Objectives. To compare outcomes by intent of nonfatal firearms-related injuries in a hospitalized population, newborn to 19 years of age, and estimate the national incidence of ensuing disability.
Methods. Descriptive statistics and comparative analysis using χ2, odds ratio, and t test were applied to data from the National Pediatric Trauma Registry (NPTR) and the National Electronic Injury Surveillance System of the US Consumer Product Safety Commission. Demographics, preinjury medical history, scene of injury, primary body part injured, severity of injury, utilization of resources, short-term and long-term disability, medical cause of disability, and disposition at discharge were studied.
Results. NPTR unintentional (n = 268) and assault-related firearms-related injuries (n = 506) were compared. In both groups, the majority of patients were male (80%). Compared with the unintentionally injured, the assaulted children were older and more frequently black (59.3% vs 32.5%). Approximately 17% in both groups had a preinjury history of medical/psychosocial problems. Unintentional injuries occurred mainly in private dwellings (75.7%), and assaults occurred in public places/street (53.8%). In both groups, injuries to multiple body regions were prevalent, and a substantial proportion sustained injuries of serious to critical level. Most children were transported by ambulance, but a significant proportion in the unintentional group were transported by helicopter. The rate of admission to the intensive care unit was ∼40% for both groups. The unintentionally injured had a higher rate of surgical intervention (66.8% vs 50.8%) and stayed in the hospital longer than the assaulted ones (median: 5 days vs 3 days). Almost half of the children in both groups were discharged with disability, and ∼87% returned to their home. Applying the NPTR disability rate to National Electronic Injury Surveillance System estimates of hospitalization suggests that ∼3200 children nationwide develop disability from firearms-related injuries annually.
Conclusions. Nonfatal firearms-related injuries in a pediatric population are associated with a high use of medical resources and lasting disability. Public policies should be developed and implemented to reduce the occurrence of these catastrophic events.
Analysis of national trends indicates that, for all ages, nonfatal and fatal firearms-related injuries are declining in the United States,1,2 although the rate of firearms-related deaths remains high. In fact, for all ages, firearms-related injuries remained the second leading cause of deaths in the United States in 1998.2
Among children and young adults newborn to 19 years of age, the Surveillance for Fatal and Nonfatal Firearm-Related Injuries 2 reported that between 1993 and 1998 in the United States, an average of 4922 deaths and 22 000 nonfatal injuries occurred annually. Of the nonfatal injuries, an average of 10 258, or almost half, led to hospitalization each year.
As the number of fatal and nonfatal firearms-related injuries are declining, so are the number of hospitalizations. More recent data from the National Electronic Injury Surveillance System (NEISS) indicated that for the calendar year 2000, the national estimate of the firearms-related hospitalizations, inclusive of transfers to another specialized medical care facility and hospitalized, was 7060 (J.L. Annest, Centers for Disease Control and Prevention [CDC], personal communication, unpublished data from the CDC Firearm Injury Surveillance Study).
Although many nonfatal injuries are minor, others have more far-reaching consequences and may result in life-long limitations. Although there are limited reports concerning the physical and psychological long-term outcomes from firearms-related injury in the adult population,3–6 no information exists to date about long-term sequelae from such injuries in the pediatric population. To overcome this gap, this study sought to 1) evaluate outcomes by intent from pediatric nonfatal firearms-related injuries that required hospitalization and 2) provide estimates of national incidence of ensuing disability. The data sources for this study are the National Pediatric Trauma Registry (NPTR) and the NEISS of the US Consumer Product Safety Commission.
The NPTR is a database that contains information about many aspects of pediatric trauma and its causes and consequences. Data were contributed voluntarily to the NPTR by pediatric trauma centers or by children’s hospitals with pediatric trauma units.
The NPTR collected data on children and adolescents who were 19 years of age and younger and were admitted to the hospital for an acute injury, including patients who were dead on arrival or died in the emergency department (ED). All types of injuries were included, except burns, poisoning, and near drowning. The operation, data management, and quality assurance of the NPTR have been described in detail elsewhere.7 For ensuring accuracy and consistency across centers, all coding for injury diagnoses, injury severity, and external causes were performed centrally by the NPTR staff.
The injuries recorded in the NPTR represent a subset of all of the injuries that occur to children because the NPTR includes only injuries that lead to hospitalization. Also, because the NPTR’s participating institutions are hospitals that specialize in the treatment of pediatric trauma, the proportion of severe cases reported to the NPTR is higher than that observed in all types of hospitals. The NPTR is a voluntary database, not a population-based sample; thus, population-wide incidence estimates cannot be inferred directly from it.
The year 2000 NEISS comprises 99 hospitals that are a stratified probability sample of all hospitals in the United States and its territories that have at least 6 beds and provide 24-hour emergency care.2 Through an agreement between the Consumer Product Safety Commission and the CDC, information about all patients who had gun-related injuries and were treated in NEISS hospital EDs has been collected since June 1992. Nonfatal firearms-related gunshot cases extracted from NEISS included only patients who were alive when discharged from the ED. The recorded disposition at ED discharge indicates whether the patient was admitted to the hospital. Each eligible nonfatal case was assigned a sample weight. Sample weights were summed to provide national estimates of nonfatal firearms-related injuries.
A firearms-related injury was defined as a gunshot wound or penetrating injury from a weapon that uses powder charge to fire a projectile. This definition includes gunshot injuries sustained from handguns, rifles, and shotguns but excludes gunshot wounds from air-powered BB and pellet guns.2
Accordingly, we extracted from the NPTR (October 1995 to February 2002, N = 50 199) all cases with an external cause of injury identified by the following E-codes: 1) 922.0, 922.1, 922.2, 922.3, 922.8, and 922.9 (unintentional); 2) 955.0, 955.1, 955.2, 955.3, and 955.4 (intentionally self-inflicted); 3) 965.0, 965.1, 965.2, 965.3, and 965.4 (assault); 4) 985.0, 985.1, 985.2, 985.3, and 985.4 (undetermined intent); and 5) 970 (legal intervention).
The first E-code in the record was used for the extraction process. This extraction process resulted in a subset of 913 children, newborn to 19 years of age, inclusive of 107 (11.7%) children who died from the injuries after hospital presentation. After the deaths were excluded, a total of 806 children remained. Children who were hospitalized for intentionally self-inflicted gunshot wounds (n = 12) and those for whom the intent was undetermined (n = 20) were excluded from this analysis. No injuries as a result of legal intervention were recorded. This selection yielded a total of 774 nonfatal injuries, 268 of which were caused unintentionally and 506 of which were caused by assaults.
The study variables included demographics, preinjury medical history, scene of injury, severity of injury, injured body regions, utilization of resources, disability, medical cause of disability, and disposition at discharge from acute care.
Utilization of resources included method of transport to the hospital, surgical intervention, admission to intensive care unit, and length of stay in the facility. Severity of injury was measured by the Injury Severity Score (ISS).8 Injured body regions were defined as those associated with the computation of the ISS: head and neck, face, thorax, abdomen, extremities, and skin. The additional category of “multiple body regions” indicates injuries to >1 of the above regions.
Disability and its duration were derived from the functional status at discharge. In the NPTR, functional status is assessed at the time of discharge by rating the child’s ability in 10 functional domains: vision, hearing, speech, self-feeding, bathing, dressing, walking, cognition, behavior, and bladder/bowel control. Using performance and neurologic tests, a clinician rates the functioning of the child in each area as either age appropriate, impaired, or unable. Moreover, for each functional limitation, the clinician assigns an expected duration (<7 months, 7–24 months, >24 months) and indicates whether it was existing before the current injury event and, if so, whether it was worsened by the current injury.9–13
Disability as a result of the current injury was defined as any degree of limitation (impaired or unable) in 1 or more functional domains that was not preexisting. Disability was considered short term when all of the functional limitations were expected to last <7 months and long term when at least 1 functional limitation was expected to last >7 months. Medical cause of disability was defined by the International Classification of Diseases Ninth Clinical Modification underlying injury diagnosis codes.
Race/ethnicity was collapsed from the original 7 categories recorded in the NPTR to white, black, Hispanic, and other by classifying white Hispanic as Hispanic, and black Hispanic as black.2 The category “other” includes Asian and American Indian.
Type of weapon was not reported because it was recorded as unknown in 66% of the unintentional injury group and in 92.1% of the assault injury subset.
Data analyses were performed using BMDP statistical software.14 χ2, odds ratio (OR), and t test were applied.
Unintentional Versus Assault-Related Injuries
Most of the injured children (∼80%) were boys in both groups (Table 1). The children who were assaulted were significantly older, with a median age of 15 years compared with a median of 12 years in the unintentional injury group. In fact, more than half (53.6%) of the older teenagers had been assaulted (Table 1). The 2 groups were also racially/ethnically different. The unintentionally injured children were predominantly white (50.0%), whereas the assaulted children were predominantly black (59.3%; Table 1).
No difference was detected between the 2 groups in terms of preinjury medical history: >75% of the children were healthy before the injury occurred (Table 1). A very small proportion (3.4% in the unintentional group and 1.6% in the assault group) had some previous disability: physical or psychosocial or both.
The injury event occurred in different places according to the intent of the injury. Unintentional injuries were more likely to occur at home than the assaults (75.7% vs 27.1%; OR: 8.41; P < .001; Fig 1). Conversely, assaults were more likely to take place in the street or in public places (53.8% vs 7.5%; OR: 14.41; P < .001; Fig 1). Almost none of the injuries, whatever the intent, took place at school.
On examination of the injured body regions (Table 2), it emerged that unintentionally injured children were more likely than the assaulted children to sustain wounds to the deeper structures of the abdomen (14.2% vs 7.9%; OR: 1.92; P < .05), whereas assaulted children were more likely to sustain wounds to the superficial skin (35.0% vs 23.1%; OR: 1.79; P < .001). Overall, however, children in both groups were more likely to sustain injuries to multiple body regions, with the extremities, either alone or with other serious concomitant injuries, being wounded in ∼30% of the children in both groups.
The severity of the injuries as measured by the ISS was similarly distributed in both groups (Table 2). The majority of the children (90%) sustained injuries of minor to moderate severity (ISS: 1–19), but a substantial proportion (7.1% and 8.9%, respectively) sustained injuries of serious to critical level (ISS: 20–75).
Assessment of the resources consumed by the 2 groups indicates interesting differences (Table 3). Unintentionally injured children were more likely than the assaulted ones to be transported to the treating hospital by helicopter (27.6% vs 8,3%; OR: 4.21; P < .001). Conversely, assaulted children were more likely to be taken to the hospital by ambulance (81.2% vs 60.8%; OR: 2.79; P < .001). The rate of admission to the intensive care unit (>40%) was similar between the 2 groups, but unintentionally wounded children were more likely to require surgical intervention (66.8% vs 50.8%). In addition, unintentionally injured children stayed in the hospital a significantly longer time than the assaulted ones (mean: 8.4 days; median 5 days vs mean 6.2 days; median: 3 days; t test: P < .001).
Almost half of the children in both groups were discharged from the hospital with some disability (Table 4). The rate of short-term disability as estimated at discharge was similar between the 2 groups (∼35%). The rate of long-term disability was higher but not significantly so among the unintentionally injured children than among the assaulted ones (8.2% vs 5.5%). The cause of short-term disability was similar between the 2 groups and mainly attributable to wounds to the extremities; the cause of long-term disability for both groups was mainly attributable to injury to the central nervous system (Table 4). Specifically, among the children with expected long-term disability, 41% in the unintentional injury group sustained injury to the head, and 50% in the assault group sustained injury to the spinal cord. The disposition at discharge from the hospital was similar between the 2 groups, with ∼87% of the children returning to their home.
Estimates of National Incidence of Disability
We then applied the rate of short-term, long-term, and unknown duration of disability as established by the NPTR to the number of hospitalizations (7060) established by NEISS for children 0 to 19 years of age during the 2000 calendar year. The rates applied were those resulting from the NPTR unintentional and assaultive injury combined. Although the number of hospitalizations reported by NEISS includes those resulting from additional causes, such as intentionally self-inflicted, legal intervention, and undetermined intent, that were excluded from the analysis of the NPTR data, it is unlikely that the trivial number of these types of injuries recorded in the NPTR would have made a substantial difference in determining disability rates.
Thus, applying a rate of 35.1% for short-term disability, 6.5% for long-term disability, and 3.9% for disability of unknown duration, the nationwide number of children who are admitted to hospitals annually for nonfatal firearms-related injuries would be ∼2478 with short-term disability, 459 with long-term disability, and 275 with disabilities of unknown duration. Altogether, an estimated 3212 children would be discharged annually with disabilities from firearms-related injuries, a number slightly higher than the 3042 deaths reported by the Web-based Injury Statistics Query & Reporting System for the year 2000.15
Injuries caused by firearms continue to plague American youths. Although injuries and fatalities related to firearms have been steadily declining since 1993,2 recent estimates indicate that firearms-related deaths are second only to motor vehicle crashes as a cause of death for adolescents and young adults,16,17 yet firearms-related fatalities represent only one aspect of this public health problem. In fact, an evaluation of the magnitude of nonfatal firearms-related injuries treated in hospital EDs indicated that the rate of nonfatal injuries was 2.6 times the national rate of fatal injuries for 1992.17
The burden of nonfatal firearms-related injuries does not stop at hospitalization. Many hospitalized survivors of gunshot injuries5,6 report significant long-term declines in physical and/or mental heath. Furthermore, psychological sequelae of firearms-related injuries extend beyond the patient him- or herself, and the incidence of posttraumatic stress disorder,4–6 for instance, is extraordinarily high in many communities.18,19
Although firearms-related injuries take a disproportionate toll on children and young adults, no information exists about their immediate and long-term outcomes in the pediatric population. Using data collected at discharge from hospitalization, this study examines the physical and neurologic consequences of nonfatal firearms-related injuries by type of intent in children newborn to 19 years of age. It also provides crude estimates of the expected disability nationwide.
As in previous studies,2 the majority (4 of 5) of the children in both unintentional and assault groups were male. As expected also, girls were more likely to be injured in an assault than unintentionally.20 Children with unintentional firearms-related injuries were younger and more likely to be white than patients with assault-related firearms injuries. There was no difference between the 2 groups in terms of preinjury medical history such as physical or psychosocial problems.
Unintentional injuries were more likely to occur in private dwellings and require air-lifting than the assaults, suggesting a less urban population than the assault group. In contrast, assault-related firearms injuries disproportionately affect young male urban blacks as reported elsewhere.21 Very few injuries occurred in school, suggesting again that, despite several well-publicized tragic episodes, schools are safe places for children.22 This is in agreement with our previous reports that most school injuries are relatively minor and are related to falls and sports.23
A high proportion of children in both unintentional and assault groups sustained substantial injury to multiple body regions, an unexpected finding because this trauma presentation is more often associated with high-impact forces such as car crashes. The small proportion of children in this sample with gunshot wounds to the head probably reflects the relatively high probability of death before reaching the hospital.24
Almost one third of the children in both groups sustained injuries to the extremities, mainly fractures. These complex skeletal injuries are associated with complications and often require additional surgery after the initial treatment.25 An interesting finding was the higher proportion of children in the unintentional group with wounds to the deeper structures of the abdomen, a fact that probably explains the higher utilization rates of surgical intervention in this group.
There was no significant difference in injury severity between the 2 groups. This may reflect the greater lethality of assault-related injuries,24 resulting in children who die at the scene and are never transported to the hospital. In fact, the CDC lists intentional injuries (homicide and suicide), many of which are firearms related, among the most frequent causes of death in children and adolescents.26
More than one third of the children in both groups developed short-term disability mainly caused by wounds to the extremities that generally reduce mobility and ability to perform activities of daily living. Injuries to the central nervous system were the dominant cause of long-term disability in both groups. Particularly noticeable is the high incidence of spinal cord injury as a cause of long-term disability in the assault group. Spinal cord injuries are not frequently observed in this age group but have been associated with firearms-related injuries in adults.3
Although no direct measure of costs was available to the study, the results indicate that they may be substantial during both the acute and postacute phases. The high incidence of Emergency Medical Services use (including air evacuation), the surgery and intensive care utilization, and the extended length of stay reported suggest that the acute medical care costs are very high. Short- and long-term disabilities may result in substantial rehabilitation costs. Further adding to these costs, many of these disabled children can be expected to require special accommodations in school.
The overall national estimates of short- and long-term disability reported here should be viewed as a crude estimate only and interpreted with caution. NEISS data, used to estimate the total number of children hospitalized for firearms-related injuries, are obtained from a random sample of hospital EDs and are subject to fluctuations and inaccuracies inherent in any sampling scheme.2
The NPTR data, used to determine the proportion of injured children discharged with new disabilities, are derived from specialty pediatric trauma centers. These centers may be providing care for the most seriously injured children who survived until hospitalization. These children may be more seriously injured than the broader population in the NEISS national sample, which is derived from ED visits. NEISS data indicate that slightly fewer than half (46.7%) of people who were aged 19 years or younger and were hospitalized for firearms-related injuries were treated in pediatric hospitals from 1997 through 2000. NPTR hospitals are likely to be able to provide highly specialized pediatric trauma care, which may be associated with improved outcomes among survivors. However, the extrapolations used in preparing the disability estimates reported here assume that outcomes would be the same as those achieved in NPTR hospitals, regardless of hospital type or trauma specialization. Additional research is required to test this assumption. Finally, disability is assessed at discharge and does not include more subtle deficiencies that may develop after hospitalization.
It is by now well established that firearms-related injuries are a major cause of mortality and morbidity among children and adolescents in the United States. This study corroborates this assertion and provides a crude estimate of the population burden of disabling injuries to children resulting form firearms. However, prevention efforts to reduce the incidence of this type of injuries are complex because of the number and variety of guns in circulation and, in the case of the pediatric population, the various levels of maturity of the at-risk population.
It has been suggested27–29 that focusing on reducing the presence of accessible guns in the child’s environment—decreasing the number of guns in homes with children, advocating for safer storage of weapons, and modifying handguns to reduce the potential for harm—may be an effective preventive approach. In fact, the availability of firearms has been associated with unintentional deaths, suicides, and homicides among 5- to 14-year-olds.28,29
In this study, the vast majority of children who sustained unintentional firearms-related injuries were injured at home. The American Academy of Pediatrics suggests that guns should not be kept in homes where children live and, if they must be kept at home, that the weapons be stored safely.29 However, only half of pediatricians report that they sometimes counsel parents about firearms, and <1 in 6 always offers counseling.30
This study provides evidence of the burden of childhood firearms-related injuries in the United States. Fortunately, injuries and deaths as a result of firearms have been decreasing since 1993.2 An assessment of the factors that are responsible for this decline is needed to design additional prevention efforts.
This study was supported in part by grant H133B950006 from the National Institute on Disability and Rehabilitation Research (Washington, DC), by funding from the Maternal and Child Health Bureau of the Health Resources and Services Administration (Rockville, MD), and by the Harvard Youth Prevention Research Center (grant R49/CCR118602-02).
- Received March 21, 2003.
- Accepted October 7, 2003.
- Reprint requests to (C.D.) Tufts-NEMC, 750 Washington St, Box 75 K-R, Boston, MA 02111. E-mail:
- ↵Gotsch KE, Annest JL, Mercy JA, Ryan WR. Surveillance for fatal and nonfatal firearm-related injuries-United States, 1993–1998. CDC Surveill Summ. 2001;50 :1– 32
- DiScala C, Osberg JS, Savage RC. Children hospitalized for traumatic brain injury: transition to postacute care. J Head Trauma Rehabil. 1997;12 :1– 10
- ↵BMDP Statistical Software. Los Angeles, CA: BMDP Statistical Software Inc; 1985
- ↵WISQARS (Web-based Injury Statistics Query & Reporting System). Available at: www.cdc.gov/ncipc/wisqars. Accessed June 24, 2003
- ↵DiScala C, Gallagher SS, Schneps SE. Causes and outcomes of pediatric injuries occurring at school. J School Health. 1997;9 :167– 176
- ↵National Center for Injury Prevention and Control. Injury Fact Book 2001–2002. Atlanta, GA: Centers for Disease Control and Prevention; 2002
- ↵Firearm-related injuries affecting the pediatric population. Committee on Injury and Poison Prevention. American Academy of Pediatrics. Pediatrics. 2000;105(suppl) :888– 895
- ↵American Academy of Pediatrics. Periodic Survey of Fellows Number 47: Firearm Safety Counseling; 2001. Available at: www.aap.org/research/ps47exs.htm. Accessed December 27, 2002
- Copyright © 2004 by the American Academy of Pediatrics