Objective. To provide national estimates of fatal and nonfatal firearm-related (FA) injuries among children ≤14 years old and to examine the circumstances under which these injuries occurred.
Methods. For nonfatal FA injuries among children, we analyzed data on emergency department (ED) visits from the National Electronic Injury Surveillance System for 1993 through 2000. National estimates of injured children who were treated in hospital EDs were examined by selected characteristics, such as age, gender, race/ethnicity of the patient, primary body part affected, intent of the injury, the relationship of the shooter to the patient, where the injury occurred, and activity at the time of injury. For fatal FA injuries among children, we analyzed mortality data from the National Vital Statistics System for 1993 through 2000. Data from both sources were used to calculate case-fatality rates.
Results. From 1993 through 2000, an estimated 22 661 (95% confidence interval [CI]: 16 668–28 654) or 4.9 per 100 000 (95% CI: 3.6–6.2) children ≤ 14 years old with nonfatal FA injuries were treated in US hospital EDs. Assaults accounted for 41.5% of nonfatal FA injuries, and unintentional injuries accounted for 43.1%. Approximately 4 of 5 children who sustained a nonfatal, unintentional FA injury were reportedly shot by themselves or by a friend, a relative, or another person known to them. During this period, 5542, or 1.20 per 100 000 (95% CI: 1.17, 1.23), children ≤14 years old died from FA injuries; 1 of every 5 children who were wounded by a firearm gunshot died from that injury. Most FA deaths were violence related, with homicides and suicides constituting 54.7% and 21.9% of these deaths, respectively. For individuals ≤14 years old, the burden of morbidity and mortality associated with FA injuries falls disproportionately on boys, blacks, and children 10 to 14 years old. Both fatal and nonfatal injury rates declined >50% during the study period.
Conclusions. Although rates of nonfatal and fatal FA injuries declined during the period of study, FA injuries remain an important public health concern for children. Well-designed evaluation studies are needed to examine the effectiveness of potential interventions aimed at reducing FA injuries among children.
Fatal and nonfatal firearm-related (FA) injuries remain an important public health concern.1–3 In 2000, FA deaths resulted in 992 701 years of potential life lost before the age of 75, or 5.1% of the total for all causes for 2000—more than congenital anomalies, cerebrovascular disease, human immunodeficiency virus disease, chronic respiratory disease, and several other leading causes of death.4 Nonfatal FA injuries also place a substantial burden of morbidity5,6 and economic cost on society.7,8 Patients who survive gunshot wounds may sustain serious spinal cord injury and paraplegia,7 vision loss,9 and lung contusion and other pulmonary trauma.10 Gunshot wounds in the United States in 1994 cost an estimated $2.3 billion in lifetime medical costs or, on average, ∼$17 000 per injury.8
FA injuries among children have long been a concern of the public health community.11 Increased epidemiologic interest and improved surveillance12,13 in recent years have produced numerous population-based studies of the distribution and characteristics of fatal and nonfatal FA injuries among children and adolescents with primary focus on adolescents and younger adults.14–16 However, exposure to injury risk factors and circumstances may differ considerably for children ≤14 years old compared with older teenagers and young adults.17,18 Also, injury prevention strategies to prevent FA injuries for this younger population could be different considering where and how they conduct their daily activities.19 This study focuses on a more detailed description of nationally representative data on fatal and nonfatal FA injuries among children ≤14 years old and characterizes the nature and circumstances of injury incidents, including injury intent, primary body part affected, the identity of the shooter, location of the shooting, case-fatality rates, and outcomes. Intervention strategies are discussed concerning unintentional injury and violence-prevention activities aimed at children ≤14 years old.
Data on nonfatal FA injuries from 1993 through 2000 were from the National Electronic Injury Surveillance System (NEISS), which is operated by the US Consumer Product Safety Commission.20 NEISS is a stratified probability sample of all US hospitals that have an emergency department (ED) and a minimum of 6 beds. During the study period, the number of NEISS hospitals ranged from 90 to 101. Before 1997, the NEISS sampling frame had 4 hospital size strata (small, medium, large, and very large) based on the annual number of ED visits. In 1997, the sampling frame was updated and a stratum for children’s hospitals was added. Our estimates were adjusted to account for changes in the sampling frame over time.
At each NEISS hospital, trained on-site coders reviewed hospital ED records and trauma logs to identify patients who were treated for FA injuries. Coders then captured data on demographic characteristics of the patient, circumstances of the injury incident (eg, primary body part affected, relationship between victim and shooter), and a brief narrative describing the injury event. An FA injury was defined as a nonfatal gunshot wound from a projectile fired from a weapon using a powder charge. BB and pellet gun-related injuries and nonprojectile injuries such as injured trigger fingers or pistol whippings were excluded. Injured individuals who were dead on arrival at the hospital or who died in the ED were excluded from our analysis.
National estimates of nonfatal FA injuries are based on data obtained on 849 FA injury patients who were ≤14 years old and were treated in NEISS hospitals. Of these cases, 17 were recorded as FA suicide attempts and 153 cases were of undetermined intent. Data for 2 legal intervention cases were combined with the 402 assaults and are subsequently referred to as assaults. The remaining 275 cases were unintentional injuries. Each case was assigned a sample weight equal to the inverse of its probability of selection, adjusted for nonresponse and hospital mergers, and ratio adjusted to account for changes in the number of annual ED visits in the sampling frame over time. These weights were summed to calculate national estimates. Details about the NEISS sample design and its representativeness of people who are treated in US hospital EDs and weighting methods are documented elsewhere.21–25
National data on FA deaths from 1993 through 2000 were from the National Center for Health Statistics’ National Vital Statistics System mortality data files. All FA deaths of US residents ≤14 years old with an International Classification of Diseases, Ninth Revision26 underlying cause of death code of E922.0 to E922.9, E955.0 to E955.4, E965.0 to E965.4, E985.0 to E985.4, or E970 for 1993 through 1998 and with an International Statistical Classification of Diseases and Related Health Problems, 10th Revision27 underlying cause of death code of W32 to W34, X72 to X74, X93 to X95, Y22 to Y24, or Y35.0 for 1999 and 2000 were included. Because there were only 13 legal intervention FA (E970 or Y35.0) deaths for 1993 through 2000 in this age range, they were combined with assaults (homicides) for analysis. Case fatality rates (CFRs), expressed as a percentage, were calculated by dividing the number of FA fatalities by the total number of nonfatal and fatal FA injuries and then multiplied by 100. Nonfatal and fatal injury rates per 100 000 people were calculated using population estimates for 1993 through 1999 and population projections for 2000 based on the 1990 census from the US Bureau of the Census.
All data analysis was conducted using SAS software.28 To account for the complex survey design of NEISS, a SAS-callable version of SUDAAN29 was used to calculate variances and confidence intervals (CIs) for national estimates of nonfatal injuries and injury rates.30 SUDAAN′s jackknife method31 was used for variance estimation. For deaths and death rates, standard errors and CIs were computed assuming a Poisson distribution.32 Differences in incidence rates between years were tested for statistical significance using t tests. Because of sample size constraints and to facilitate comparison with previous studies on FA injuries that commonly use standard 5-year age groupings, we grouped cases into 2 age intervals: 0 to 9 years and 10 to 14 years of age.
To analyze the circumstances of unintentional FA injuries, 2 independent researchers reviewed each case. Using the narrative describing the injury event and other coded data in the NEISS record, researchers assigned each case to 1 or more of 18 injury categories using a standardized protocol. These categories represented various activities in which the victim participated at the time of injury (eg, playing with the gun, hunting, gun maintenance). A third researcher reviewed all coding and helped to determine the final category for cases for which researchers disagreed about code assignments. After the coding process, the 18 activity categories were collapsed into 5 mutually exclusive, hierarchic groups (in descending order): injured from a gun used during sports (eg, hunting, sports shooting, target shooting), playing with a gun (eg, using in jest by self or with another child), handling the gun (eg, examining, cleaning, holstering, and loading/unloading the gun), and other (eg, sitting on the gun, falling and accidentally pulling trigger, stray bullets), and unknown.
From 1993 through 2000, an estimated 22 661 (95% CI: 16 668-28 654) children who were ≤14 years old and had nonfatal FA injuries were treated in US hospital EDs. Children 12 to 14 years old accounted for an estimated 66.8% of nonfatal injuries for all children in this age range and 86.3% of nonfatal injuries for those 10 to 14 years old. During this same time period, 5542 (95% CI: 5396–5688) FA deaths occurred among US children ≤14 years old. Of these deaths, 63.7% were children 12 to 14 years old. Among the 10- to 14-year-old age group, 87.7% of deaths were children 12 to 14 years old.
The overall incidence rates for this study population were 4.9 nonfatal FA injuries per 100 000 individuals (95% CI: 3.61–6.20) and 1.20 deaths per 100 000 individuals (95% CI: 1.17–1.23; Table 1). Overall, 1 of every 5 children in this age range who were shot with a firearm died from the injury.
For children ≤14 years old, the distribution of nonfatal and fatal injuries varied by age, gender, and race/ethnicity. Children <10 years old accounted for 23.5% of all fatal and nonfatal injuries combined. Boys were 4 times more likely to experience a nonfatal injury and nearly 3 times as likely to sustain a fatal injury as girls. However, girls who were shot were more likely to die from their wounds (CFR: 25% for girls vs 18% for boys). Although blacks had the highest rates of fatal and nonfatal injuries compared with all other race/ethnicity categories, they were 7% more likely to survive than non-Hispanic whites (CFR: 17.6% for blacks vs 24.6% for non-Hispanic whites). Unintentional injuries accounted for 43.1% of nonfatal FA injuries as well as 20.7% of fatal FA injuries. Although assaults accounted for 41.5% of nonfatal FA injuries, they represented the majority (54.7%) of fatal FA injuries. Suicide attempts accounted for <3% of nonfatal FA injuries but 21.9% of FA deaths.
Between 1993 and 2000, the nonfatal FA injury rate among children <10 years old dropped by 2.12 injuries per 100 000 population (P < .05), representing a decrease of 68% from 3.10 injuries per 100 000 in 1993 to 0.98 in 2000 (Fig 1). A similar decrease in the nonfatal injury rate was observed for children 10 to 14 years old, whose rate dropped by 9.79 injuries per 100 000 (P < .05), or 58% during the same time period. The fatal FA injury rate likewise decreased during the period of analysis. For children <10 years old, the fatal FA injury rate dropped by 51%; this was a decrease of 0.34 deaths per 100 000 (P < .0001) between 1993 and 2000. The decline in the fatal FA injury rate was even larger for children 10 to 14 years old, with the rate dropping 2.24 deaths per 100 000 (P < .0001), or 59%, during the same time period. For nonfatal injuries, annual estimates of rates were too unstable to analyze trends separately by intent of injury. The decline for fatal FA injury rates between the 1993 and 2000 was statistically significant (P < .0001) for each intent category (unintentional, homicide, and suicide).
Across all age groups and intents, lower extremities were the most common sites of nonfatal FA injuries (Table 2). However, 1 of 5 gunshot wounds were to the head or neck region. Among children <10 years old, 28.0% (95% CI: 16.9%–39.2%) of all nonfatal injuries involved trauma to the head or neck, compared with 17.8% (95% CI: 10.8%–24.7%) for children 10 to 14 years old. Almost 52% (95% CI: 30.1%–73.5%) of children <10 years old and 39.6% (95% CI: 23.9%–55.3%) of children 10 to 14 years old were hospitalized. Most children with FA injuries (54.1%; 95% CI: 35.5%–72.8%) arrived at the hospital ED by emergency medical services. For unintentional injuries, the relationship of the offender to the child was stated for 87.7% of cases. Approximately 4 of 5 children who sustained a nonfatal, unintentional FA injury were reportedly shot by themselves or someone known to them (friend, acquaintance, relative). Of the family members who shot a child unintentionally, 79.5% (95% CI: 46.5%–112.6%) were male relatives (eg, brother, father, grandfather, stepfather, uncle). For assault-related injuries, the victim-offender relationship was missing for >40% of cases. Among cases for which a location of injury was specified, most nonfatal injuries occurred in the home.
The majority of unintentional nonfatal FA injuries for which preinjury circumstances were specified occurred while children were handling the firearm (eg, holding, cleaning, loading, and holstering a gun), playing with the firearm (eg, showing the firearm to a friend), or participating in other nonplay and nonsport activities (Fig 2). Injuries that occurred while hunting, sport shooting, or target shooting accounted for 14.1% (95% CI: 7.4%–20.8%) of all nonfatal unintentional injuries.
Our analysis provides national estimates of the incidence and characteristics of fatal and nonfatal FA injuries among children ≤14 years old. For these children, the burden of morbidity and mortality falls disproportionately on boys, blacks, and children 10 to 14 years old. Overall, 1 of every 5 FA injuries among these children resulted in death.
Although fatal FA injuries among children ≤14 years old were predominantly homicides, nonfatal FA injuries were split more evenly between unintentional and violence-related injuries. Averaged over the study period, FA suicides accounted for 152 deaths, or one fifth of all FA deaths for children in this age range each year. However, FA suicides accounted for about one third of all suicides for children ≤14 years old.4
For nonfatal FA injury cases for which the location of the incident could be determined, most injuries occurred at home regardless of intent. Among nonfatal FA injury cases for which the identity of the shooter could be determined, most FA injuries were reported as unintentionally self-inflicted or caused by someone known to the victim. Approximately half of the children who were treated for FA injuries were transported to the ED by emergency medical services, and 40% were hospitalized.
Our analysis demonstrates a significant decline in the incidence of fatal and nonfatal FA injuries among children ≤14 years old from 1993 through 2000. Several possible factors have been offered to explain this decline. Growing prevention efforts aimed at reducing unsupervised access to guns by children and reducing youth violence coincide with the decline in national BB/pellet gun-related injury rates and FA injury and death rates in the United States.33 Part of the decline in FA fatalities may be the result of improved trauma care.34 For assault-related FA injuries, possible explanations for the downward trend include improved economic conditions and the decline of the illicit drug trade over the study period.35 Although other legislative, behavioral, and environmental interventions may have contributed to the decline, there is little direct evidence of such impacts.
Three general approaches—behavior oriented, product design oriented, and legislative strategies—are being taken to help prevent unintentional FA injuries to children. Behavior-oriented strategies include educating parents and caregivers about gun safety practices (eg, safe storage practices, such as separating children from firearms using physical barriers such as lock boxes or gun safes; conducting gun safety programs for children).36 Efforts to promote safe gun storage practices are consistent with the Healthy People 2010 objective of reducing the proportion of people who live in homes with firearms that are loaded and unlocked.37 Our findings support recommendations to encourage parents to reduce their children’s risk of unsupervised access to firearms through storing firearms safely in their home and discussing unsupervised access to firearms and safe storage practices with their relatives and with the parents of their children’s friends.38,39 Our findings also suggest that prevention efforts to reduce unintentional FA injuries among children should be aimed at brothers, fathers, and other male caregivers. Counseling about firearm safety practices and FA injury prevention delivered by health care providers during child and adolescent health care encounters has been recommended.40,41 However, the effectiveness of firearm safety counseling in primary health care settings in regard to safe gun storage is inconclusive.42–44
Product design-oriented approaches include designing firearms with safety devices to protect children from using a firearm without adult supervision. Passive or automatic interventions that have been proposed include magazine safeties that prevent the discharge of rounds that remain in the chamber after the ammunition clip is removed,45 grip safeties that prevent child-sized hands from successfully pulling the trigger,3,46 trigger locks that are permanently built into the gun,46 and personalizing guns with integrated devices that prevent unauthorized firing.47 The effects of these product-oriented approaches to reducing gun violence and unintentional FA injuries in children are unknown.
Child access prevention laws, which hold parents or guardians legally responsible for injuries or deaths resulting from minors using their firearms, have been enacted in 19 states.33 Some studies have suggested that these laws are effective only in states where violations of the law are a felony as opposed to a misdemeanor,48,49 but additional research on this topic is needed.
Strategies for preventing assault-related and suicide-related FA injuries in children have been discussed extensively elsewhere.50–52 Some behavior-oriented and product design-oriented approaches overlap with those aimed at reducing unintentional FA injuries and are discussed above. Other prevention efforts aimed at reducing violence-related FA injuries that need additional evaluation include strategies intended to prevent interpersonal violence among children, such as preventing the expression of violent behavior (eg, social skills training, positive youth development programs); strategies to limit how and where firearms are used, such as eliminating the possession and use of weapons in and around schools; and strategies aimed at youth suicide prevention, such as counseling/treatment for substance abuse and depression, and crisis management.
Our study results are subject to several limitations. First, NEISS collects data only on patients who are treated in hospital EDs; therefore, children who were injured and treated in physicians’ offices or clinics were not included. Second, classification of activities at the time of injury for unintentional FA injury incidents was based on a brief narrative abstracted from the medical record. Approximately 31% of the narratives obtained from medical records had little or no information on the activity. Third, underlying cause and manner of death for data from 1993 through 1998 were classified using the International Classification of Diseases, Ninth Revision, whereas those for 1999 and 2000 were classified using the International Classification of Disease and Related Health Problems, 10th Revision.27 No adjustments were made to ensure comparability across these coding systems because data from the National Center for Health Statistics’ Comparability Study were not yet available. However, preliminary estimates across all ages suggest that the adjustment ratios for FA injuries are small (0.9969 for FA homicides and 1.0579 for unintentional FA injuries).53 Because these ratios are relatively small, our study findings and conclusions should not be notably affected by the change in coding systems. Fourth, sample size limitations resulted in unstable rates in some demographic groups and activity categories. In particular, trend analysis of the decline in annual nonfatal FA injury rates by intent of the injury could not be conducted because of the small number of cases in some intent categories. Fifth, some patients who were treated in NEISS hospital EDs may have died of their injuries at a later point in their treatment at the hospital or other medical facility. However, a previous study suggested that most patients who are hospitalized for FA injuries are discharged from the hospital alive.25 Finally, classification of nonfatal FA injuries by intent of injury is based on information recorded in the medical record as described by the attending nurse or physician or in the emergency medical services report. The intent of injury for nonfatal cases was not confirmed by police reports or other criminal justice records.
Nonfatal and fatal FA injury rates for children ≤14 years old declined significantly from the early 1990s until 2000. These declines are consistent with those among adolescents and adults. However, children remain at risk of sustaining an FA injury. Prevention efforts are needed to reduce further the number of FA injuries associated with unsupervised access to firearms, interpersonal violence, and self-directed violence. Various approaches have been taken to reduce the risk of FA injuries among children, but evidence of their effectiveness is limited. Possible interventions, including their combined effects, need to be evaluated to assess their effectiveness to prevent FA injuries among US children.
We acknowledge the helpful comments by Lynda Doll, PhD, and statistical support of Karen Gotsch, MPH, Patricia Holmgreen, MS, and Kevin Webb of the National Center for Injury Prevention and Control, CDC. We thank Tom Schroeder, MS, Cathy Downs, Art McDonald, MA, and other staff of the Division of Hazard and Injury Data Systems, US Consumer Product Safety Commission for assistance in collecting these data.
- Received April 24, 2003.
- Accepted September 10, 2003.
- Reprint requests to (J.L.A.) Office of Statistics and Programming, National Center for Injury Prevention and Control, Centers for Disease Control and Prevention, 4770 Buford Hwy (MS-K59), Atlanta, GA 30341-3724. E-mail:
- ↵Schneider M-J. Introduction to Public Health. Gaithersburg, MD: Aspen Publishers; 2000:268–271
- ↵Karlson TA, Hargarten SW. Reducing Firearm Injury and Death: A Public Health Sourcebook. New Brunswick, NJ: Rutgers University Press; 1997:xii-xvii
- ↵Centers for Disease Control and Prevention. Web-based injury statistics query and reporting system (WISQARS). Available at: www.cdc.gov/ncipc/wisqars. Accessed August 21, 2002
- ↵Wheatley GM. Childhood accidents. In: American Public Health Association (ed). Accident Prevention: The Role of Physicians and Public Health Workers. New York, NY: McGraw-Hill Book Company; 1961;69–79
- Hootman JM, Annest JL, Mercy JA, Ryan GW, Hargarten SW. National estimates of non-fatal firearm related injuries other than gunshot wounds. Inj Prev.2000;6 :268– 274
- ↵Agran PF, Winn D, Anderson C, Trent R, Walton-Haynes L. Rates of pediatric and adolescent injuries by year of age. Pediatrics.2001;108(3) . Available at: pediatrics.org/cgi/content/full/108/3/e45
- ↵Consumer Product Safety Commission. The National Electronic Injury Surveillance System: A Tool for Researchers. Bethesda, MD: Consumer Product Safety; 2000
- ↵Schroeder T, Ault K. The NEISS Sample (Design and Implementation) From 1979 to 1996. Bethesda, MD: Consumer Product Safety Commission; 2001
- Schroeder T, Ault K. The NEISS Sample (Design and Implementation): 1997 to Present. Bethesda, MD: Consumer Product Safety Commission; 2001
- ↵World Health Organization. Manual of the International Statistical Classification of Diseases and Injuries, and Causes of Death. 9th rev. Geneva, Switzerland: World Health Organization; 1977
- ↵World Health Organization. Manual of the International Statistical Classification of Diseases and Injuries, and Causes of Death. 10th rev. Geneva, Switzerland: World Health Organization; 1992
- ↵SAS Institute. SAS. Version 8.2. Cary, NC: SAS Institute; 2001
- ↵Research Triangle Institute. SUDAAN: Software for the Statistical Analysis of Correlated Data. Version 8.0.0. Research Triangle Park, NC: Research Triangle Institute; 2001
- ↵Marker D, Lo A, Brick M, Davis W. Comparisons of National Estimates From Different Samples and Different Sampling Frames of the National Electronic Injury Surveillance System (NEISS): Final Report. Rockville, MD: Westat; 1999
- ↵Research Triangle Institute. SUDAAN User Manual Release 8.0. Research Triangle Park, NC: Research Triangle Institute; 2001:85–88
- ↵Nguyen MH, Annest JL, Mercy JA, Ryan GW, Fingerhut LA. Trends in BB/pellet gun injuries in children and teenagers in the United States, 1985–99. Inj Prev.2002;8 :185– 191
- ↵Harris AR, Thomas SH, Fisher GA, Hirsch DJ. Murder and medicine: the lethality of criminal assault 1960–1999. Homicide Stud.2002;6 :128– 66
- ↵Department of Health and Human Services. Healthy People 2010. Washington, DC: US Government Printing Office; 2000:15-15-15-16
- ↵American Academy of Pediatrics. Firearm-related injuries affecting the pediatric population. Pediatrics.2000;105 :888– 895
- ↵Grossman DC, Cummings P, Koepsell TD, et al. Firearm safety counseling in primary care pediatrics: a randomized, controlled trial. Pediatrics.2000;106 :22– 26
- Albright TL, Burge SK. Improving firearm storage habits: impact of brief office counseling by family physicians. J Am Board Fam Pract.2003;16 :40– 46
- ↵Stevens MM, Olson AL, Gaffney CA, Tosteson TD, Mott LA, Starr P. A pediatric, practice-based, randomized trial of drinking and smoking prevention and bicycle helmet, gun, and seatbelt safety promotion. Pediatrics.2002;109 :490– 497
- ↵Webster DW, Starnes M. Reexamining the association between child access prevention gun laws and unintentional shooting deaths of children. Pediatrics.2000;106 :1466– 1469
- ↵Mercy J, Rosenberg M. Firearms violence in and around schools. In: Hamburg B, Elliott D, Williams K (eds). Violence in American Schools. New York, NY: Cambridge University Press; 1998:159–187
- US Department of Health and Human Services. Surgeon General’s Call for Action to Prevent Suicide. Washington, DC: US Department of Health and Human Services; 1999
- ↵US Department of Health and Human Services. Youth Violence: A Report of the Surgeon General. Rockville, MD: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Injury Prevention and Control; Substance Abuse and Mental Health Services Administration, Center for Mental Health Services, and National Institutes of Health, National Institute of Mental Health; 2001
- Copyright © 2004 by the American Academy of Pediatrics