OBJECTIVE: The Standardized Assessment of Concussion (SAC) is a validated tool for identifying the effects of mild traumatic brain injury (mTBI). Previous research focused on sport-related sideline evaluation of adolescents and adults. Our goal was to evaluate performance of the SAC among subjects with and without head injury in a pediatric emergency department (ED).
METHODS: This was an observational study of children 6 to 18 years of age who presented to an ED with blunt head injury (case-patients) or minor extremity injury (controls). SAC and graded-symptom-checklist scores were compared. American Academy of Neurology concussion grades, presence of loss of consciousness and posttraumatic amnesia were also compared with SAC and graded-symptom-checklist scores among case-patients.
RESULTS: Three hundred forty-eight children were enrolled. SAC scores trended lower (greater cognitive deficits) for case-patients compared with controls but did not reach significance. Graded-symptom-checklist scores were significantly higher among case-patients. Presence of altered mental status magnified this effect. There was no correlation between SAC scores and other indicators of mTBI. There was a positive correlation between graded-symptom-checklist scores and posttraumatic amnesia and American Academy of Neurology concussion grade.
CONCLUSIONS: The graded symptom checklist reliably identified mTBI symptoms for all children aged 6 years and older. SAC scores tended to be lower for case-patients compared with controls but did not reach significance. Patients with altered mental status at the time of injury manifest an increased number and severity of symptoms. Additional research into strategies to identify cognitive deficits related to mTBI and classify mTBI severity in children is needed.
WHAT'S KNOWN ON THIS SUBJECT:
Mild traumatic brain injury (mTBI) is common among children, but there are no evidence-based tools that systematically quantify cognitive deficits and symptom severity for those in this age group. Better instruments to quantify severity and monitor recovery are needed.
WHAT THIS STUDY ADDS:
Results of this study reveal that the graded symptom checklist in the SAC systematically quantifies the severity of symptoms of mTBI. They also reveal the interplay between somatic complaints and other measures of severity.
Blunt head trauma is common among children and accounts for nearly 600 000 emergency department (ED) visits annually. Most of these patients suffer mild traumatic brain injuries (mTBIs), also known as concussions.1,2 Although most children will recover from mTBI, they may initially display subtle symptoms and cognitive changes not identified by informal orientation questions.3,4 Concussion-grading scales (eg, American Academy of Neurology [AAN])5 are not designed to detect these subtle changes; they ignore many common mTBI symptoms and focus almost exclusively on posttraumatic amnesia and loss of consciousness (LOC). Their relationship to cognitive deficits is unclear. To date, no concussion-grading scale has been prospectively evaluated.6 The Glasgow Coma Scale (GCS), a widely used measure of TBI, provides no assessment of symptoms or cognitive deficits.
Identifying effects of mTBI soon after injury has at least 2 benefits. Patients educated on expected postconcussive symptoms report fewer symptoms than those who do not receive this education.7 The symptoms of mTBI are nonspecific and mimic other common pediatric diseases such as attention-deficit/hyperactivity disorder, depression, and learning disabilities.8 Providing the primary care physician with objective evidence of mTBI sequelae should help minimize prescribing unnecessary medication and allow for more appropriate referral. As the number of children admitted for mTBI declines,9 the responsibility of managing mTBI sequelae will increasingly fall to primary care providers. An instrument easily applied in a busy practice setting that identifies and quantifies symptoms and cognitive deficits and objectively demonstrates their resolution is critical for improving the care of patients with mTBI.
Currently, there is no validated screening instrument for identifying cognitive deficits attributable to mTBI in pediatric patients that could readily be applied in the ED or office. The Standardized Assessment of Concussion (SAC) is a brief sideline screening tool that requires no formal neuropsychological training.10 It focuses on the cognitive domains and symptoms most commonly associated with mTBI.11 It has been shown to be effective in detecting cognitive deficits in older teenagers and young adults suffering sports-related concussion; it has not been well studied in younger children.10,12,–,14 One adult ED study demonstrated the sensitivity of the SAC to the acute cognitive changes associated with mTBI.15 Recently, the authors of the SAC highlighted the utility of systematic evaluation of mTBI by using a concussion-symptom inventory for initial diagnosis of and monitoring recovery from mTBI.16
Our primary goal was to identify a more informative, quickly administered evidence-based tool that could be adapted for use in the pediatric ED. Our secondary goal was to investigate what, if any, relationship existed between the AAN concussion grade, the GCS, and the SAC. The primary outcome was comparison of SAC scores and symptom severity in head-injured case-patients versus non–head-injured controls. The secondary outcomes were comparison of SAC and graded-symptom-checklist scores to traditional indicators of mTBI severity, such as LOC, posttraumatic amnesia, concussion grades, and GCS among case-patients. We hypothesized that children with head injury would demonstrate lower SAC scores and greater mTBI symptoms than children with minor extremity injuries.
PATIENTS AND METHODS
We conducted an observational cohort study in the ED of a regional pediatric trauma center with an annual volume of 47 000 visits. We prospectively enrolled patients aged 6 to 18 years. Children were enrolled as case-patients if they had suffered blunt head trauma in the previous 24 hours and had an initial GCS of ≥13. We categorized case-patients as those with and without unambiguous evidence of altered mental status (AMS), which we defined as (1) any history of LOC, (2) any history of posttraumatic amnesia, or (3) a GCS of ≤14. Children were enrolled as controls if they had suffered minor extremity trauma (contusions, sprains, strains, minor nondisplaced fractures) without concomitant head injury.
Patients were excluded for any of the following: open head injury, intoxication, use of narcotics for pain control before the administration of the SAC, suspicion of nonaccidental trauma, admission to the neurosurgical service or ICU, multisystem injuries, or the presence of preexisting central nervous system abnormalities. Non–English-speaking patients were excluded, because the SAC has only been validated in an English-speaking population. English-speaking children of Spanish-speaking parents were included; consent was obtained from the parents in Spanish. Patients were enrolled by the principal investigator (J.A.G.) or research assistants during peak hours between noon and midnight 10 hours/day, 7 days/week, excluding some holidays, from July 1, 2007, to June 30, 2008. The study was approved by the Colorado Multiple Institutional Review Board.
A manual for administration and scoring accompanies the SAC. Three research assistants who enrolled subjects participated in a tutorial led by the principal investigator, who detailed SAC administration and assignment of concussion grades.
The SAC assesses 4 cognitive domains: orientation to time; immediate memory; concentration; and delayed recall. The SAC also incorporates a graded symptom checklist on which the severity of symptoms considered to be representative of mTBI is documented. The SAC and graded symptom checklist are scored separately. Administration takes ∼5 minutes. Responses to each item on the SAC are dichotomous: 1 point for each correct answer, 0 points for each incorrect answer. Lower scores on the SAC (possible range: 0–30 points) indicate greater cognitive deficit. The graded symptom checklist measures patient-reported severity of 15 symptoms on a 3-point scale. Higher checklist scores (possible range: 0–45 points) indicate an increased number and/or severity of symptoms. Version A of the SAC, including the graded symptom checklist, is included in the Appendix.
We modified the SAC in the following ways. We did not record educational level; our study population all had less than a high-school education. Instead, subjects were stratified into 4 age categories to ensure comparison of case-patients with controls of similar developmental level: 6 to 8, 9–11, 12–14, and 15–18 years. This process is consistent with that of previous studies in which the SAC in younger children was evaluated.17,18 We adapted the questions of the graded symptom checklist so that the language was age-appropriate. For example, rather than asking whether the subject had nausea, we asked if he or she “felt like throwing up.” These changes were standardized and asked in the same manner for all subjects. No changes were made to the cognitive components of the SAC. We incorporated parental responses only to questions relating to posttraumatic amnesia and LOC. We did not attempt to quantify the duration of posttraumatic amnesia, because some children were discharged before complete resolution of their amnesia. We recorded the presence or absence of amnesia at any time after injury and evaluated it as a dichotomous variable. All subjects were evaluated by using the SAC version A. GCS scores and AAN concussion grades were assigned to case-patients only. Assignment of concussion grades relied on the subject's and/or witness's recollection of the duration of symptoms for grades 1 and 2 (<15 or >15 minutes, respectively). Any child with LOC was assigned a concussion grade of 3 (most severe). If there was no witness to the head injury or the witness did not know if the subject suffered a LOC, no grade was assigned. GCS scores were assigned following the traditional adult classification system.19
Data Collection and Processing
Responses to the SAC and graded symptom checklist were elicited by the research assistants or principal investigator. Before data analysis, the principal investigator reviewed each subject entry for missing or discrepant data. Any discrepancies prompted a review of the medical record and comparison with the data-collection form, and final determination was made by the principal investigator.
We chose a difference between case-patients and controls of 15% (∼3 points) on the SAC to be clinically relevant. This difference was based on previous research results,10,12,13,17 which have shown that nearly 90% of subjects will score 2 or more points below preinjury baseline after mTBI and that decrements generally range between 2 and 4 points. Sample-size calculations showed that for each age group, 22 case-patients and 22 controls were required to detect such a difference with >80% power, assuming a common SD of 3.4 (based on research results10,12) using a 2-sided 2-group t test. Two-group t tests were used to compare the primary outcome measures between case-patients and controls. Spearman correlation coefficients were calculated to compare SAC and graded-symptom-checklist scores to AAN concussion grade, LOC, and posttraumatic amnesia in head-injured patients. The average graded-symptom-checklist score was calculated for grade 1, 2, and 3 concussions on the AAN scale. Logistic regression was used to examine the association between AMS (witnessed LOC, posttraumatic amnesia, or a GCS of 13 or 14) and graded-symptom-checklist scores, for which the median value of graded-symptom-checklist scores for case-patients with AMS (12 points) was used to categorize the outcome variable. The model was adjusted for age. The association between graded-symptom-checklist scores as a continuous outcome and AMS was also examined by using linear regression and adjusted for age. SAS 9.2 software (SAS Institute Inc, Cary, NC) was used for all analyses.
During the study period, 445 patients were approached for enrollment. Forty-one patients declined participation, and 66 patients met exclusion criteria. The remaining 348 patients (165 case-patients and 183 controls) were enrolled (Table 1). There was no difference in age or gender between case-patients and controls. Falls accounted for most injuries in children younger than 12 years. Sports-related injuries predominated in children aged 12 years and older. The mean time since injury for the controls, case-patients without AMS, and case-patients with AMS was 5.3, 3.8, and 4.4 hours, respectively, and not significantly different.
Total SAC scores trended lower among case-patients aged 9 years and older but were not significant (Fig 1). The 12- to 14-year age group showed the only significant difference in SAC scores (P = .001). When only those patients with unambiguous evidence of AMS were compared with the control group, only the 12- to 14-year age group demonstrated a significant difference similar to that of the entire case-patient group (data not shown). Graded-symptom-checklist scores for case-patients were substantially higher than those of controls. These differences were statistically significant for all ages (P ≤ .0001) (Fig 2).
Spearman correlation coefficients and P values that compared SAC and graded-symptom-checklist scores to traditional indicators of mTBI severity (AAN concussion grade, LOC, posttraumatic amnesia) and time since injury were calculated. We found significant positive correlations of graded-symptom-checklist scores with posttraumatic amnesia (P ≤ .0001), AAN grade (P = .03), and time since injury (P = .003). That is, higher graded-symptom-checklist scores correlated with the presence of posttraumatic amnesia, higher AAN grade, and longer time since injury. No significant correlation existed between the SAC and traditional indicators of mTBI severity. No correlation analysis was performed for the GCS scores, because 96% of the case-patients had a GCS of 15. LOC did not correlate with either SAC scores or graded-symptom-checklist scores.
To further investigate the correlation between AMS and symptoms, we calculated odds ratios for symptom scores on the basis of the presence or absence of AMS. Subjects with unambiguous evidence of AMS were 2.8 times more likely (95% confidence interval: 1.4–5.5) to have a graded-symptom-checklist score of >12 and had a mean graded-symptom-checklist score 5 points greater than those without AMS, when adjusting for age.
Given the lack of prospective data for AAN concussion grade, we conducted further comparisons against the graded symptom checklist. We calculated Spearman correlation coefficients for each of the 15 individual symptoms in the checklist. Three symptoms showed a correlation (P < .05): dizziness, photophobia, and memory disturbance.
The mean graded-symptom-checklist score was 9.2 (95% CI: 7.7–10.7) for case-patients with an AAN grade 1 concussion, 15.4 (95% CI: 11.5–19.3) for AAN grade 2 concussion, and 11.3 (95% CI: 9.1–13.5) for AAN grade 3 concussion. The difference between grade 1 and 2 injuries was significant, but the difference between grade 1 and 3 injuries was not.
We found no difference in SAC scores between case-patients and controls. With the exception of the youngest age group, there was a trend toward lower scores among case-patients, and the difference in the 12- to 14-year-old group did reach significance. When we compared only those patients with unambiguous evidence of AMS to controls, we were still unable to detect a difference in SAC scores.
Although the SAC scores of those in the 12- to 14-year-old age group did differ significantly, this finding is likely attributable to chance, given that the 15-to 18-year-old group did not demonstrate a statistically significant difference. Previous research has consistently demonstrated the validity and reliability of the SAC in this age group.10,13,14,20 Because we defined a case-patient as any child with blunt head trauma regardless of symptoms, we may have included subjects without actual underlying concussive brain injury, and the subsequent differences may not have been large enough to reach significance. However, this is unlikely to fully account for our findings, because differences did not reach significance in the subset of case-patients with unambiguous evidence of AMS. Another possible explanation relates to the fact that our study was powered on the basis of previous studies of baseline and postinjury scores of the same individuals, whereas in our study we compared scores among different individuals. More variability would be expected between individuals; thus, we may have included too few subjects to reach significance.
The graded-symptom-checklist scores differed considerably between case-patients and controls in all age groups. This difference was even more appreciable when we compared subjects with unambiguous evidence of AMS to those without AMS. There was a strong correlation between the presence of posttraumatic amnesia and greater symptom severity. Subjects with any posttraumatic amnesia (but not LOC) had a greater number and/or more severe symptom complex compared with children without posttraumatic amnesia. Furthermore, the mean graded-symptom-checklist score for case-patients with an AAN grade 3 concussion (any LOC) was actually lower than that found for case-patients with a grade 2 concussion, which suggests that LOC does not necessarily indicate a greater injury. This result is in line with earlier literature that revealed that posttraumatic amnesia is a reliable indicator of mTBI severity and is predictive of somatic, cognitive, and emotional symptoms.21,–,23 Because the disposition of patients was left to the discretion of treating physicians, we could not quantify the total duration of posttraumatic amnesia and evaluate the impact of its duration on this effect. It should be noted that the results of a recently published large multicenter study of children with TBI revealed that LOC is a risk factor for more severe TBI.24
Our results showed a correlation between AAN concussion grade and graded-symptom-checklist composite scores. However, only 3 symptoms (dizziness, photophobia, and memory disturbance) showed a correlation when considered individually. A few other symptoms approached significance. Given that not all subjects will suffer all 15 symptoms, it is likely that our sample size was too small to accurately detect all significant correlations.
Nearly all of our case-patients (96%) had a GCS of 15 at the time of evaluation despite having significantly more symptoms than controls. This finding highlights the important observation that a normal level of alertness and interaction with the environment does not exclude mTBI. A GCS of 15 should not reassure providers that no brain injury occurred, and it should not be construed to obviate the need for more detailed evaluation. The GCS is not a useful tool for classifying mTBI severity.25,26
The positive association between time since injury and graded-symptom-checklist scores is interesting. Results of previous sport-related concussion research with the SAC revealed the greatest symptoms shortly after injury.14,18 We expected decreasing symptoms over time rather than the increasing symptoms noted. There are 2 possible explanations for the unexpected finding. We studied a pediatric ED population that largely depends on adults to seek care on their behalf. There may have been a delay in seeking care among some subjects until the parents became concerned about the severity of symptoms or that symptoms were not clearing as quickly as expected. Second, the checklist elicits all symptoms present since the injury, not simply those present at the time of ED evaluation. Therefore, it is an additive rather than cross-sectional measure.
The results of this study must be interpreted in light of some limitations. Unlike studies using the SAC on the sideline immediately after a concussion, we evaluated patients up to 24 hours after injury, which allowed time for cognitive disturbances to resolve. Case-patients were compared with controls rather than their own preinjury baseline, which introduced a greater degree of score variability. As mentioned above, we also included as case-patients any patient who suffered blunt head trauma regardless of whether there was clear evidence of underlying concussive injury. The subgroup of case-patients who did not demonstrate unambiguous mental status changes may have contained some patients who suffered head injuries but not brain injuries and, therefore, performed similarly to controls on the SAC. It would be ideal to compare SAC scores to those of another cognitive test previously validated in a pediatric mTBI population as a reference standard. However, no test that could be quickly administered in an ED setting exists. Finally, subjects were not followed longitudinally, and we cannot draw any conclusions regarding the predictive value of these instruments.
Current methods of identifying the effects of mTBI on the pediatric brain are suboptimal. The GCS has no role in assessing the severity of mTBI. Our results suggest that the AAN concussion-grading scale may not accurately reflect the severity of injury and should be used with caution, especially when making decisions about returning athletes to competition.
We have demonstrated that the graded symptom checklist within the SAC systematically identifies the symptoms of mTBI in a school-aged pediatric population. Posttraumatic amnesia was found to predict greater symptom severity. Future efforts should focus on creating a rapid, easily administered tool for detecting the cognitive effects of mTBI in children that accounts for developmental differences and provides an assessment of the likelihood for developing postconcussive syndrome. Such a tool should be amenable to monitoring recovery after the acute event and easily administered in busy practice environments.
The authors would like to thank James P. Kelly, MD and Michael McCrea, PhD for their technical advice and review of the manuscript.
- Accepted June 15, 2010.
- Address correspondence to Joseph A. Grubenhoff, MD, Children's Hospital, 13123 E 16th Ave, B-251, Aurora, CO 80045. E-mail:
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
- ED =
- emergency department •
- TBI =
- traumatic brain injury •
- mTBI =
- mild traumatic brain injury •
- AAN =
- American Academy of Neurology •
- LOC =
- loss of consciousness •
- GCS =
- Glasgow Coma Scale •
- SAC =
- Standardized Assessment of Concussion •
- AMS =
- altered mental status •
- CI =
- confidence interval
- 6.↵American Academy of Neurology, Quality Standards Subcommittee. Practice parameter: the management of concussion in sports (summary statement). Neurology. 1997;48(3):581–585
- Ponsford J,
- Wilmott C,
- Rothwell A,
- et al
- Iverson GL
- Bowman SM,
- Bird TM,
- Aitken ME,
- Tilford JM
- McCrea M,
- Kelly JP,
- Kluge J,
- Ackley B,
- Randolph C
- Randolph C,
- Millis B,
- Barr WB,
- et al
- Shores EA,
- Lammél A,
- Hullick C,
- et al
- Kuppermann N,
- Holmes JF,
- Dayan PS,
- et al.
- Teasdale GM
- Copyright © 2010 by the American Academy of Pediatrics