals: January 2018–December 2020

Figure 2 displays a map of ZCTA area changes in IGW visits between 2018 and 2020. There were 24 ZCTAs with reductions in IGW visits and 27 with no changes in visits. Although 85% of the ZCTAs with no changes had less than 5 IGW visits in each study year, one of those ZCTAs had 26 IGW visits in each year. There were 37 ZCTAs with less than 5 visits, 34 with less than 20 visits, 16 with 20 to 59 visits, and 7 with 60 or more additional visits between 2018 and 2020. Hot spots included a North and South Lawndale (Chicago) ZCTA with 115 additional 2020 visits and an Austin ZCTA with 107 additional visits. The 23 ZCTAs with more than 20 IGW visits accounted for 61.6% of all visits among county residents.

FIGURE 2—

Changes in Zip Code Tabulation Area Emergency Department and Inpatient Intentional Gunshot Wound Visits by Cook County Residents at 89 Illinois Hospitals: January 2018–December 2020

Data on the characteristics of patients with IGW visits are presented in Table 1. IGW visits increased the most among those from more affluent ZCTAs while decreasing among those from the poorest ZCTAs (from 61.8% in 2018 to 49.4% in 2020). Although statistically significant, yearly age group and race/ethnicity differences were modest. Overall, 8.8% of IGW visits occurred among patients younger than 18 years, 75.6% among non-Hispanic Black patients, and only 3.5% among non-Hispanic White patients. Medicaid coverage increased from 50.9% in 2018 to 59.7% in 2020, whereas IGW visits among uninsured patients fell from 28.9% to 23.0%. There were no significant yearly differences with respect to weekend visits (which were overrepresented at 35.5% across the study years).

TABLE 1—

Characteristics of Patients With Emergency Department or Inpatient Visits for Intentional Assault Gunshot Wounds: Cook County, Illinois; January 2018–December 2020

	% Intentional Gunshot Wound Visits
All (n = 7122)	2018 (n = 1553)	2019 (n = 2538)	2020 (n = 3031)
Sex
Female	11.5	9.9	11.0	12.5
Male	88.5	90.1	89.1	87.5
ZCTA % poor households
< 5	2.6	2.5	2.5	2.7
5–9.99	10.1	9.1	9.1	11.3
10–19.99	35.8	26.7	40.0	36.6
≥ 20	51.5	61.8	48.3	49.4
Age group, y
0–17	8.8	9.7	9.8	7.6
18–24	32.6	33.5	33.2	32.1
25–29	21.2	18.7	21.8	22.0
30–45	28.4	28.1	27.2	29.1
> 45	9.0	10.0	8.1	9.2
Insurance status
Private	15.1	16.9	15.3	14.2
Medicaid	56.6	50.9	54.8	59.7
Medicare	3.0	3.4	2.7	3.1
Uninsured	25.2	28.9	27.2	23.0
Race/ethnicity
Other or unknown	9.4	10.7	11.0	7.8
Non-Hispanic White	3.5	3.0	3.4	3.8
Non-Hispanic Black	75.6	74.8	76.1	75.6
Hispanic	11.5	11.5	9.5	12.9
Weekend visit	35.5	35.1	36.2	35.3
Residence
Chicago	84.7	83.2	84.9	85.1
Suburban Cook County	15.3	16.8	15.1	14.9
Inpatient admission	36.7	44.4	27.2	39.6
Patient died	6.5	7.2	5.5	7.0
Annual total charges, $	342 778 141	67 205 271	94 542 986	181 029 884
Total annual hospital days	24 894	6 452	7 023	11 419

Note. ZCTA = zip code tabulation area. The sample consisted of 7122 visits by Cook County residents to 89 Illinois hospitals. All comparisons by year were significant at P < .01 except weekend visit (P = .80), residence (P = .34), patient died (P = .09), and sex (P = .02).

More than one third of IGW visits resulted in inpatient hospitalizations. There were 6452 hospital days for IGW patients in 2018 versus 11 419 days in 2020. Combined inpatient and emergency department visits accounted for approximately $67 million in hospital charges in 2018, increasing to more than $181 million in 2020. Over the entire 36-month study period, IGW visits accounted for more than $342 million in charges and 24 894 inpatient hospital days.

Table 2 presents Poisson regression results for the likelihood of inpatient admission and hospital death. Men were almost twice as likely to die in the hospital as women. Patients from higher-poverty ZCTAs had a significantly reduced likelihood of death. Uninsured patients had a 40% lower likelihood of being admitted than privately insured patients and a 44% higher likelihood of death; conversely, Medicaid patients had a higher likelihood of inpatient admission and a lower likelihood of dying in the hospital. Hispanic patients were 42% less likely to die in the hospital than non-Hispanic Black patients, but there were no other significant differences between years or by patient race/ethnicity.

TABLE 2—

Poisson Regression Results for Likelihood of Inpatient Admission or Death Among Patients Visiting Hospitals for Intentional Assault Gunshot Wounds: Cook County, Illinois; January 2018–December 2020

	Inpatient Admission, %	Inpatient Admission, IRR (95% CI)	Died, %	Died, IRR (95% CI)
Sex
Male	36.8	1.12 (0.75, 1.66)	6.9	1.97 (1.08, 3.60)
Female	31.4	1 (Ref)	3.3	1 (Ref)
Day of visit
Weekend	33.5	0.91 (0.81, 1.02)	6.8	1.04 (0.91, 1.18)
Weekday	37.7	1 (Ref)	6.3	1 (Ref)
Residence
Suburban Cook County	37.2	1 (Ref)	6.0	1 (Ref)
Chicago	34.8	1.12 (0.75, 1.66)	6.9	0.93 (0.72, 1.21)
ZCTA % poor households	37.2
<5	42.2	1 (Ref)	15.7	1 (Ref)
5–9.99	41.7	0.99 (0.79, 1.23)	6.6	0.45 (0.23, 0.87)
10–19.99	33.5	0.84 (0.65, 1.08)	7.0	0.44 (0.26, 0.77)
≥20	36.7	0.88 (0.74, 1.07)	5.7	0.35 (0.22, 0.57)
Age group, y
0–17	32.1	0.85 (0.75, 0.96)	5.7	1.17 (0.94, 1.44)
18–24	34.4	1 (Ref)	5.8	1 (Ref)
25–29	33.4	0.99 (0.90, 2.08)	6.7	1.13 (0.93, 1.38)
30–45	38.2	1.08 (0.98, 1.19)	7.1	1.27 (1.00, 1.61)
>45	46.2	1.15 (0.95, 1.38)	7.3	1.37 (0.88, 2.13)
Insurance status
Private	37.1	1 (Ref)	6.9	1 (Ref)
Medicaid	41.2	1.15 (1.02, 1.29)	4.9	0.71 (0.55, 0.92)
Medicare	62.1	1.54 (1.29, 1.82)	4.2	0.57 (0.25, 1.33)
Uninsured	22.0	0.60 (0.44, 0.82)	9.9	1.44 (1.11, 1.87)
Race/ethnicity
Non-Hispanic White	40.7	1.06 (0.89, 1.25)	5.2	0.66 (0.31, 1.41)
Other or unknown	44.3	1.34 (1.04, 175)	10.0	1.44 (0.77, 2.67)
Non-Hispanic Black	34.4	1 (Ref)	6.4	1 (Ref)
Hispanic	40.5	1.21 (1.02, 1.45)	4.6	0.58 (0.37, 0.92)
Year
2018	44.4	1 (Ref)	7.2	1 (Ref)
2019	27.2	0.62 (0.49, 0.77)	5.5	0.75 (0.56, 1.02)
2020	39.6	0.87 (0.67, 1.13)	7.0	1.02 (0.68, 1.52)

Note. CI = confidence interval; IRR = incident rate ratio; ZCTA = zip code tabulation area. The sample consisted of 7122 visits by Cook County residents to 89 Illinois hospitals.

The CPD reported 9584 nonfatal gunshot injuries in Chicago during the years covered by our study.⁶ The ratio of hospital IGW visits to CPD nonfatal gunshot victims varied from 54% in 2018 to 96% in 2019 and 75% in 2020. This variation resulted in IGW-coded visits among county residents increasing by 95% between 2018 and 2020, as compared with only a 42% increase in CPD-reported nonfatal victims.

DISCUSSION

Chicago has a long history of gun violence and had experienced a previous spike in gun homicides in 2016. The underlying social conditions in poor, segregated Chicago neighborhoods, and increasingly in south suburban Cook County, reflect generations of structural racism and disinvestment. Although these long-lasting conditions are well-known underlying causes of violence, there had been a 20-year trend of declining gun violence in US cities including Chicago.^8,9 Chicago gun violence had peaked in the 1990s and, following nationwide trends, declined by more than 50% by 2014, when there were 416 gun homicides. An increase in gun homicides began in 2015 and there was a sudden and sharp spike in January 2016, with 764 deaths occurring in that year. In 2016 the percentages of gun homicide victims who had at least 1 previous arrest (80%), a prior gun-related arrest (30%), and more than 10 prior arrests (40%) remained similar to previous years.¹⁰

The 2016 gun violence spike began the same month that the Laquan McDonald police shooting video was finally made public, and some attributed the spike in violence to a “Ferguson effect” in which trust in the police collapsed in some neighborhoods.¹¹ It occurred during a state budget crisis that cut violence prevention and social service funding. That same month the CPD introduced new documentation requirements that significantly lowered investigatory street stops. The homicide “clearance rate,” which reflects the number of murder arrests, declined from an average of 45% in 2013–2014 to 26% in 2016 and has remained dismal in the years since, fueling the likelihood of retaliatory killings and an unwillingness of witnesses to come forward.¹⁰ However, a University of Chicago Crime Lab study showed that there was little direct evidence supporting any particular theory regarding the cause of the 2016 spike.¹⁰ After falling in 2017–2018, gun violence reached a new high during the COVID-19 epidemic.

Gun Violence During the COVID-19 Pandemic

The March 2020 shelter-in-place orders followed by other restrictions shut down restaurants, theaters, schools, and other public spaces, leading to large service sector job losses, social displacement, and isolation. Gun and alcohol sales spiked, as did opioid overdoses.^12,13 Community violence prevention programs, which are dependent on person-to-person contact, also shut down. A national wave of gun violence ensued, with Chicago among the most affected cities.^14,15 Although 1 national study revealed no evidence that the COVID-19 pandemic surge in firearm purchases was related to increased gun violence at the state level,¹⁶ that finding might not hold true specifically for the Chicago area, where household gun ownership has increased dramatically.¹⁷ In 2016, Chicago police recovered illegal guns at 6 times the rate of New York City, including a higher proportion of the most lethal, higher-caliber, larger-magazine weapons.¹⁰

Epidemiological Model of Gun Violence

Social network researchers have used crime data to study how gun violence is transmitted through interpersonal ties in social networks.¹⁸ According to one study, social contagion accounted for more than 63% of Chicago shooting episodes from 2006 to 2014, with numerous shooting “cascades” wherein individuals were shot shortly after a social network peer “infector” had been shot.¹⁹ The social contagion approach is similar to models of the spread of infectious diseases and attempts to apply a public health lens and community outreach methods to mitigate the gun violence epidemic.²⁰ This has led to promising use of professional violence “interrupters,” trusted individuals with gang and prison reentry backgrounds who are capable of reaching young potential offenders and offering opportunities for lifestyle changes.²¹ However, at present only a few hundred intermittently funded violence interrupters are practicing in Chicago.

LIMITATIONS AND CONCLUSIONS

A recent study from Indianapolis, Indiana, that matched individual police gunshot injury victims to electronic health records revealed that 83% of victims were matched to clinical records within 48 hours of a shooting.⁵ Some of the unmatched victims may have refused medical attention; however, most of the discordant cases reflected apparent ICD coding errors. In our study, it is clear that IGW data significantly underrepresented the true incidence of nonfatal gun injuries likely to require hospital care.²² The fact that differences between emergency department visits and CPD data were especially pronounced in 2018 led to inflated hospital data estimates of percentage changes between 2018 and 2020. However, there is no reason to believe that the areas we identified as having the largest increases in IGW emergency department visits were not the same areas experiencing the largest spikes in gun violence.

Despite the undercount of nonfatal gun injuries, the findings presented here regarding the terrible costs of gun violence provide an important perspective on the cost effectiveness of future economic investments in communities with high levels of gun violence. The $342 million in hospitalization charges for IGWs during the study period is only a fraction of the total medical care costs of gun violence, which extend to lifetime care for many individuals crippled by shootings and trauma-related health conditions for entire families and communities.

While police and criminal justice system expenditures soar, dissatisfaction with the “warrior policeman” approach to gun violence has grown, with many urging new approaches to community-based security that can solve what has been described as simultaneous overpolicing and underpolicing.^23–25 For example, expensive Chicago police “shot spotter” audio sensing technology has been criticized for its ineffectiveness in achieving shooting arrests and its effect on police behavior after alerts.²⁶ Increased patrols and frequent illegal gun arrests of citizens with no previous criminal record remain controversial, and murder investigations suffer from lack of resources, hostility related to years of racist policing, and an inability to protect witnesses.²⁷ The nationwide increase in shootings in 2020 has cast doubt about what has been seen as increasingly effective, evidence-based violence prevention.²⁸

Coordinating crime and clinical data on IGW victimization will require greater public health resources with respect to accurate coding to match crime incidents to all related clinical information. Errors likely work both ways, with some gunshot wound medical treatment cases not found in police records despite legal reporting requirements.²⁹ In the context of the national spike in gun violence, the health care system remains a fertile location for violence prevention outreach. Hospital emergency departments can play an important role, for instance, by allowing interrupters to persuade victims’ families and associates to not seek retaliation and providing social services to help victims and their families.^3,30 Longer-term approaches will require large investments in new programs and a community-based, professional public safety workforce.