Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
November 2004


Author Affiliations

Author Affiliations: Division of General Pediatrics (Dr Limbos), Division of Children, Youth, and Families (Dr Ramirez), Childrens Hospital Los Angeles, and Southern California Injury Prevention Research Center, Department of Epidemiology, University of California Los Angeles School of Public Health (Drs Ramirez and Kraus and Ms Park), Los Angeles, Calif; and the Department of Occupational and Environmental Health, College of Public Health, University of Iowa, Iowa City (Dr Peek-Asa).


Copyright 2004 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2004

Arch Pediatr Adolesc Med. 2004;158(11):1057-1061. doi:10.1001/archpedi.158.11.1057

Background  Injuries to the head comprise 20% to 39% of all school-related injuries. Head injuries among special education students have not been adequately described.

Objectives  (1) To examine the incidence and characteristics of head injuries in children enrolled in special education and (2) to determine the factors that increase the risk of sustaining a head injury compared with an injury to another part of the body.

Methods  Pupil Accident Reports for 6769 students enrolled in 17 of 18 special education schools in 1 large urban school district during the academic years 1994-1998 were reviewed, and information on the nature of injury, external cause, and activity was abstracted. Head-injured and nonhead-injured cases were identified and compared by race, sex, age, characteristics of injury, and disability category.

Results  Six hundred ninety-seven injury events were reported during the 4-year study period. The overall injury rate was 4.7 injuries per 100 student-years. Two hundred five children (29.4%) sustained injuries to the head, and the rate of head injury was 1.3 injuries per 100 student-years. Falls were the leading cause of injury. Head injuries were most commonly associated with physical education and unstructured play and usually occurred on the playground. Disproportionately more head than nonhead injuries were sustained in the classroom (12% vs 8%) and the bathroom (9% vs 3%). Compared with children with emotional/mental disabilities, children with multiple disabilities had the highest risk of a head injury (incidence density ratio, 2.4 [95% confidence interval, 1.6-3.5]), followed by children with physical disabilities (incidence density ratio, 1.8 [95% confidence interval, 1.1-3.1]). There appeared to be no significant difference in the rate of head injury by sex and age.

Conclusions  Modifications of the classroom, bathroom, and playground environments might reduce the risk of head injuries in children enrolled in special education. Special modifications and increased supervision may, in particular, reduce the risk of head injury for children with physical and multiple disabilities.

Almost 20% of children younger than 18 years have a chronic physical, developmental, behavioral, or emotional condition and require more than the usual level of health and related services.1 Children with disabilities have higher injury rates than their peers without disabilities.2 Approximately 6 million children with developmental and physical disabilities are enrolled in US public schools.3 These children sustain 17% of all school-related injuries4 but comprise only 2% of the student population.3,5 Ramirez et al6 reported an overall rate of 4.7 school-related injuries per 100 student-years among children in special education. Children with multiple and physical disabilities had a higher risk of injury compared with other disabled groups.

Injuries to the head comprise from 20% to 39% of all school-related injuries among the general population of school-aged children.2,4,7 National Pediatric Trauma Registry data reveal that injuries to the head and spinal cord account for the most severe injuries that occur at school. Head injuries among special education students have not been adequately described. Ramirez et al6 found that almost half of all injuries to children with disabilities were to the face and 16% to the head. In a day-care setting, Leland et al8 found that children with disabilities sustain 74% of injuries to the head. Head injuries also accounted for 42% of injuries among children with disabilities in a hospital sample, most commonly in children with epilepsy and cerebral palsy.9

Because of the frequency of head injuries and their potential for severe injury, we compared injuries to the head with other injuries sustained by children enrolled in special education schools in 1 large school district. In the special education population in particular, head injuries may result in disproportionately disabling sequelae because of the potential difficulty in assessing the severity of the injury and possible interaction with preexisting medical and developmental conditions. Describing characteristics of head injuries sustained at school by children with disabilities will focus prevention strategies and assist health care professionals in prescribing special precautions. The objectives of our study were to examine the incidence and characteristics of head injuries in this unique population and to determine the factors that increase the risk of sustaining a head injury compared with another injury.


Data for this analysis were obtained from a database on school-related injuries to children with disabilities in special education schools.6 The study population included 6769 students from 17 of all 18 special education schools in the Los Angeles Unified School District in California evaluated during the 1994-1998 academic years. One school was excluded because it is a medical provider to homebound students with severe medical needs.


Data collection procedures are detailed elsewhere,6 but we describe relevant methods here. An injury was defined as an event resulting in physical trauma to a student sustained during a school-sponsored activity. Categories of injury types were created based on text descriptions by school staff on injury reports. We identified 697 injuries from Pupil Accident Reports completed during the 1994-1998 academic years. Head-injury cases were defined as physical injury to the head and scalp, including the frontal area, because these regions overlay the brain and injury to these areas has the potential for associated intracerebral injury. Nonhead-injury cases had trauma to other regions of the body exclusive of the head, scalp, and frontal area. We linked head-injured and nonhead-injured cases to demographic and disability data from the school district’s Information Technology Division. Annually, students with possible disabilities are evaluated for special education eligibility and classified using California Department of Education definitions.10 Based on these, we created 5 broad categories of disability types: (1) developmental disability; (2) cognitive disability (autism, learning disability, mental retardation, and emotional disturbance); (3) physical disability (orthopedic, aphasia, blindness, deafness and blindness, hearing impairment, language/speech impairment, partial sight); (4) multiple disability; and (5) other disability (established medical disability, other health impairment, undetermined disability, and traumatic brain injury). Students with multiple disabilities have diagnosed disabilities in more than 1 category.

The external cause of injury was the first event in the chain of injury events. For example, if a child hit a pole and then fell, the external cause was a collision. Intentionality was not assigned because of the limited cognition of many students. Thus, bites, pushes, or kicks were labeled “injuries by other students.”

The population at risk was measured using the district’s yearly enrollment. Person-time was corrected for absenteeism by multiplying accrued person-time by 1 minus the average absentee rate for each school. Incidence was calculated per 100 student-years.

This study was approved by the institutional review boards of the University of California, Los Angeles, and the Los Angeles Unified School District.


All 6769 students were included in regression models that examined head injury risk by age, sex, and disability category. Other variables, such as activity at the time of injury, could not be included in models because stratum-specific student time was not available. We calculated rate ratios for head injury using generalized estimating equations to account for repeated subjects during the 4-year study period. A Poisson link function was used to adjust for varying time at risk. The deviance test was used to assess model fit. The final model did not include interactions because of unstable estimates.

Our head-injury case definition included 52 cases with injuries to the head and nonhead body regions. We conducted a sensitivity analysis excluding these 52 cases from our regression models and found no differences in estimates. Therefore, children who sustained both head and nonhead injuries were excluded. We purport that the causal pathway to head injury is similar among cases with trauma only to the head and cases with trauma to the head plus other body regions.


During the 4-year study period, 6769 enrolled students experienced 697 events resulting in a physical injury, for an overall injury rate of 4.7 per 100 student-years. The rate of head injury in this population was 1.3 per 100 student-years. Table 1 presents the characteristics of students with head injury and nonhead injury. One hundred nineteen head-injury cases (about 60%) had either emotional/mental (59 [28.8%]) or multiple (60 [29.3%]) disabilities. In contrast, 210 nonhead-injured cases (more than 40%) had emotional/mental disabilities. Table 2 presents the characteristics of children with head injuries. The crude rate of head injury by disability was greatest for children with other types of disabilities (2.2 per 100 student-years) and multiple disabilities (2.1 per 100 student-years). Children with emotional/mental disabilities had the lowest rates (0.9 per 100 student-years). Boys had a slightly higher rate (1.4 per 100 student-years) of head injury compared with girls (1.1 per 100 student-years) (Table 2).The 697 injury events led to 888 distinct physical insults to the body. One body region was affected in 547 events and multiple regions in 150 events. The distributions of injury type (Table 3) and external cause, associated activity, and school location (Table 4) differed significantly between head-injured and nonhead-injured cases.

Table 1. 
Image not available
Selected Characteristics of 697 Students With Head Injury and Nonhead Injury, 1994-1998 Academic Years*
Table 2. 
Image not available
Number of Students With Head Injury and Head Injury Rates by Selected Characteristics, 1994-1998 Academic Years
Table 3. 
Image not available
Number and Percentage of Injury Types Among 801 Students With Head and Nonhead Injuries, 1994-1998 Academic Years*
Table 4. 
Image not available
Number and Percentage of Primary External Cause, School Activity, and School Location Among Injured Students, 1994-1998 Academic Years*

The final model was mutually adjusted for disability, sex, age, year in school, and school of attendance (Table 5). Residual analysis indicated adequate fit. Compared with children with emotional/mental disabilities, children with multiple disabilities had the highest rate of a head injury (incidence density ratio, 2.4 [95% confidence interval, 1.6-3.5]), followed by children with physical disabilities (incidence density ratio, 1.8 [95% confidence interval, 1.1-3.1]). There appeared to be no substantial difference in the rate of injury by sex (incidence density ratio, 1.2 [95% confidence interval, 0.9-1.7]) or age. Rate ratios for injury comparing younger to older children ranged from 1.2 to 0.6, with wide confidence intervals.

Table 5. 
Image not available
Rate Ratios (Incidence Density Ratios) and Confidence Intervals (CIs) for Sustaining an Injury to the Head, 1994-1998 Academic Years*

Head injuries were frequent injuries in this population of children in special education schools. The head injury rate in this special education population was similar to the rate for children without disabilities. We found few age and no sex differences in the head injury rate. Studies of school-related injuries in children without disabilities report that boys have 2 to 3 times the injury rate of girls the same age.11 In contrast, Dunne et al2 found no sex difference in injury rates among children with developmental disabilities. However, none of these studies specifically addressed head injuries in the special education setting.The use of existing records can lead to the potential for reporting bias. District policy requires reporting injuries during or related to school activities, however slight. Despite this policy, there may be underreporting or overreporting of injuries within each individual school. Generally, however, school administrators acknowledge that children with special needs are much more closely monitored and reporting may be higher than for students in regular education settings. Furthermore, reporting across special education schools is likely consistent.

We define injuries to the head as cases resulting in any physical trauma to the head, scalp, and frontal region overlaying the intracerebral contents. This definition was designed to describe the epidemiology of impacts to the head, which have the potential for more severe outcomes like traumatic brain injury. Ultimately, uncovering the pattern of mild head injury events is a basis for understanding how to prevent more severe traumatic head injury events. However, this is not a standard classification used among populations of children who are hospitalized or fatally injured, in whom it is more common to examine traumatic brain injury.

We were unable to examine more severe outcomes because severity was not described in these records that were completed at school following the event. Cases may have received additional evaluation by a physician. Detailed medical records were not available to us but could be an important linkage for future research. Data on school location and activity were missing in at least 28% of cases. No significant missing data patterns were found by disability or demographics, and thus, we suspect no substantial biases from missing data.

Despite these limitations, we found several risk factors that have implications for prevention. First, students with physical and multiple disabilities had a significantly higher risk of sustaining a head injury than students with different disabilities. Ramirez et al6 also reported that students with multiple disabilities had the highest risk of injury, followed by students with physical disabilities. This difference in risk may be related to physical limitations such as impaired balance and ambulation and/or to differences in exposure. Students with orthopedic handicaps and sensorimotor deficits may be less able to negotiate the school environment and protect themselves from falls, the leading cause of head injury. Additionally, their normal cognition may allow them to explore the environment as children without disabilities would, thus exposing them to potential injury hazards. Students with emotional/mental disabilities had the lowest risk of head injury. These children, although having impaired cognitive and social functioning, had no physical disabilities, suggesting that physical limitations more so than cognitive limitations place children at risk for head injury.

Contrary to studies of injury in children without disabilities, this study did not find sex and age to be associated with risk of head injury, which is consistent with 2 studies of special needs children.2,8 Dunne et al2 theorized that this population may be overprotected by caretakers, leading to fewer exposures typically associated with sex in the able-bodied population. We propose the same overprotection effect also results in restricted activity regardless of age. Hence, the presence of an impairment may reduce sex-injury and age-injury relationships, which typically appear in the general education population.

Our findings suggest modifications of the classroom environment might reduce the risk of head injuries in children enrolled in special education. Classrooms for students with physical and multiple disabilities might include padding for furniture edges and for floor surfaces. We found, in particular, that injuries sustained in the bathroom and during toileting were more likely to be injuries to the head rather than to other parts of the body. Interventions that might be implemented in this setting include fall-resistant surfaces and grab handles, in addition to padding of toileting and changing areas. Helmets might also be considered for children with greater risk for head injury. Staffing could be structured to maximize supervision for students, activities, and locations associated with greatest risk.

School is an important part of cognitive and social development for all children. However, children enrolled in special education are at increased risk of injury at school, in particular, head injury. Classroom modifications and policies regarding safety and supervision practices may help reduce the incidence of head injuries in this population and allow students with even complex medical conditions to more fully participate in the school experience.

What This Study Adds

Head injuries comprise a significant number of the injuries sustained by children in the school setting and account for the most severe injuries that occur at school. Few data exist about the risk, characteristics, and circumstances of injury to the head among children with disabilities enrolled in special education. This study provides new information about the incidence and characteristics of head injuries in a large population of children with physical and/or developmental disabilities in the special education setting and determines the factors that increase the risk of sustaining a head injury. The study results have implications for developing prevention strategies in the classroom and playground environments of this unique population of children.

Back to top
Article Information

Correspondence: Mary Ann P. Limbos, MD, MPH, Childrens Hospital Los Angeles, 4650 Sunset Blvd, MS 76, Los Angeles, CA 90027 (

Accepted for Publication: May 13, 2004.

Newachek  PWStrickland  BShonkoff  JP  et al.  An epidemiologic profile of children with special health care needs. Pediatrics 1998;102117- 121
Dunne  RAsher  KRivara  F Injuries in young people with developmental disabilities: comparative investigation from the 1988 National Health Interview Survey. Ment Retard 1993;3183- 88
Snyder  THoffman  C Digest of Education Statistics, 2001.  Washington, DC National Center for Education Statistics2002;Report NCES 2002130
Di Scala  CGallagher  SSchneps  S Causes and outcomes of pediatric injuries occurring at school. J Sch Health 1997;67384- 389
Aron  LLoprest  PSteuerle  C Serving Children with Disabilities: A Systematic Look at the Programs.  Washington, DC Urban Institute Press1996;
Ramirez  MPeek-Asa  CKraus  J Child disability and risk of school-related injury. Inj Prev 2004;1021- 26
Gibson  HKlassen  TP How safe are our schools? Can J Public Health 1996;87106- 108
Leland  NGerrard  JSmith  D Comparison of injuries to children with and without disabilities in a day-care center. J Dev Behav Pediatr 1994;15402- 408
Berggreen  S Accidents and surgical emergencies in a population of mentally retarded children. Acta Paediatr Scand 1973;62289- 296
 California Special Education Programs: A Composite of Laws.  Sacramento California Dept of Education, Special Education Division1998;
 Injuries in the School Environment: A Resource Guide (2nd ed).  Newton, Mass Children's Safety Network at Education Development Center Inc1997;