BACKGROUND AND OBJECTIVE: Previous studies have reported concerning numbers of injuries to children in the school setting. The objective was to understand temporal and demographic trends in intentional injuries in the school setting and to compare these with intentional injuries outside the school setting.
METHODS: Data from the National Electronic Injury Surveillance System–All Injury Program from 2001 to 2008 were analyzed to assess emergency department visits (EDVs) after an intentional injury.
RESULTS: There were an estimated 7 397 301 total EDVs due to injuries sustained at school from 2001 to 2008. Of these, an estimated 736 014 (10%) were reported as intentional (range: 8.5%–10.7% for the study time period). The overall risk of an EDV after an intentional injury in school was 2.33 (95% confidence interval [CI]: 1.93–2.82) when compared with an EDV after an intentional injury outside the school setting. For intentional injury–related EDVs originating in the school setting, multivariate regression identified several demographic risk factors: 10- to 14-year-old (odds ratio [OR]: 1.58; 95% CI: 1.10–2.27) and 15- to 19-year-old (OR: 1.69; 95% CI: 1.01–2.82) age group, black (OR: 4.14; 95% CI: 2.94–5.83) and American Indian (OR: 2.48; 95% CI: 2.06–2.99) race, and Hispanic ethnicity (OR: 3.67; 95% CI: 2.02–6.69). The odds of hospitalization resulting from intentional injury–related EDV compared with unintentional injury–related EDVs was 2.01 (95% CI: 1.50–2.69) in the school setting. These odds were found to be 5.85 (95% CI: 4.76–7.19) in the outside school setting.
CONCLUSIONS: The findings of this study suggest a need for additional prevention strategies addressing school-based intentional injuries.
- CI —
- confidence interval
- ED —
- emergency department
- EDV —
- emergency department visit
- NEISS —
- National Electronic Injury Surveillance System
- NEISS-AIP —
- National Electronic Injury Surveillance System–All Injury Program
- OR —
- odds ratio
What’s Known on This Subject:
Injuries sustained by children in the school setting have a significant public health impact. A concerning subgroup of school injuries are due to intentional and violent etiologies. Several studies have identified a need for further research to understand intentional school-based injuries.
What This Study Adds:
This study discusses national estimates and trends over time and risk factors of intentional injury–related emergency department visits due to injuries sustained in the school setting.
American children spend ∼180 days each year at school, making it the most common place other than their home where they spend time.1–3 Consequently, it is perhaps unsurprising that they sustain an estimated 10% to 25% of all injuries in the school setting.1,4,5 Whereas most injuries in this setting are documented to be the result of unintentional mechanisms, including sports or other physical activity,4–13 a concerning subgroup of school injuries are due to intentional and violent etiologies, with 10% to 17% of school injuries resulting from intentional mechanisms.4–6 Additionally, results from the Youth Risk Behavior Surveillance System for 2011 indicate that 20.1% of youth nationwide reported being bullied on school property.14
A number of studies have shown significant health-related issues, including an increase in risk-taking behaviors, poor school performance, and absenteeism related to bullying and violence.15–23 As a consequence, there have been numerous efforts to reduce injuries in the school setting, including increased supervision of physical activities, improving student:teacher ratio, teaching students how to deescalate bullying, adding safety officers, teaching school staff how to address violence, and extra activities for students such as music and arts classes to avoid unstructured or unsupervised time in school.2,10,22,24
Pediatric intentional and unintentional injuries have been thoroughly discussed in the literature,20,25–30 and several studies have identified a need for further research to understand intentional school-based injuries.1–7,10,24,25,31–33 With the increasing awareness of physical and mental health issues associated with intentional injuries resulting from interpersonal violence and bullying in children, it is important to examine the overall trend for such injuries in the school setting to tailor appropriate measures for safety and prevention.1,6,24,31,32 Therefore, the goal of this study was to examine the national demographic and temporal trends of intentional injuries in the school setting. Our specific objective was to compare emergency department visits (EDVs) for intentional injuries sustained in the school setting to EDVs resulting from intentional injuries sustained outside of the school setting to better guide preventive efforts.
This was a retrospective cohort study using data from the National Electronic Injury Surveillance System (NEISS)–All Injury Program (NEISS-AIP) from January 1, 2001, through December 31, 2008. The study was classified as exempt by the Rhode Island Hospital Committee on the Protection of Human Subjects.
Study Setting and Population
Data from NEISS-AIP were obtained at www.icpsr.umich.edu/icpsrweb/ICPSR/series/00198. NEISS-AIP is a database of EDVs for injury reported by 66 hospitals (of the 100 NEISS database hospitals) selected from throughout the United States. Sampling is performed to derive a stratified probability sample of the 5000 hospitals in the United States with emergency departments (EDs) that have at least 6 beds and are operational 24 hours per day. A sampling weight is assigned to each case on the basis of the inverse probability of selection to provide national estimates. NEISS-AIP is operated by the US Consumer Product Safety Commission in collaboration with the National Center for Injury Prevention and Control, Centers for Disease Control and Prevention.34 A NEISS-trained researcher at each institution is responsible for reviewing the chart and entering data related to each injury visit, including the patient's age, gender, race, and ethnicity; the date of the injury; the location where the injury occurred in the community; the body part to which the injury occurred; the intent and perpetrator (if intentional); the patient's diagnosis and disposition; and a detailed account of the injury event as gleaned from the medical records. Trained quality assurance coders at the NEISS center code from the narrative and provide the “cause of injury.” Data values with <20 records, with weighted estimates <1200, or with a coefficient of variation >30% are considered by NEISS to be unstable and were interpreted accordingly when making decisions about variable recoding and regression modeling.
Our study examined nonfatal injury-related EDVs from January 2001 through December 2008. The injury-related EDV rate was calculated by using the average census for the age group from the US census data over the study time period (http://www.census.gov/popest/data/historical/2000s/vintage_2008/). Intent of injury is recorded as “intentional” or “unintentional/unknown” in the data set and was used as such for this study. NEISS-AIP defines intentional injuries as “an injury [that] was caused by an act carried out on purpose by oneself or by another person(s), with the goal of injuring or killing”.35 Unintentional injuries were defined as those due to “not deliberate means.” Because NEISS data are not intended to capture all deaths related to injuries, EDVs associated with fatality or “dead on arrival” were excluded from our analysis.
The location of injury was identified as “school setting” when documented as “school,” whereas injuries that occurred in any location other than school were coded as “outside school setting.” Those EDVs where the location was unknown were excluded from the analysis. Visits by patients between the ages of 5 and 19 years were included in the analysis. Age was categorized into 3 groups: 5 to 9 years, 10 to 14 years, and 15 to 19 years, which approximate the traditional elementary, middle, and high school age groups, respectively. NEISS coding for race and ethnicity defines “white” as non-Hispanic, “black” as both non-Hispanic and Hispanic black, and “Hispanic” as “Hispanics for all races other than black.” Although it is noted by NEISS-AIP that ∼17% of the race/ethnicity data are missing for the overall data set, these data were found to be missing for 7% for our study sample.
Diagnosis at discharge was restricted to the top 5 categories and “other.” We created a “traumatic brain injury” category for the diagnoses of concussion and fracture or internal injury when “head” was the injured body part.5,36 All other diagnoses were left as originally coded in the NEISS-AIP database. Immediate cause of injury was also restricted to the top 5 causes and “other.” Disposition was recoded as follows: “discharged” (those patients who were “treated and released”), “hospitalized/transferred/observed” (those patients who were “treated and transferred”, “hospitalized,” or “held for observation”), and “other” (those patients who “left without being seen/left against medical advice,” or whose disposition was not specified or was “unknown”).
Data were analyzed with SAS (version 9.3; SAS Institute, Inc, Cary, NC) by using the SURVEYFREQ and SURVEYLOGISTIC procedures. A specific weighting factor was assigned to each NEISS hospital related to the inverse probability of selection, permitting generalization of the data to the national population. The probability value for statistical significance was set at an α level of 0.05. National estimates, frequencies, and 95% confidence intervals (CIs) were calculated on the basis of the sample weights and clusters. Tests for trend for intentional injuries over time were performed by using PROC SURVEYLOGISTIC with treatment year scored as a linear sequence. Separate multivariable logistic regression models were fit to the in-school and out-of-school data to model the probability of an intentional injury–related EDV, adjusted for gender, age, race, and ethnicity.
During the study time period (2001–2008), NEISS-AIP data identified an estimated 44 721 462 injury-related EDVs for youth 5 to 19 years of age where the location of injury was known. Of these, an estimated 7 397 301 (12.5%) injuries occurred in the school setting (an average of 924 662 EDVs per year). This finding represents an annualized injury-related EDV rate of 45 per 1000 children in the school setting, based on the average 2001–2008 US census (census mean = 20 480 716 for 5–19 years of age).
There were a total of 736 014 (average = 92 001 per year) intentional injury–related EDVs from the school setting, with an annual rate of 4.5 per 1000 children. For the “outside school” setting, the total number of intentional injury–related EDVs was 2 293 922 from 2001 to 2008 (average = 286 740 per year), with an annual rate of 14 per 1000 children. For the unintentional injuries, the annual rate of EDVs was found to be 41 per 1000 children in the school setting (average = 832 660 per year) and 214 per 1000 children in the “outside school” setting (average = 4 378 770 per year) for the study time period. The proportion of EDVs due to intentional injury at school varied over time, from 10.7% of all injuries in 2001–2002 to 8.5% of all injuries in 2007–2008 (Table 1, Fig 1). A test for trend revealed a decreasing trend for intentional injuries inside school (Ptrend = .046) but not for outside school (Ptrend = .462).
The demographic characteristics for patients with intentional injury–related EDVs from both in-school and out-of-the-school settings are summarized in Table 2. The odds for an EDV due to an intentional injury (with reference to unintentional injury) in the school setting compared with outside the school setting were found to be 1.76 (95% CI: 1.45–2.12) in the univariate logistic models. Male gender and the 10- to 14-year age group were identified as risk factors for intentional injury–related EDVs in the school setting and female gender and the 15- to 19-year age group were identified as risk factors for intentional injury–related EDVs in the outside school setting (Table 3). Patients of black and American Indian race and of Hispanic ethnicity were also found to have higher risk of intentional injury–related EDVs in both settings when compared with unintentional injuries (Table 3).
For intentional injury–related EDVs from the school setting (Table 4), various demographic risk factors were identified as follows: 10- to 14-year and 15- to 19-year age groups, black and American Indian race, and Hispanic ethnicity (odds ratio [OR]: 3.67; 95% CI: 2.02–6.69). When contrasting the 2 settings (Table 4), girls were less likely to have an intentional injury–related EDV compared with boys in the school setting but were more likely than boys to have an intentional injury–related EDV in the outside-school setting. The risk of an intentional injury–related EDV was found to be lower for the 15- to 19-year age group in the school setting when compared with the outside school setting. Black race and Hispanic ethnicity were both found to confer higher risks for the school setting when compared with the outside-school setting (Table 4).
A description of the EDVs is presented in Table 5 for the 2 settings. Fractures were identified more often in the school setting, whereas lacerations were identified more often in the outside-school setting. The proportion of injuries due to traumatic brain injury was higher in the school setting compared with the outside-school setting. Poisoning was identified as 1 of the top 5 diagnoses for the outside-school setting. Conversely, strain/sprain was common in the school setting but comparatively not as common outside of school. The proportion of EDVs due to assault as an immediate cause was higher in the school setting, and those due to self (intentional self-harm) or legal intervention were more common in the outside-school setting.
With regard to disposition, the overall risk of hospitalization (hospitalized/transferred/observed) in all settings was 4.91 (95% CI: 4.03–5.97) for any intentional injury–related EDV compared with an unintentional injury–related EDV. The likelihood of hospitalization after an intentional injury–related EDV (versus an unintentional injury–related EDV) was lower for the school setting (OR: 2.01; 95% CI: 1.50–2.69) when compared with outside the school setting (OR: 5.85; 95% CI: 4.76–7.19) (Table 3).
Despite increased emphasis on safety at school and evidence of a decreasing trend, children aged 5 to 19 years still experienced a substantial number of intentional injuries at school over the time period studied. A recent study has also noted that ED utilization by pediatric patients due to violence (not limited to a specific setting) has not changed.29 Additionally, our study has shown that the proportion of injury visits originating from school that are intentional continues to be higher than for those suffered out of the school setting, suggesting the need for additional interventions to enhance safety at school.
This study identified a gender disparity in intentional injury–related EDVs for the 2 settings, with boys more likely to have intentional injury–related EDVs when originating in the school setting and girls more likely in the outside-school setting. This gender difference is interesting, especially because previous studies have identified male gender to be a risk factor for injuries in a variety of settings.6,11,12,14,21,26 Further research is needed to understand why this disparity exists.
Another important finding from this study is that middle school–aged children (10- to 14-year-old group) carry a significant burden of intentional injury–related EDVs in the school setting. This is in contrast to the out-of-school setting where high school–aged students (15- to 19-year-old group) were found to have the highest burden. This distribution persisted through the study time period and was consistent with our earlier study.5
Previous studies have documented that African American race and Hispanic ethnicity are associated with increased likelihood of bullying and its associated mental and physical health impacts.18,19,21 Our study also suggests that racial disparity persists for the risk of intentional injury–related EDVs. In addition to students of black race and Hispanic ethnicity, American Indian youth were found to be high risk. Interestingly, these demographic risks are much higher in the school setting compared with the outside-school setting. This finding suggests the need for developing interventions that meet the needs of these populations. Of note, there is no information in NEISS-AIP regarding the racial and ethnic background of perpetrators or of the overall racial/ethnic composition of the school where an injury occurred, thus limiting understanding of the racial disparity phenomenon.
When considering the types of injuries sustained due to intentional mechanisms, it is important to note that the risk of sustaining traumatic brain injuries was similar to that of sustaining abrasions/contusions. The head and neck region was also found to be the most common body region affected, followed by the upper extremities. This finding is in contrast to earlier studies that documented upper extremity to be the most frequently affected body part in all school injuries.4,7 Possibly related to this phenomenon, most of the intentional injuries in our study sample were reported to be to the result of assault, and ∼90% reported the perpetrator to be either a friend or an acquaintance in the school setting. Other studies have shown that a bullied victim frequently knew the bully and that the incident usually involved a past disagreement.28 When considering the perpetrators, it is notable that 10% of the EDVs were reportedly due to multiple perpetrators in the school setting. Intentional injuries and violence have immense mental health effects on the victims, and studies have reported that victims may be involved with future violence, either as victims or as perpetrators.21,22 With the knowledge that intentional injuries continue to be a significant problem in the school setting, and in light of the limited information that these data offer regarding perpetrators, it is clear that more work needs to be done to understand the victim-bully relationship to design preventive efforts.
The risk of hospitalization in our study was found to be higher for intentional versus unintentional injury–related EDVs in either setting. If being hospitalized is a proxy for sustaining a serious injury, these data document that children sustained more serious injuries resulting in hospitalization when the injury was intentionally inflicted compared with unintentional injuries. This finding further emphasizes the importance of the problem and need for preventive efforts.
This study was limited in terms of its ability to completely capture the full range of injuries suffered by children, both in and out of the school environment. Because the NEISS-AIP data source is predicated on the injury resulting in an EDV, it does not include patients who were directly admitted to a hospital. This limitation also points toward the inherent selection bias dependent on the sampling design and a caregiver’s decision to seek medical care in the ED. For example, studies have shown that there may be overreferral to the ED from school after an injury.9 Additionally, care may have been sought at a primary care office, urgent care center, or other medical provider, and those visits would not have been captured here. Another limitation is the data's wide CIs that result from national estimates that are based on a restricted sample of 66 hospitals. It is also important to note that NEISS data do not code for the acuity of a visit and do not report data on fatalities in the school setting, which may limit understanding of the most serious injuries.
Although the NEISS-AIP database collects information on a wide range of patients, it lacks the granularity that would be necessary to more fully answer questions one might have about an individual injury. This, in turn, makes it difficult to ascertain the specific circumstances under which each of the injuries occurred. A plethora of characteristics related to the school environment, including the physical environment and location of a school, supervision provided while in the school environment, demographic characteristics of perpetrators, and many others, all may have an effect on the occurrence of an individual intentional injury. Precise detail about the intentional injury event can assist in developing preventive strategies especially related to using the principles of the Haddon Matrix.37
There are substantial numbers of intentional injury–related EDVs from the school setting. Our study identifies age, gender, and racial and ethnic disparities associated with school-based intentional injuries, emphasizing the need for the development of culturally appropriate preventive strategies.
- Accepted November 15, 2013.
- Address correspondence to Siraj Amanullah, MD, MPH, 593 Eddy St, Department of Emergency Medicine, Claverick Building, Second Floor, Providence, RI 02903. E-mail:
Drs Amanullah and Linakis conceptualized and designed the study, carried out the initial analysis, and drafted the initial manuscript; Drs Heneghan, Steele, and Mello helped in the design of the study, analysis, and reviewed and revised the manuscript; and all authors approved the final manuscript as submitted.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: No external funding.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- ↵Hull J. Time in school: how does the United States compare? The Center for Public Education; December 2011. Available at: www.centerforpubliceducation.org/Main-Menu/Organizing-a-school/Time-in-school-How-does-the-US-compare. Accessed December 1, 2012
- Feldman W,
- Woodward CA,
- Hodgson C,
- Harsanyi Z,
- Milner R,
- Feldman E
- Dale M,
- Smith ME,
- Weil JW,
- Parrish HM
- Boyce WT,
- Sprunger LW,
- Sobolewski S,
- Schaefer C
- ↵Ramirez M, Wu Y, Kataoka S, et al. Youth violence across multiple dimensions: a study of violence, absenteeism, and suspensions among middle school children. J Pediatr. 2012;161(3):542–546
- Peguero AA
- Maitra AK,
- Sweeney G
- Laflamme L,
- Menckel E
- ↵National Center for Health Statistics. NEISS-AIP, National Electronic Injury Surveillance System–All Injury Program. Available at: http://healthindicators.gov/Resources/DataSources/NEISS-AIP_88/Profile. Accessed December 2012
- ↵Centers for Disease Control and Prevention. Injury center. Available at: www.cdc.gov/ncipc/wisqars/nonfatal/definitions.htm#nonfatalintent. Accessed Decmeber 2012.
- ↵Hostetler SG, Xiang H, Smith GA. Characteristics of ice hockey-related injuries treated in US emergency departments, 2001–2002. Pediatrics. 2004;114(6). Available at: www.pediatrics.org/114/6/e661
- Baker SP,
- Li G
- Copyright © 2014 by the American Academy of Pediatrics