Understanding the contours of these injuries—where, to whom, and how—is essential for efforts to develop solutions. The Centers for Disease Control and Prevention (CDC), the federal agency responsible for collecting data to protect the US from dangerous health threats, has faced challenges with the expansion of its injury data collection, an essential requirement for obtaining more reliable numbers on this under-recognized aspect of gun violence. The CDC’s estimate for 2015 totaled 84,997 nonfatal firearm injuries, but this estimate was qualified by a confidence interval of 36,636 to 133,357 injuries, too large a range to estimate the true burden of gun injury with accuracy or to know whether year-to-year increases or decreases indicate true changes or rather, are an artifice of data limitations. In the face of increasing scrutiny of the reliability of this data, the CDC recently removed 2016 and 2017 nonfatal injury data for firearms and all other injury types from its website. Everytown for Gun Safety strongly supports equipping the CDC with the funds necessary to thoroughly measure and examine gun violence.
In the absence of a national database tracking shootings, or reliable nonfatal injury estimates, Everytown has calculated the total number of nonfatal gun injuries and conducted original analysis using the federal Nationwide Emergency Department Sample (NEDS) data set. The NEDS is the largest publicly available emergency department database in the US, a part of the Healthcare Cost and Utilization Project (HCUP). This data set includes data from approximately 950 hospitals, 16 times the number included in the CDC survey, and provides the most reliable nonfatal injury estimates currently available.
Using the HCUP data set, which contains roughly 30 million hospital discharge records each year, we estimate the total number of nonfatal firearm injuries and injuries by demographic group. Scroll down to read more about our methodology.
For every one person shot and killed by a gun, two more are wounded.
Approximately 73,330 people are shot and injured by firearms every year in the US, an average of 200 gun injuries sustained every day.1 A yearly average was developed using three years of data: 2013, 2014, and 2016. The three-year sum of gun injuries is 219,989 (95% CI, 196,810 to 243,168). This means that for every gun death, there are two more gun injuries.2There are an average 36,383 gun deaths a year. A yearly average was developed using five years of most recent available data: 2013 to 2017. Centers for Disease Control and Prevention. National Center for Injury Prevention and Control, Web-based Injury Statistics Query and Reporting System (WISQARS) Fatal Injury Reports.
The lower the household income in a community, the higher the chance of gun injury.
Despite accounting for a quarter of the US population, individuals living in zip codes with a median household income at the bottom fourth of the income scale—$40,333 or less—account for over half (52.9%) of nonfatal firearm injuries.4$40,333 represents a three year average of the uppermost bounds of the lowest median household income quartile: 2013, 2014, and 2016. The 25th percentile of income scale was ≤$38k in 2013, ≤$40k in 2014, and ≤$43k in 2016. Across all years, the rate of injuries among those living in these zip codes is over seven times higher than among those living in zip codes with median household incomes at the top fourth of the income scale—$67,000 or more.5$67,000 represents a three year average of the lowermost bounds of the highest median household income quartile: 2013, 2014, 2016. The 75th percentile of the income scale was ≥$64k in 2013, ≥$66k in 2014, and ≥$71k in 2016. Rates of injuries per 100,000 people range from 47.7 in the bottom fourth of the income distribution to 6.5 injuries per 100,000 people in the top quartile. Gun injuries decrease as median household income by zip code increases. In short, for those living in low-income areas, the likelihood of sustaining a gun injury is many orders of magnitude higher than for those living in higher-income neighborhoods.
Men and young adults are disproportionately at risk of gun injury.
Men are nearly eight times more likely to be wounded by guns than are women (41.1 and 5.3 injuries per 100,000 people, respectively). When it comes to age, the nonfatal gun injury rate increases as age increases, peaks among 20-to-24-year-olds (76.7 injuries per 100,000 people), and decreases thereafter.
One in six gun injuries involves a child or teenager.
Approximately 12,706 nonfatal gun injuries are sustained by America’s children and teens every year. Children and teens account for 17.3% of nonfatal firearm injuries, compared to 7.7% of firearm deaths. This means that while roughly one in 13 firearm deaths occurs among children and teens, one in six gun injuries is among people under the age of 20. Approximately every 40 minutes, a child or teenager is wounded with a gun.
Gun injury rates are similar across rural and urban areas.
Nonfatal gun injury rates are relatively stable across levels of urbanization. The most urban counties (large central metro) have a nonfatal gun injury rate of 28.3 injuries per 100,000 people, whereas the rate in the most urban counties (noncore) is 25.7 injuries per 100,000 people. The only county type with a markedly lower nonfatal gun injury rate than the rest were city suburbs (large fringe metro) with a rate of 15.6 injuries per 100,000 people.
Half of all gun injuries occur in the South.
While gun injury rates vary little between US rural and urban counties, there are considerable differences by region. Half of all injuries took place in one region, the American South. In fact, the South has the highest rate of nonfatal firearm injuries, with 30.5 injuries per 100,000 people—about 2.5 times that of the lowest region, the Northeast. These findings are consistent with other research finding higher rates of gun injury in the South.6Shilpa J. Patel, Gia M. Badolato, Kavita Parikh, Sabah F. Iqbal, and Monika K. Goyal, “Regional Differences in Pediatric FirearmRelated Emergency Department Visits and the Association with Firearm Legislation,” Pediatric Emergency Care (February 2019), doi: 10.1097/PEC.0000000000001779. The second highest rate is in the Midwest, with 24.2 injuries per 100,000 people. The West is third overall, with a rate of 17.4 injuries per 100,000 people. The Northeast has the lowest rate of gun injuries, with a rate of 12.7 injuries per 100,000 people.
“In order to effectively study anything, researchers need accurate, comprehensive data. That’s true with infectious diseases. It’s true with environmental threats, and with drug addiction. And it’s true with gun violence.”– Dr. Daniel Webster, Johns Hopkins Bloomberg School of Public Health
Unlike gun deaths, of which a full census of national, state, and county data is available from the death certificate-based National Vital Statistics System, standardized nonfatal gun injury data at the state or local level are limited. Rather, researchers and the public rely on estimates from a sample of hospital emergency departments.
The limitations of CDC firearm injury data stem in part from the small sample size of such emergency departments. Currently, the agency’s survey includes only about 100 hospitals, of which 66—less than 2% of all hospitals in the US—submit data related to gunshot wounds.7Sean Campbell, Daniel Nass, and Mai Nguyen, “The CDC Is Publishing Unreliable Data on Gun Injuries. People Are Using It Anyway,” FiveThirtyEight, October 4, 2018, https://53eig.ht/2Paa485. But missing a key trauma hospital within a city with high rates of gun assault can skew the survey, because gun assaults tend to be concentrated in very small areas within specific city neighborhoods. Further, as hospitals transition out of the data set, the replacement hospital may treat a very different mix of injuries, which can further disrupt the estimates.8Sean Campbell and Daniel Nass, “How One Hospital Dramatically Skewed the CDC’s Estimate of Nonfatal Gun Injuries,” The Trace, August 13, 2019, https://bit.ly/2OVPvAC.
Researchers and many others have called for federal and state government agencies to provide funding to collect gun violence data and conduct gun violence research necessary for policy-making that can save lives. This type of data collection is costly. The current CDC nonfatal injuries data come from collaboration with an existing surveillance system operated by the Consumer Product Safety Commission based on a small sample of hospitals. CDC officials are exploring an expansion of the roster of reporting hospitals but have made it clear that this would require additional funding.9Robert Redfield, official correspondence to US Department of Health and Human Services, “Centers for Disease Control and Prevention’s Responses to Questions About Firearm Injury Data,” May 3, 2019, https://bit.ly/2km0WDb.
A Clearer Picture
The tens of thousands of Americans injured by firearms each year face many difficulties, including, but not limited to, physical disability, psychological trauma, increased healthcare costs, and the risk of additional injury in cycles of violence.10Carla DiScala and Robert Sege, “Outcomes in Children and Young Adults Who Are Hospitalized for Firearms-Related Injuries,” Pediatrics 113, no. 5 (May 2004): 1306-1312; Carla DiScala and Robert Sege, “Outcomes in Children and Young Adults Who Are Hospitalized for Firearms-Related Injuries,” Pediatrics 113, no. 5 (May 2004): 1306-1312; James Garbarino, Catherine P. Bradshaw, and Joseph A. Vorras, “Mitigating the Effects of Gun Violence on Children and Youth,” The Future of Children 12, no. 2 (Summer-Autumn 2002): 73-85. Angela Scarpa, Jimmy D. Hurley, Howard W. Shumate, Sara Chiara Haden, “Lifetime Prevalence and Socioemotional Effects of Hearing about Community Violence,” Journal of Interpersonal Violence 21, no. 1 (January 1, 2006): 5-23. Nikeea Copeland-Linder, Sara B. Johnson, Denise L. Haynie, Shang-en Chung, and Tina L. Cheng, “Retaliatory Attitudes and Violent Behaviors among Assault-Injured Youth,” Journal of Adolescent Health 50, no. 3 (March 2012): 215-20. Catherine Juillard, Laya Cooperman, Isabel Allen, Romain Pirracchio, Terrell Henderson, Ruben Marquez, Julia Orellana, Michael Texada, and Rochelle Ami Dicker, “A Decade of Hospital-Based Violence Intervention,” Journal of Trauma and Acute Care Surgery 81, no. 6 (December 2016): 1156-1161; Corinne Peek-Asa, Brandon Butcher, and Joseph E. Cavanaugh, “Cost of Hospitalization for Firearm Injuries by Firearm Type, Intent, and Payer in the United States,” Injury Epidemiology 4 (December 2017): 20; Carnell Cooper, Dawn M. Eslinger, and Paul D. Stolley, “Hospital-Based Violence Intervention Programs Work,” The Journal of Trauma and Acute Care Surgery 61, no. 3 (September 2006): 534-537.
While the data in this paper are collected when a patient is discharged from the emergency department or inpatient care, the impact of a gun injury does not end there. Every gun death is a tragic loss of life. But gun injuries represent loss too: the loss of a sense of safety, of mental wellbeing, of physical ability, a loss of income for those whose injury affects their work, and a loss of innocence—particularly for the thousands of American children who are injured with guns.
As the crisis of gun violence in this country continues, so must the CDC’s investment in nonfatal injury data. While Everytown has endeavored to fill this gap by providing this analysis of nonfatal firearm injury, ultimately the federal government has an important role to play. The CDC has historically conducted rigorous research on many causes of injuries and diseases in fulfillment of its mission to “protect our nation against expensive and dangerous health threats.”
Make no mistake—firearm injury is one such threat.
Estimated number, rate, and three-year average of nonfatal, hospital treated firearm injuries: US 2013, 2014, and 2016.
The analysis in A More Complete Picture: The Contours of Gun Injury in the United States provides estimates of nonfatal gun injuries in the US by demographics categories.
Everytown uses data for this analysis from the Nationwide Emergency Department Sample (NEDS), the largest publicly available all-payer emergency department (ED) database in the US, consisting of approximately 31 million discharge records each year. The NEDS is a part of the Healthcare Cost and Utilization Project (HCUP) of the US Department of Health & Human Services’ Agency for Healthcare Research and Quality.
Described fully elsewhere,12Agency for Healthcare Research and Quality, “Introduction to the HCUP Nationwide Emergency Department Sample (NEDS) 2016,” September 2018, https://bit.ly/2BrJ26O. the NEDS represents a 20 percent–stratified sample of US hospital-owned EDs and captures encounter-level data on all visits that do not result in an admission (i.e., treat-and-release visits and transfers to another hospital) as well as discharge-level data on patients seen in the ED and then admitted to the same hospital. Federal government hospitals such as Veterans Affairs (VA) hospitals, military hospitals, and Indian Health Service facilities are not included. Each year, approximately 950 hospitals located in 36 states and the District of Columbia contribute to the NEDS.
Discharge-level weights provided in the NEDS allow a nationally representative estimation of the number and rate of ED visits.
Each discharge record in the NEDS includes variables such as:
- Patient demographics (e.g., gender, age, urban-rural designation of residence, median household income of residential zip code).13Gender of patient coding is gender: (0) male, (1) female; age in years coded 0-90 years; any ages greater than 90 are set to 90; urban-rural designation for patient’s county of residence: (1) large central metropolitan, (2) large fringe metropolitan, (3) medium metropolitan, (4) small metropolitan, (5) micropolitan, and (6) not metropolitan or micropolitan; median household income quartiles for patient’s zip code in 2013 are defined as: 1) $1–$37,999 2) $38,000–$47,999 3) $48,000–$63,999, and 4) $64,000 or more; median income household income quartiles for patient’s zip code in 2014 are defined as: 1) $1–$39,999 2) $40,000–$50,999 3) $51,000–$65,999, and 4) $66,000 or more; median income household income quartiles for patient’s zip code in 2016 are defined as: 1) $1–$42,999, (2) $43,000–$53,999, (3) $54,000–$70,999, and (4) $71,000 or more. The NEDS makes patient race data available only at the state level and therefore that data is not used in this analysis;
- Patient disposition (e.g., died in ED, transferred to another short-term hospital, admitted to the same hospital);
- Hospital characteristics (e.g., US Census region) and;
- Nature of the visit (e.g., International Classification of Disease Ninth Revision (ICD-9-CM)/Tenth Revision (ICD-10-CM) diagnoses and external cause of injury codes [E-codes]).14ICD-10-CM E-codes can be found at https://bit.ly/2o8OgkY.
Everytown abstracts hospital discharge records for patients who arrived at EDs with firearm-related injuries at hospitals across the US from the NEDS data from 2013, 2014, and 2016. The NEDS data from 2015 are excluded because of a switch from the International Classification of Disease Ninth Revision (ICD-9-CM), used since the 1970s, to a more complex and detailed coding system, ICD-10-CM/PCS.15ML Barrett, “2015 Healthcare Cost and Utilization Project (HCUP) National Inpatient Sample: Change in Structure and Data Elements Caused by Transition to ICD-10-CM/PCS,” US Agency for Healthcare Research and Quality, November 13, 2017, https://bit.ly/2pDVvSp. Records are included if at least one of the four E-code fields, (included in all HCUP records) contains either ICD-9-CM E-codes16For NEDS data from 2013 and 2014: E922.0-E922.3, E922.8, E922.9, E955.0-E955.4, E955.9, E965.0-E965.4, E979.4, E985.0-E985.4. or ICD-10-CM E-codes.17For NEDS data from 2016: W32, W33, W34.00, W34.09, W34.10, W34.19, X72, X73, X74.8, X74.9, X93, X94, X95.8, X95.9, Y22, Y23, Y24.8, Y24.9, Y35.001-Y35.023, Y35.41-Y35.093, Y38.4. See Table for detailed information regarding ICD-9-CM and ICD-10 CM E-codes that are used to identify firearm-related injuries. For the NEDS 2016, only ICD-10-CM “initial encounter” codes are used (i.e., the seventh character of the E-code is “A”), to eliminate double-counting of injuries. Firearm-related injuries that resulted in death during the ED visit or during subsequent hospitalization are excluded.
The distribution of gun injury intent is difficult to determine from medical record documentation alone. ICD diagnoses and E-codes limit intent categories related to gun injuries to five categories: assault, unintentional, self-harm, undetermined, and shootings by law enforcement. In cases where the intent of a shooting injury is unclear, and in the absence of affirmative documentation on the incident, medical coders will often mark the cause of injury as “unintentional.” As with other studies that solely used ICD coding to classify injury intent, the findings of intent distribution may not accurately reflect the true distribution of nonfatal firearm injuries and therefore are not included in this analysis.18Kirsten McKenzie, Lois Fingerhut, Sue Walker, Adam Harrison, James E. Harrison, “Classifying External Causes of Injury: History, Current Approaches, and Future Directions,” Epidemiologic Reviews 34, no. 1 (January 2012): 4–16, https://doi.org/10.1093/epirev/mxr014.
Everytown estimates the number and percent of nonfatal firearm injuries overall and by patient demographics (see above).
Everytown calculates rates of nonfatal firearm injuries per 100,000 people and three-year annual averages of nonfatal firearm injuries overall and by demographics using total and demographic-specific population figures obtained from the CDC’s Wide-ranging Online Data for Epidemiologic Research (WONDER) website.19Centers for Disease Control and Prevention, National Center for Health Statistics, “Underlying Cause of Death, 1999-2017,” on CDC WONDER online database, http://wonder.cdc.gov/ucd-icd10.html. Population data for quartiles of median household income come from Claritas.20Marguerite Barrett, Rosanna M. Coffey, and Katharine Levit, “Population Denominator Data Sources and Data for Use with the HCUP Databases (Updated with 2017 Population Data),” US Agency for Healthcare Research and Quality, HCUP Methods Series Report # 2018-03 (October 4, 2018), https://hcup-us.ahrq.gov/reports/methods/2018-03.pdf. The 2013 NCHS Urban-Rural Classification for Counties is used to classify urbanization.21Deborah D. Ingram and Sheila J. Franco, “2013 NCHS Urban–Rural Classification Scheme for Counties,” National Center for Health Statistics, Vital and Health Statistics 2, no. 166 (April 2014).
Everytown conducts all descriptive analyses using the Stata IC 14 SVY suite (StataCorp LLC, College Station, Texas) and weighted commands (using the weight variables provided by HCUP) to generate national estimates of nonfatal firearm injuries.
The NEDS, like every dataset and surveillance system, has advantages and shortcomings. One important challenge with the NEDS is related to the International Classification of Diseases (ICD) system used for coding of incidents. The ICD system is used by doctors and healthcare providers in hospitals across the country. In cases where the intent of a shooting injury is unclear, and in the absence of affirmative documentation on the incident, medical coders will often mark the cause of injury as “unintentional.” Therefore, many potential assault injuries may instead be coded as unintentional. The other challenge is that while the current CDC nonfatal injuries data exists through a user-friendly publicly-available interface called Web-based Injury Statistics Query and Reporting System (WISQARS), analysis of the NEDS requires specialized data analysis software and skills.
Everytown Research & Policy is a program of Everytown for Gun Safety Support Fund, an independent, non-partisan organization dedicated to understanding and reducing gun violence. Everytown Research & Policy works to do so by conducting methodologically rigorous research, supporting evidence-based policies, and communicating this knowledge to the American public.