Brain Injury In Children
Burden of Illness of TBI in Children and Adolescents
The estimated incidence of TBI doubles between the ages of 5 and 14 years, and peaks for both males and females during adolescence and early adulthood to approximately 250 per 100,000 (Centers for Disease Control and Prevention, 1998). Motor vehicle accidents are the cause of one-third to one-half of TBI in adolescents. In younger children, pedestrian traffic accidents are the leading cause of TBI, followed by falls.
The greatest proportion of people who survive TBI are children and adolescents. Because the lives of most survivors of moderate to severe TBI involve chronic, life-long disabilities with varying degrees of dependence (Chesnut, Carney, Maynard, et al., 1999), the cost in individual suffering, family burden, and financial burden to society may be greater for those who have more years to live.
For children with TBI who have multiple functional deficits, limitations in bathing, dressing, and walking are observed in between 50 percent and 90 percent, depending upon and directly proportional to the number of functional deficits (DiScala, Osberg, and Savage, 1997). For children with four or more functional deficits, impairments in self-feeding, cognition, and behavior can be observed in 75 percent, as well as impairments in speech in 67 percent, vision in 29 percent, and hearing in 16 percent (DiScala, Osberg, and Savage, 1997).
The opinion has been expressed by a number of clinicians and researchers that one of the most important problems in child and adolescent TBI is underdiagnosis (Bergman, 1998; DePompei, 1998; Glang, 1999; Gordon, 1999; Sohlberg, 1999; Ylvisaker, 1998). They believe that many children with TBI are either not identified as having a problem or receive the wrong diagnosis. The concern is that the problems of these children remain unrecognized or are treated with methods developed for other pathologies that are inappropriate to the special needs of TBI. In the 18th Annual Report to Congress on the Implementation of the Individuals with Disabilities Education Act (U.S. Department of Education, 1996) the number of children receiving services because of their disabilities was tabulated by diagnosis. For the 1994-1995 school year, only 7,188 students ages 6 to 21 years were identified within the school system as having a traumatic brain injury. The gap between this number and reports of incidence of TBI among children and adolescents suggests that many children with TBI may be misidentified or unidentified. Consequently, we do not have a complete picture of the burden of illness, either on an individual level or a population basis.
For children who have been accurately identified, comparative studies of TBI need to include mediators of outcome that account for variations in outcome. Individuals with TBI are part of a diverse group. For example, with respect to severity, the range of outcome is from death or persistent unresponsiveness at one extreme to full recovery within a relatively short period of time at the other extreme. Between these two extremes are different profiles of ability and need -- at every level of severity -- depending on preexisting factors, the exact location and nature of the brain injury, and other factors. In addition to type and location of injury and the child's developmental stage, factors such as injury severity, time from injury to evaluation, age at injury, and pre-injury characteristics appear to be important predictors of recovery.
The National Pediatric Trauma Registry (NPTR) is a multi-institutional database of information on children and adolescents (0 to 19 years of age) who have been admitted to a hospital for an injury since 1985. A review of data from the NPTR on 24,021 cases of TBI that occurred over a period of 7.5 years revealed that 16.6 percent of the children sustained severe injuries based on the Injury Severity Score (ISS), and 11.5 percent of the children sustained severe head injuries based on the Glasgow Coma Scale score (GCS) (DiScala, Osberg, and Savage, 1997). The same study indicated that 6.1 percent of children with brain injuries in the sample died, compared with 0.5 percent of children without brain injuries from the entire data set of all traumas. This study also revealed a relationship between multiple injuries and degree of functional limitation. For children with injuries to the head, extremities, and other areas of the body, 54 percent had four to nine functional limitations, compared with 32 percent who had injuries to the head and other body regions (but not to extremities), and 14 percent with head injury only.
Two studies about long-term outcomes from mild (Klonoff, Clark, and Klonoff, 1993) and severe (Boyer and Edwards, 1991) brain injury sustained during childhood or adolescence suggest a better outcome for children who suffer less severe brain injuries. In the first study, 159 children from an original sample of 231 children (average age 8 years) admitted to one of two university hospitals for brain injury (90 percent mild) over a 15-month period were located and interviewed 23 years after injury. Fifty-nine percent of participants were married, and 80 percent were employed full-time; 1 percent were disabled, and 3 percent were unemployed. In the second study, the severely injured sample (n = 220) was prospectively followed for 3 years after discharge from one inpatient rehabilitation program that receives referrals from regional trauma centers. At discharge, the following proportions by age group returned home: 0-5 years, 75 percent; 6-10 years, 100 percent; 10-15 years, 80 percent; 16-21 years, 60 percent. After 3 years, 79 percent were home, and 10 percent were in long-stay nursing facilities. At 1-year followup, 8 percent were in regular education, 39 percent in special education, and 17 percent in cognitive therapy; 26 percent were not in an educational program, and 1 percent were in day care. The disposition of the remaining 9 percent was not specified. Glasgow Outcome Scale scores (GOS) after 3 years were 3 percent dead, 14 percent vegetative, 14 percent severe, 17 percent moderate, and 52 percent good.
A series of six prospective studies from one cohort of patients treated at one of two regional hospitals suggests an indirect relationship between severity of injury and scores on neuropsychological tests (Fay, Jaffe, Polissar, et al., 1993; Fay, Jaffe, Polissar, et al., 1994; Jaffe, Fay, Polissar, et al., 1992; Jaffe, Fay, Polissar, et al., 1993; Jaffe, Polissar, Fay, et al., 1995; Yorkston, Jaffe, Polissar, et al., 1997). Children injured between the ages of 6 and 15 years who were admitted to the hospital were prospectively followed beyond 3 years. Results showed consistent association of severity with neuropsychological test scores at early outcome, 1 year, 3 years, and > 3 years. The written language skills of the more severely injured children were worse than those of their less severely injured counterparts. However, no significant difference was evident on test scores between the less severely injured children (GCS 13-15) and controls.
The weakness in these studies is the failure to stratify by other factors that interact to influence outcome; thus, their results should be viewed with caution when attempting to infer a direct relationship between severity and outcome. For example, pooling age groups or not accounting for pre-injury characteristics may mask variability in outcomes associated with severity. Although children with less severe injuries may have better outcomes, there may be many exceptions to the general "dose response" rule (Ylvisaker, 1999).
Time from Injury to Evaluation
A review that focused on age-related differences in child and adolescent TBI (Taylor and Alden, 1997) found that with longer time since injury, children's cognitive and academic skills decreased, suggesting that deficits may appear or become worse over time. The question of deterioration over time and its measurement is complex and must be examined with all potential influences in mind. Assessments conducted sooner after injury may be less reliable in measuring or predicting long-term outcome than those taken at a later time. Many assessment tools have built increases into their age-related norms for scoring. With these tools, an observed decrease in scores may actually reflect no change (no decrease or increase in abilities). Additionally, factors that interact to produce deterioration over time are (1) neurologic maturation (part of the brain that is injured may need to mature physically to support later developmental acquisitions, but this maturation is blocked by the earlier injury); (2) psychoreactivity (related to personal loss, loss of friends, etc.); and (3) teaching and behavior management strategies that are not appropriate for children with TBI (Ylvisaker, 1999).
The Taylor and Alden study (1997) also suggests that assessment during puberty may be particularly unreliable because of influences during that phase such as social difficulties, verbal memory impairment, and behavioral problems. A second study describing pediatric TBI rehabilitation medical management (Jaffe and Hays, 1986) indicated that special problems with brain injury incurred during childhood and adolescence include hypopituitarism, growth impairment, and isosexual precocious puberty.
Age at Injury
A retrospective, population-based study (Michaud, Rivara, Grady, et al., 1992) of severe admissions for head injury (GCS < 8, N = 75, < 16 years of age) to a level 1 trauma center found that as age increased, the proportion of good recovery also increased, and the proportion of death decreased. The Taylor and Alden review (1997) reported that outcomes were worse for children < 7 years old compared with those > 7 years but did not specify whether severity of injury was accounted for in the studies reviewed. Therefore, the less favorable outcome they reported in younger children may reflect (1) a higher probability of severe injuries for children < 7 years, or (2) a lower probability of recovery from injuries for children younger than 7 compared with children 7 or older, after accounting for injury severity.
Two studies, one prospective (Boyer and Edwards, 1991) and one descriptive (Jaffe and Hays, 1986), indicate that a greater proportion of children than adults who sustain TBI have premorbid behavioral and emotional problems. In the Boyer and Edwards (1991) sample (n = 220), 35 percent had learning disabilities, attention deficit, or emotional difficulties prior to injury. Emotionally or psychologically compromised children may be particularly vulnerable to TBI.
In a study about the effects of family on recovery from TBI in children (Taylor, Drotar, Wade, et al., 1994), a predictive model that includes measures of pre-injury child and family status was developed on 96 children with TBI and 71 orthopedic controls, ages 6 to 12 years at time of injury. Pre-injury factors including child behavior, marital adjustment, family functioning, health and daily living, and parent behavior accounted for 57 percent of the variance in postinjury behavior problems, 37 percent of the variance in adaptive behavior, and 13 percent of the variance in achievement