Published online April 2, 2007
PEDIATRICS Vol. 119 No. 4 April 2007, pp. 749-758 (doi:10.1542/10.1542/peds.2006-1186)
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Swaine, B. R.
Right arrow Articles by Pless, I. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Swaine, B. R.
Right arrow Articles by Pless, I. B.
Related Collections
Right arrow Office Practice
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Facebook   Add to Reddit   Add to Technorati   Add to Twitter  
What's this?

ARTICLE

Previous Head Injury Is a Risk Factor for Subsequent Head Injury in Children: A Longitudinal Cohort Study

Bonnie R. Swaine, PhDa,b,c, Camille Tremblay, MScc, Robert W. Platt, PhDd,e, Guy Grimard, MDb,f, Xun Zhang, PhDe and I. Barry Pless, MDd,e

a School of Rehabilitation
b Faculty of Medicine, Université de Montréal, Montréal, Québec, Canada
c Centre de Recherche Interdisciplinaire en Réadaptation du Montréal Métropolitain, Institut de Réadaptation de Montréal, Montréal, Québec, Canada
d Departments of Pediatrics and Epidemiology and Biostatistics, McGill University, Montreal, Quebec, Canada
e McGill University Health Centre, Montréal Children's Hospital Research Institute, Montréal, Québec, Canada
f Hôpital Ste-Justine, Montréal, Québec, Canada


   ABSTRACT
TOP
 ABSTRACT
METHODS
 RESULTS
 DISCUSSION
 STUDY LIMITATIONS
 CONCLUSIONS
 REFERENCES
 
OBJECTIVE. The objective of this study was to determine whether children who sought care for a head injury were at greater risk of having a subsequent head injury within the following 6 and 12 months compared with children who sought care for an injury other than to the head.

DESIGN/SETTING. This was a longitudinal cohort study conducted in the emergency departments of 2 Montreal (Quebec, Canada) pediatric hospitals.

PARTICIPANTS. The parents of 11867 injured children aged 1 to 18 years were interviewed by telephone at 6 (n = 10315) and 12 (n = 9486) months after their child's injury to ascertain outcome (ie, subsequent head injury) and to provide information on potential risk factors (age, gender, chronic medical condition, activity level, and socioeconomic status).

MAIN OUTCOME MEASURE. The outcome of interest was a head injury requiring medical attention within the following year ascertained by parental recall or physician claims data.

RESULTS. A total of 245 and 386 previously head-injured children sustained a subsequent head injury within 6 and 12 months, respectively. Children who sought care for an initial head injury (n = 3599) were at higher risk of having a subsequent head injury within 6 months than children who sought care for an injury not to the head (n = 6716). The adjusted odds ratio suggested weak confounding by age, gender, and history of previous head injury. Results were consistent on the basis of physician claims data and 12-month follow-up interview data.

CONCLUSIONS. These results provide evidence that having a head injury increases a child's risk of having a subsequent head injury. Although age, gender, and history of previous head injury confound the relationship, the effect remains substantial.

Key Words: children • recurrence • injury • head trauma • rehabilitation

Abbreviations: HI—head injury • ED—emergency department • CHIRPP—Canadian Hospital Injury Research and Prevention Program • RAMQ—Régie de l'assurance-maladie du Québec • SES—socioeconomic status • CI—confidence interval • OR—odds ratio • aOR—adjusted odds ratio

Circumstantial evidence suggests that among children a previous head injury (HI) increases the risk of subsequent HI, the consequences of which may be serious. This study provides evidence that having an HI is associated with an increased risk of having a subsequent HI.

HI is an important public health problem among children and youth. Schneier et al1 estimated that there were 50658 hospital admissions for HI among children ≤17 years of age in the United States in 2000. Others have reported that HI accounts for >10% of all emergency department (ED) visits.2 Between 1994 and 2000, 24% of injuries to Canadian children (≤12 months old) for whom care was sought at an ED involved the head.3

Most HIs in children are mild, that is, the children are able to return to their previous activities.4 Some, however, are serious, causing an extended period of unconsciousness and cognitive, psychosocial, or physical sequelae.5–7 Furthermore, some researchers believe 1 HI may increase the risk of a subsequent HI, although there is still no strong evidence to support this view.8–10 Studies with children have been retrospective and methodologically flawed, usually because they lacked adequate comparison groups. Also, injury reoccurrence has typically been ascertained solely on the basis of nonstandardized questionnaires. Finally, only anecdotal evidence suggests that an athlete is more prone to future concussive injury after having sustained a sport-related concussion.11 Here, the likelihood of repeat injury may simply reflect the amount of playing time or a player's style of play (ie, using dangerous game strategies) rather than any inherent risk.12

However, the evidence from animal and human studies concerning the potential consequences of repeated concussion is conflicting.13–17 Given the possible consequences of subsequent HI among children, this issue warrants additional investigation. The objective of this study was, therefore, to determine whether previous HI is a risk factor for subsequent HI among children. Secondary objectives were to (1) examine the effect of other factors (eg, age, gender, socioeconomic status, preexisting disabling conditions, comorbidity, and injury severity) on subsequent HI and (2) estimate the time to subsequent HI among injured children.


   METHODS
TOP
 ABSTRACT
 METHODS
RESULTS
 DISCUSSION
 STUDY LIMITATIONS
 CONCLUSIONS
 REFERENCES
 
Participants
Participants in this longitudinal cohort study were children aged 1 to 18 years who sought care for an injury at the ED of either of the 2 pediatric hospitals in Montreal from December 2000 to March 2003. Both hospitals are designated provincial neurotrauma centers that serve most head-injured children and youth in Montreal. Each year, ~30000 children seek care for an injury at the ED of these centers, and HI accounts for >3000 visits.

Children were recruited by using data routinely collected through an ED-based injury surveillance system, the Canadian Hospital Injury Research and Prevention Program (CHIRPP).18 The CHIRPP database contains information about the circumstances of the injury provided by the patient or an accompanying adult on a 1-page self-administered questionnaire. On the back of the same form, diagnostic codes, injured body part, and level of treatment provided are recorded. The child's unique health insurance number, date of birth, gender, postal code, and date of visit are also recorded.

The institutional review boards at both hospitals approved this study. Parents gave verbal consent to participate in the interviews, as well as approval to be contacted on the forms. Permission to access the physician claims database was obtained from the Quebec Commission for Access to Information.

Two groups of children were studied. The exposed group included all children seeking care for an HI during the study period. This group was defined by using a diagnosis recorded on the CHIRPP form: minor HI, skull fracture, intracranial injury, or concussion. Also included were children with injuries to the eyes, facial fractures, dental injuries, or facial lacerations if the mechanism of injury involved being struck forcefully against a hard surface, a fall from a height, or both. Athletes who sustain a fractured mandible or maxilla almost always sustain a coexistent concussion19 because the forces required to fracture bones often exceed the impact threshold to cause a mild HI or concussion.20 This broad definition of HI was used to include children with definite HI and those with probable HI. The CHIRPP form does not contain information on amnesia or loss of consciousness or the Glasgow Coma scale, making it difficult to obtain a more precise diagnosis, including the presence or absence of brain injury. This definition was used by others in earlier descriptive studies on repetitive HI among children.9,10

Two unexposed (without an HI) children were selected randomly to serve as controls for each exposed child.21 These were children seeking care for a musculoskeletal injury to the extremities during the same period. Children with at least 1 of the following diagnoses recorded on the CHIRPP form were recruited: fracture, laceration, sprain, and soft tissue injury to an upper or lower extremity. The choice of this comparison group (ie, injured children as opposed to those with medical disorders) was made to control for potential confounders, such as exposure to hazards and children's developmental abilities.

CHIRPP data were entered into the study database each week. Only children whose parents indicated on the CHIRPP form their consent to receive a follow-up telephone call were recruited. The child's first visit to the ED during the study period was identified as the index visit.

Main Outcome Measure
The outcome of interest was ≥1 medical care visits for a subsequent HI at 6 and 12 months after the index visit. It was measured separately by using 2 data sources: parental report and physician claims data.

Parent Report
During standardized telephone interviews with established interrater and intrarater reliability,22 1 of 2 trained individuals asked parents to respond to the following question: Since the injury in (month of index injury or 6 months ago), has (name of child) had another injury that required medical attention? In the case of a repeat injury, parents were asked about the type and nature of injury, when it occurred, whether they considered it to be "serious" (eg, a concussion or skull fracture), and whether and for how long the child was hospitalized. In addition, they were asked about the reason(s) that prompted them to seek care for the subsequent HI (eg, it was serious or because you were worried about what you were told during the last visit and wanted to be sure that everything was ok). Parents also provided details about discharge recommendations regarding activity restrictions (for the index injury), usual level of activity of their child compared with others of the same age and gender, and other child-, injury-, and parent-related variables (Table 1). Answers were recorded on digitized response forms to maximize data quality.


View this table:
[in this window]
[in a new window]

  		TABLE 1 Baseline Characteristics of Respondents in the 6-Month Telephone Interview According to Exposure Status

 
The choice of questionnaire items was based on pediatric injury literature and on clinical experience with the patient population.23–26 The reliability of the questionnaire and the scoring of parental responses was investigated with 42 parents of injured children (33% with an HI). Parents participated in 2 telephone interviews separated by a 2-week interval; the first interviews were tape recorded to be reassessed 4 weeks later. {kappa} values exceeding 0.75, indicating "substantial" reliability27 were found for the majority of the questions, including the one about subsequent HI.

Physician Claims
Provincial health records for all 11867 children (exposed and unexposed, including those not participating in telephone interviews) were accessed from the database of the provincial health insurance board of Quebec, Canada (Régie de l'assurance-maladie du Québec [RAMQ]). These data were acquired through linkage by using the health insurance number on the CHIRPP form. The RAMQ database contains diagnostic codes of all insurable health care services provided to eligible Quebec residents and is used to reimburse fee-for-service providers (eg, physicians).

A file of CHIRPP-based data (each child's index date, diagnostic group, and RAMQ number) was sent to the RAMQ where health insurance numbers were linked to subsequent medical care visits for each child over the 12 months after the index visit. The returned file contained the children's age, gender, the first 3 digits of the postal code of residence, and the complete record of services paid to fee-for-service physicians who provided care during the study period. The postal code data served as a proxy for socioeconomic status (SES) by giving the median household income in the child's area of residence. The following variables were included for each medical visit: diagnostic codes, medical procedures and their costs, type of physician, and facility type. Confidentiality was maintained by using a scrambled personal insurance number. Consequently, we were unable to link children's RAMQ data with those based on the parent report; therefore, the physician claims data were analyzed separately.

Ascertaining subsequent HI using the physician claims data was based on the results of earlier work examining the concordance between diagnoses recorded in the injury surveillance system (CHIRPP) and those recorded in the physician claims for 3049 injured children.28 Briefly, we demonstrated that using a combination of diagnostic and procedure codes in the RAMQ physician claims database was a valid method of estimating HI occurrence among children. Substantial concordance (weighted {kappa}: 0.66; 95% confidence interval [CI]: 0.63–0.69) was found between the 2 data sources. The sensitivity of diagnostic and procedure codes in the RAMQ database for identifying HI and musculoskeletal injury were 0.61 (95% CI: 0.57–0.64) and 0.97 (95% CI: 0.96–0.98), respectively, whereas the specificities for identifying HI and musculoskeletal injury were 0.97 (95% CI: 0.96–0.98) and 0.58 (95% CI: 0.56–0.63), respectively.

The data sources for the other potential predictors of subsequent HI included parent responses to the telephone interview, physician claims, and the CHIRPP records.

Statistical Analyses
Exposed and unexposed groups (and respondents and nonrespondents) were compared with regard to categorical and continuous variables using {chi}2 and t tests, respectively. Bivariate analyses were performed to explore factors associated with subsequent HI. Child-related factors included age, gender, parental perception of child's level of physical activity compared with children of same age and gender (more active, about the same, or less active than other children), chronic medical condition (yes/no), general health (excellent, good, or poor), and history of previous HI in the year preceding the index injury (yes/no). Those related to the child's family included family structure (single parent, yes/no; oldest child, yes/no; mother's age in categories), maternal education (highest level education completed), parental perception of being anxious (yes/no), and median household income (a proxy for SES). Injury-related factors included injury type (ie, exposure group), involvement in a recreational activity at time of injury (yes/no), receipt of discharge instructions about activity restrictions (yes/no), place of injury (home, daycare, or elsewhere), and hospitalization for index injury (yes/no).

Using the Mantel-Haenszel method,29 crude and adjusted relative risks and 95% CIs were calculated for exposed and unexposed groups. Multivariable analysis included fitting 2 logistic regression models to evaluate factors associated with risk of subsequent HI while adjusting for potential confounders: 1 used parental responses to questionnaires and the other used physician claims data. Potential confounders (covariates) included child-, parent-, and index injury-related variables. Their selection was data driven but based on substantive knowledge where feasible.

Additional exploratory logistic regressions were calculated to examine the possibility of a dose-response and the sensitivity of the result to errors in classification. These analyses determined whether the risk of subsequent HI could vary according to the type/severity of the index HI (ie, definite versus probable HI) or according to the type of the subsequent HI (ie, definite versus probable HI). In the context of this study, definite HI referred to a concussion/mild HI, skull fracture or intracranial injury to the head (ie, high degree of certainty of sustaining a genuine HI). Probable HI included injuries to the eyes, facial fractures, dental injuries, or facial lacerations if the mechanism of injury involved being struck forcefully against a hard surface, a fall from a height, or both. These analyses were conducted by using interview and physician claims data from the 12-month follow-up to provide sufficient numbers of children with subsequent HI. Goodness of fit statistics were calculated for all models.

Finally, a Kaplan-Meier survival analysis30 was performed for each exposure group to determine the probability of subsequent HI over time using each data source. Cox regression was also performed with the 2 groups to estimate the (hazard) risk ratio of subsequent HI at any given time.31 Data were analyzed using SAS 8.2 (SAS Institute, Cary, NC).

A power calculation, based on previously observed rates of subsequent HI (exposed 2.5%, unexposed 1.0% per year), determined that a minimum of 9000 children (6000 unexposed, 3000 exposed) were required at the 12-month follow-up to detect a relative risk of ≥2.5 ({alpha}: .05) with at least 90% power using logistic regression.21


   RESULTS
TOP
 ABSTRACT
 METHODS
 RESULTS
DISCUSSION
 STUDY LIMITATIONS
 CONCLUSIONS
 REFERENCES
 
CHIRPP records for 11867 children were obtained. Response rates of 87% and 80% were attained for the 6- and 12-month interviews, respectively, and 90% of parents participating in the 6-months interview participated at 12 months. The main reason for loss to follow-up was the inability to obtain an accurate family telephone number (63% of losses). Only 278 and 19 parents refused to participate in the 6- and 12-month interviews, respectively. Physician claims data were unattainable for 176 children because of missing RAMQ numbers, thus providing RAMQ data for a total of 11691 children. A participant flow diagram describes the records included in the various analyses (Fig 1).


Figure 1
View larger version (28K):
[in this window]
[in a new window]

  		FIGURE 1 Participant flow diagram. MSK indicates musculoskeletal.

 
The mean age of respondents in the 6- and 12-month follow-up interviews was slightly higher (but within 6 months) than that of nonrespondents. Respondents and nonrespondents had similar proportions of anglophone and francophone parents and of male and female children. Exposed and unexposed children among participants in the 6- and 12-month interviews differed with respect to age (9.5 years for unexposed vs 6.5 years for exposed; P < .001). The child's mother was the most frequent respondent.

Within 6 months after the index injury, 245 (2.4%) of the 10315 children were reported to have had a subsequent HI requiring medical attention. Among those with an index HI, the percentage of children with a second HI was 3.2%. Within 12 months, 386 children (4.1%) had had an HI, 12 of whom sustained a subsequent HI within the first 6-month interval. Among those with an index HI, the rate of subsequent HI at 12 months was 5.5%. The children's characteristics are presented in Table 1 according to exposure status. Only data for respondents of the 6-month interview are presented; the 12-month data are almost identical (ie, within 1–2 percentage points).

The results of the bivariate analyses regarding the predictors of subsequent HI within 6 and 12 months are presented in Table 2. Children who sought care for an HI (n = 3599) were at higher risk of having a subsequent HI at 6 months than were children who sought care for a non-HI (n = 6716; odds ratio (OR) at 6 months: 1.7; 95% CI: 1.3–2.2). Similar results were found at 12 months. Other predictors included having sustained the HI during a recreational activity (6 months only), injury sustained at daycare, child's age, being male, having had a previous HI, and being less or equally active as a child of the same age and gender (6 months only).


View this table:
[in this window]
[in a new window]

  		TABLE 2 Injury-, Child-, and Parent-Related Predictors of Subsequent HI Within 6 and 12 Months: Bivariate Analysis

 
Table 3 presents the adjusted ORs (aORs) for subsequent HI at 6 and 12 months using both data sources. At 6 months, the aOR (parent interviews) was 1.6 (95% CI: 1.2–2.0), suggesting some confounding by age, gender, and history of previous HI. Results were generally consistent at the 12-month follow-up (aOR: 1.5; 95% CI: 1.2–1.9) and were comparable to those based on analysis of physician claims data for services received during the following 12 months. (aOR: 1.7; 95% CI: 1.4–2.0).


View this table:
[in this window]
[in a new window]

  		TABLE 3 Logistic Regression Models: Association Between Injury-, Child-, and Parent-Related Predictors and Subsequent HI Within 6 and 12 Months After Injury

 
The exploratory subgroup analyses (ie, dose effect) using data from both sources yielded the following results (Table 4): Children who sought care for a definite HI were at higher risk of having a subsequent definite HI than children who sought care for a non-HI, (parental report: aOR: 2.1; 95% CI: 1.5–2.9; physician claims: aOR: 2.6; 95% CI: 1.9–3.6). In general, it seems that having a definite HI compared with an injury not to the head increases a child's risk of any type of subsequent HI within 12 months. Children and youth who had a definite HI are even more likely to have a definite subsequent HI, regardless of their age. Goodness-of-fit statistics were nonsignificant for all models (P > .3).


View this table:
[in this window]
[in a new window]

  		TABLE 4 Exploratory Regression Models of the Association Between Injury- and Child-Related Predictors and Subsequent Definite and Probable HI Within 12 Months After Injury According to the Type of Index HI

 
Finally, among those reporting a subsequent HI (all categories) within 6 months, 72% of parents considered the subsequent HI "minor," whereas 23% considered it "serious." Approximately 42% of parents reported the subsequent injury to the head as being an HI or concussion, with the remainder reporting open wounds or "bumps to the head." Parents reported consulting medical care either because they thought the HI was serious (46%), were worried (51%), or for other reasons (8%). Only 8 children spent ≥1 night in hospital. Finally, among parents who remembered the exact date of the subsequent injury, 65% reported it as occurring within the fourth and sixth month after the index visit, and 20% reported it occurring within the first month. Similar results were found on the basis of the analyses of the responses from parents of children who had sustained a subsequent HI within 12 months.

The estimated time (in months) to subsequent HI shows the differential probability of such an event over the 12-month period among exposed and unexposed children, accounting for time and for censoring because of loss to follow-up. Figures 2 and 3 demonstrate (by using interview and RAMQ data) that the survivor function for the exposed (HI) group consistently lies below that for the unexposed group (at all points of follow-up). Cox regression models calculated with interview data (hazard ratio: 1.5; 95% CI:1.2–1.9) and RAMQ data (hazard ratio: 1.7; 95% CI: 1.4–2.2) were consistent with the results of the logistic regression models.


Figure 2
View larger version (10K):
[in this window]
[in a new window]

  		FIGURE 2 Estimated survival function of time to subsequent HI based on parental report (n = 10315).

 

Figure 3
View larger version (10K):
[in this window]
[in a new window]

  		FIGURE 3 Estimated survival function of time to subsequent HI based on physician claims data (n = 11691).

 

   DISCUSSION
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
STUDY LIMITATIONS
 CONCLUSIONS
 REFERENCES
 
To our knowledge, this is the first study of its kind using 2 data sources for a large cohort of injured children to provide evidence that having an HI is associated with an increased risk of having a subsequent HI. Our results further indicate that age, gender, and history of previous HI confound the relationship. Age and gender are well-established determinants of injury among children; boys have more injuries than girls and risk of injury typically increases with age.32–36 Children with a history of previous HI demonstrated more than a twofold risk for subsequent HI, indicating that some children seem to be predisposed to having recurrent injuries similar to that of the original injury. In a study where HI was the most frequent injury sustained by the children involved, the number of injuries before age 5 years was found to be the best predictor of injuries reported between 5 and 10 years later.37

Because the cohort consisted of children and youth with different types of HI, we were able to explore the possibility of a dose-response relationship. Clearly, having had a concussion, skull fracture, or intracranial injury compared with a musculoskeletal injury is associated with an increased risk of subsequent HI (any type) within 12 months. These children and youth are even more likely to have a definite subsequent HI, regardless of their age. The association between probable HI and subsequent HI is, however, less clear, and no definitive conclusions can be made about the issue of age within this context.

The study results were consistent using both data sources. The risk of subsequent HI calculated with the RAMQ data were somewhat higher than that calculated with the interview data, probably because of the absence in the model of a variable for "previous HI." This variable was not available from the physician claims data. These findings are consistent with exploratory work based on injury surveillance data alone (ie, CHIRPP) examining repeat injuries at 1 hospital.38 They are not in agreement, however, with those of an early report of equal rates of previous HI among children with an HI compared with school-based controls.10 Our selection of the control group (injured-based controls) may explain the inconsistency. The observed rates of subsequent HI among children with an HI of 3.2% and 5.5% within the following 6 and 12 months., respectively, however lie within the range of previous descriptive studies (1.6%–23%).8,10

Despite asking parents questions about a broad range of topics to examine the factors associated with subsequent HI, this study was unable to further delineate specific characteristics of children with repeated HIs. The mechanisms related to subsequent HI remain unclear and suggest that the phenomenon is likely to be multifactorial involving a complex interaction between children and their personal and social environments.

In terms of human factors, it is reasonable that children with an HI who do not regain their preinjury state of health could be at risk for another injury. An HI (even mild HI) can result in a wide range of physical, cognitive, and behavioral sequelae.4,7,39–42 Even subtle deficits in coordination, balance, or endurance, combined with cognitive and behavioral limitations, might reduce a child's ability to meet the demands of a difficult task, thus increasing a child's risk for reinjury. In terms of environmental factors, children who sustain an HI may be more exposed to hazardous environments than other injured children.23 Children with HI might participate more in contact sports or have lower SES status (eg, less parental supervision, less use of prevention strategies) that puts them at greater risk for HI. This study was not able to support this notion.

The observed increased risk of subsequent HI may simply be a reflection of parents' health-seeking behavior, that is, a lower threshold for bringing younger children to medical attention,43 or of higher rates of service use among parents of children with HI. With regards to the former, information received at the ED about an HI and when to call or return to the hospital44,45 may have sensitized parents to the potential seriousness of an HI. This may prompt parents to seek care for another HI for which they may not otherwise have sought care. In an attempt to address this issue, parents were asked whether they considered themselves to be anxious. Such parents may have greater difficulty interpreting symptoms associated with a mild HI and worry more, making them more likely to seek care for another HI. Self-reported parental anxiety, however, was not found to be a predictor of subsequent HI.

One could also argue that the results reflect the increased chances of parents of a child who had an HI to remember that the child had a subsequent one, which is a reporting issue. This is also unlikely because the results were corroborated with physician claims data that are free of recall bias.

Finally, one cannot rule out the possibility that other factors, not measured in our study, may be responsible for the observed association. More research is needed to understand the causal pathways that lead to the increased risk of a second HI. In-depth interviews are currently underway with the parents of the children with repeated HIs in an attempt to clarify the complex interactions between the child and his personal and social environments.

Several possible clinical implications arise from this study. The actual number of children with a subsequent HI was small. Nonetheless, it is important to prevent these injuries in children, particularly given that so little is known about their cumulative effects. On the basis of these findings, health care providers should be able to better identify the children who are at highest risk of subsequent HI. Individual-focused prevention interventions (including parental and child counseling on injury risk) may be beneficial for this group of children. Considerable thought needs to be given to the specific content of these interventions, because most current prevention strategies, aimed in part at reducing the risk of a subsequent HI (or the duration of post concussion symptoms),46 do not seem to have been successful. In fact, our study found that children who received instructions about activity restrictions tended to have an increased risk. Moreover, given that the majority of subsequent HIs were found to occur 5 or 6 months after the initial HI, a 4-week restriction of activity may need to be reexamined.


   STUDY LIMITATIONS
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 STUDY LIMITATIONS
CONCLUSIONS
 REFERENCES
 
Neither data source we used is perfect, and it is possible that misclassification of injury exposures and of subsequent HI occurred. For example, those with multiple injuries, of which at least 1 was an HI, were classified into the HI group. The method of detection of outcome by parent report and by using physician claims data did not, however, differ by exposure status. There were adequate data to describe the baseline state of the cohort, and measurable differences between the groups were controlled for at the analysis stage. Sampling bias may also have been a limitation. To recruit a representative sample from the population, all children seeking care for an HI at the 2 centers were identified and highly acceptable response rates were obtained in both groups and at both follow-up times. Some attrition was inevitable, but the losses did not occur differentially by exposure and outcome.


   CONCLUSIONS
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 STUDY LIMITATIONS
 CONCLUSIONS
REFERENCES
 
This study provides strong evidence that having an HI increases a child's risk of having a subsequent HI. Although age, gender, and history of previous HI confound the phenomenon, when considering all types of HI, the effect remains substantial. Children who sustain a concussion, skull fracture, or intracranial injury are more than twice as likely to have a subsequent HI of similar type within 12 months than are children seeking care for an injury not to the head, regardless of their age.


   ACKNOWLEDGMENTS
 
This study was funded by the Canadian Institutes of Health Research. Dr Swaine is supported by a scientist award from the Fonds de recherché en santé du Québec.


   FOOTNOTES
 
Accepted Dec 13, 2006.

Address correspondence to Bonnie R. Swaine, PhD, École de réadaptation, Faculté de Médecine, Université de Montréal, C.P. 6128 Succursale Centre-ville, Montréal, Québec, Canada H3C 3J7. E-mail: bonnie.swaine{at}umontreal.ca

These results were presented, in part, at the 6th World Congress on Brain Injury; May 6–8, 2005; Melbourne, Australia.

The authors have indicated they have no financial relationships relevant to this article to disclose.


   REFERENCES
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 STUDY LIMITATIONS
 CONCLUSIONS
 REFERENCES
 

  1. Schneier AJ, Shields BJ, Hostetler SG, Xiang H, Smith GA. Incendence of pediatric traumatic brain injury and associated hospital resource utilization in the United States. Pediatrics. 2006;118 :483 –492[Abstract/Free Full Text]
  2. Yeates KO, Luria J, Bartkowski H, Rusin J, Martin L, Bigler ED. Postconcussive symptoms in children with mild closed head injuries. J Head Trauma Rehabil. 1999;14 :337 –350[Web of Science][Medline]
  3. Pickett W, Streight S, Simpson K, Brison RJ. Injuries experienced by infant children: a population-based epidemiological analysis. Pediatrics. 2003;111(4 Pt 1) . Available at: www.pediatrics.org/cgi/content/full/111/4/e365
  4. Bijur P, Haslum M, Golding J. Cognitive and behavioral sequelae of mild head injury in children. Pediatrics. 1990;86 :337 –344[Abstract/Free Full Text]
  5. Klonoff H, Clark C, Klonoff P. Long-term outcome of head injuries: a 23 year follow-up study of children with head injuries. J Neurol Neurosurg Psychiatry. 1993;56 :410 –415[Abstract/Free Full Text]
  6. Jaffe K, Massagli T, Martin K, Rivara J, Fay G, Polissar N. Pediatric traumatic brain injury: acute and rehabilitation costs. Arch Phys Med Rehabil. 1993;74 :681 –686[CrossRef][Web of Science][Medline]
  7. Jaffe K, Polissar N, Fay G, Liao S. Recovery trends over three years following paediatric traumatic brain injury. Arch Phys Med Rehabil. 1995;76 :17 –26[CrossRef][Web of Science][Medline]
  8. Annegers J, Grabow J, Kurland L, Laws E. The incidence, causes, and secular trends of head trauma in Olmsted County, Minnesota, 1935–1974. Neurology. 1980;30 :912 –919[Abstract/Free Full Text]
  9. Partington M. The importance of accident-proneness in the etiology of head injuries in children. Arch Dis Child. 1960;35 :215 –223[Medline]
  10. Klonoff H. Head injuries in children: predisposing factors accident conditions, accident proneness and sequelae. Am J Public Health. 1971;61 :2405 –2417[Free Full Text]
  11. McCrory P. The eighth wonder of the world: the mythology of concussion management. Br J Sports Med. 1999;33 :136 –137[Free Full Text]
  12. Gerberich S, Priest J, Boen J, Staub C, Maxwell R. Concussion incidence and severity in secondary school varsity football players. Am J Public Health. 1983;73 :1370 –1375[Abstract/Free Full Text]
  13. Gronwall D, Wrightson P. Cumulative effect of concussion. Lancet. 1975;22 :995 –997
  14. Jordon B, Zimmerman R. Computed tomography and magnetic imaging comparisons in boxers. JAMA. 1990;263 :1670 –1674[Abstract/Free Full Text]
  15. Saunders R, Harbaugh R. Second impact in catastrophic contact-sports head trauma. JAMA. 1984;252 :538 –539[Abstract/Free Full Text]
  16. McCrory P, Berkovic S. Second impact syndrome. Neurology. 1998;50 :677 –683[Abstract/Free Full Text]
  17. Cantu R, Voy R. Second impact syndrome: a risk in any contact sport. Phys Sports Med. 1995;23 :27 –34
  18. Mackenzie SG, Pless IB. CHIRPP: Canada's principal injury surveillance program: Canadian hospitals injury reporting and prevention program. Injury Prevention. 1999;5 :208 –213[Free Full Text]
  19. McCrory P. Were you knocked out? A team physician's approach to initial concussion management. Med Sci Sports Exerc. 1997:S207 –S211
  20. Johnston K, McCory P, Mohtadi G, Meeuwise W. Evidence based reviews of sport-related concussion: clinical science. Clin J Sports Med. 2001;11 :150 –160[CrossRef][Web of Science][Medline]
  21. Breslow N, Day N. Statistical methods in cancer research. In: The Design and Analysis of Cohort Studies. Vol II. Lyon, France: International Agency for Research on Cancer; 1987
  22. Swaine B, Pless IB, Grimard G, Platt RW, Tremblay C, Peeters C. Development and reliability of a telephone questionnaire to determine whether children who have sustained a head injury are at risk for subsequent head injury. Presented at: 6th World Conference Injury Prevention and Control, May 12–15, 2002; Montréal, Quebec, Canada
  23. Rivara F. Epidemiology and prevention of pediatric traumatic brain injury. Pediatr Ann. 1994;23 :12 –17[Web of Science][Medline]
  24. Bijur P, Golding J, Haslum M, Kurzon M. Behavioral predictors of injury in school-age children. Am J Dis Child. 1988;142 :1307 –1312[Abstract/Free Full Text]
  25. Dowd M, Langley J, Koepsell T, Soderberg R, Rivara F. Hospitalizations for injury in New Zealand: prior injury as a risk Factor for assaultive injury. Am J Public Health. 1996;86 :929 –934[Abstract/Free Full Text]
  26. Eminson C, Jones H, Goldacre M. Repetition of accidents in young children. J Epidemiol Commun Health. 1986;40 :170 –173[Abstract/Free Full Text]
  27. Landis J, Koch G. The measurement of observer agreement for categorical data. Biometrics. 1977;33 :159 –174[CrossRef][Web of Science][Medline]
  28. Kostylova A, Swaine B, Feldman D. Concordance between childhood injury diagnoses from two sources: an injury surveillance system and a physician billing claims database. Inj Prev. 2005;11 :186 –190[Abstract/Free Full Text]
  29. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst. 1959;22 :719 –748[Web of Science][Medline]
  30. Kaplan E, Meier P. Nonparametric estimation for incomplete observations. J Am Stat Assoc. 1958;53 :457 –481[CrossRef][Web of Science]
  31. Cox D. Regression models and life-tables (with discussion). J Royal Stat Soc Series B. 1972;34 :187 –220
  32. Conn JM, Annest JL, Gilchrist J. Sports and recreation related injury episodes in the US population, 1997–99. Inj Prev. 2003;9 :117 –123[Abstract/Free Full Text]
  33. Scheidt P, Harel Y, Trumble A, Jones D, Overpeck M, Bijur P. The epidemiology of nonfatal injuries among US children and youth. Am J Public Health. 1995;85 :932 –938[Abstract/Free Full Text]
  34. Pless IB, Millar W. Unintentional Injuries in Childhood: Results From Canadian Health Surveys. Ottawa, Ontario, Canada: Health Canada; 2000
  35. Brudvik C. Child injuries in Bergen, Norway. Injury. 2000;31 :761 –767[CrossRef][Web of Science][Medline]
  36. Rivara FP. Epidemiology of childhood injuries. I. Review of current research and presentation of conceptual framework. Am J Dis Child. 1982;136 :399 –405
  37. Bijur P, Golding J, Haslum M. Persistence of occurrence of injury: can injuries of preschool children predict injuries of school-aged children? Pediatrics. 1988;82 :707 –712[Abstract/Free Full Text]
  38. Swaine BR, Yang H, Pless B, Platt R. Do children with a head injury have an increased risk of subsequent injury? Presented at: Third World Congress on Brain Injury; June 12–17, 1999; Quebec City, Quebec, Canada
  39. Casey R, Ludwig M, McCormick. Morbidity following minor head trauma in children. Pediatrics. 1986;78 :497 –502[Abstract/Free Full Text]
  40. Swaine R, Pless I, Freidman D, Montes J. A preliminary investigation of the effectiveness of a head injury program provided in a pediatric trauma center. Am J Phys Med Rehabil. 2000;79:412–420
  41. Asarnow R, Satz P, Light R, Lewis R, Neumann E. Behavior problems and adaptive functioning in children with mild and severe closed head injury. J Pediatr Psychol. 1991;16 :543 –555[Abstract/Free Full Text]
  42. Gagnon I, Swaine BR, Friedman D, Forget R. Children demonstrate decreased dynamic balance following a mild traumatic brain injury. Arch Phys Med Rehabil. 2004;85 :444 –452[CrossRef][Web of Science][Medline]
  43. Cernan J, Shapiro E, Kriel R. Neurologic and neuropsychological aspects of minor head injury in children. In: Rizzo M, Tranel D, eds. Head Injury and Postconcussive Syndrome. New York, NY: Churchill Livingston; 1996:441–456
  44. Warren D, Kisson N. Usefulness of head injury instruction forms in home observation of mild head injuries. Pediatr Emerg Care. 1989;5 :83 –85[Medline]
  45. King N, Wenden F, Wade DT. Interventions and service needs following mild and moderate head injury: the Oxford Head Injury Service. Clin Rehabil. 1997;11 :13 –27[Abstract/Free Full Text]
  46. Swaine BR, Friedman D. Activity restrictions as part of the discharge management for children with a head injury. J Head Trauma Rehabil. 2001;16 :292 –301[Web of Science][Medline]
PEDIATRICS (ISSN 1098-4275). ©2007 by the American Academy of Pediatrics

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Facebook Facebook   Add to Reddit Reddit   Add to Technorati Technorati   Add to Twitter Twitter    What's this?


This article has been cited by other articles:


Home page
 TraumaHome page
J. Bayreuther and I. Maconochie
The evidenced-based care behind the early management of head injured children
Trauma, April 1, 2008; 10(2): 85 - 92.
[Abstract] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Swaine, B. R.
Right arrow Articles by Pless, I. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Swaine, B. R.
Right arrow Articles by Pless, I. B.
Related Collections
Right arrow Office Practice
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Facebook   Add to Reddit   Add to Technorati   Add to Twitter  
What's this?