Prehospital Versus Trauma Center Glasgow Coma Scale in Pediatric Traumatic Brain Injury Patients

j o u r n a l o f s u r g i c a l r e s e a r c h  s e p t e m b e r 2 0 1 9 ( 2 4 1 ) 1 1 2 e1 1 8

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.JournalofSurgicalResearch.com

Prehospital Versus Trauma Center Glasgow Coma Scale in Pediatric Traumatic Brain Injury Patients Joseph D. Drews, MD, MS,a,b Junxin Shi, MD, PhD,c Dominic Papandria, MD,a,1 Krista K. Wheeler, MS,c Eric A. Sribnick, MD, PhD,a,c,d and Rajan K. Thakkar, MDa,b,c,* a

Department of Surgery, Nationwide Children’s Hospital, Columbus, Ohio Department of Surgery, The Ohio State University Wexner Medical Center, Columbus, Ohio c Center for Pediatric Trauma Research, The Research Institute at Nationwide Children’s Hospital, Columbus, Ohio d Department of Neurosurgery, Nationwide Children’s Hospital, Columbus, Ohio b

article info

abstract

Article history:

Background: Traumatic brain injury (TBI) is a major source of morbidity and mortality in

Received 5 December 2018

children. The Glasgow Coma Scale (GCS) can be challenging to calculate in pediatric pa-

Received in revised form

tients. Our objective was to determine its reproducibility between prehospital providers

6 March 2019

and pediatric trauma hospital personnel.

Accepted 22 March 2019

Materials and methods: The institutional trauma database for a level 1 pediatric trauma

Available online xxx

center was queried for patients aged
18 y who presented with a TBI. Demographics, mechanism, prehospital GCS, and trauma center GCS were collected. Agreement was

Keywords:

evaluated with weighted kappa (k) coefficients (0 ¼ agreement no better than that expected

Traumatic brain injury

by chance alone, 1 ¼ perfect agreement).

Glasgow coma scale

Results: The inclusion criteria were met by 1711 patients, 263 of whom were aged <3 y.

Pediatric

Prehospital GCS and trauma center GCS differed in 766 patients (44.8%). Agreement be-

Agreement

tween prehospital GCS and trauma center GCS was moderate for all patients (k ¼ 0.61, 95% confidence interval [CI] 0.57-0.64). Agreement was slightly better than chance alone in patients with trauma center GCS between 9 and 12 y (k ¼ 0.09, 95% CI 0.03-0.15) and was lower for children aged 0-2 y (k ¼ 0.51, 95% CI 0.42-0.61) than for those aged between 3 and 18 y (k ¼ 0.63, 95% CI 0.59-0.66). Younger children were more likely to have score differences of at least 3 points (21.3% versus 13.6% of 3- to 18-y-olds, P < 0.001). Conclusions: Prehospital and trauma center GCS scores frequently disagree in children, particularly in TBI patients aged <3 y and those with moderate TBI. Centers should consider the inconsistency of the pediatric GCS when triaging TBI patients. ª 2019 Elsevier Inc. All rights reserved.

Background Trauma is the leading cause of mortality in children aged >1 y.1 Traumatic brain injuries (TBIs) are a major source of

morbidity and mortality, prompting over 50,000 pediatric admissions and generating more than $1 billion in hospital charges annually.2 Rapid triage of patients at risk for clinically important TBI as a means to guide early intervention has been

* Corresponding author. Department of Pediatric Surgery, Nationwide Children’s Hospital, 700 Children’s Drive, FOB 6C, Columbus, OH 43205. Tel.: þ1 614 722 0448; fax: þ1 614 722 3903. E-mail address: [email protected] (R.K. Thakkar). 1 Present address: Department of Surgery, Emory University School of Medicine, 1364 Clifton Road NE, Atlanta, GA 30322. 0022-4804/$ e see front matter ª 2019 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jss.2019.03.038

113

drews et al  pediatric gcs agreement

a consistent focus of research of this potentially devastating injury.3-6 Since its introduction in 1974 by Teasdale and Jennett,7 the Glasgow Coma Scale (GCS) has been broadly adopted as a critical adjunct in the management of patients with TBI. GCS is used to triage TBI patients and rapidly identify patients with severe injuries, and the scale has been modified for use in children aged <3 y (Table 1).8,9 Current practice uses this pediatric GCS for nonverbal patients presenting with TBI.9 GCS values determined by prehospital providers (PH-GCS) are commonly used to activate a tiered level response for patient arrival in the emergency department of the trauma center (TC-GCS), with low scores triggering the highest levels of resource mobilization. In adults, agreement between PH-GCS and TC-GCS is poor for patients with GCS 9-12.10,11 Little is known regarding such discrepancies in pediatric trauma systems. We hypothesized that agreement between prehospital and trauma center GCS scores would be poor in children aged <3 y and in all pediatric TBI patients with moderate GCS scores of 9-12.

hemorrhage) and other significant closed head injuries (e.g., skull fracture) as well as mild TBI (e.g., concussion, cerebral contusion). In cases of missing ICD-9 codes, diagnoses were extracted manually from the electronic medical record. A full list of the included diagnosis codes is provided in Supplementary Table 1. Patients transferred from other hospitals and children with chronic abuse diagnoses were excluded.

Data collected Patient demographics and pertinent history were abstracted for further analysis. This included patient age, sex, mechanism of injury, transport time, admission and discharge disposition, and all ICD-9 diagnosis codes associated with hospital admission. The first GCS documented on the Emergency Medical Services run sheet (PH-GCS) and the first GCS recorded at the trauma center (TC-GCS) were collected, along with Injury Severity Score and head Abbreviated Injury Scale values. Patients were categorized as having mild (GCS 13-15), moderate (GCS 9-12), or severe (GCS 3-8) TBIs in accordance with previous studies.10,12-16

Materials and methods Statistical analysis Data source A retrospective review was conducted of trauma registry and electronic health record data (1994-2016) collected at a freestanding children’s hospital designated as an American College of Surgeons level 1 pediatric trauma center. This study was approved by the Nationwide Children’s Hospital Institutional Review Board, and informed consent was waived.

Study population Patients aged
18 y with a diagnosis of TBI were included. TBI was identified using International Classification of Diseases, Ninth Revision (ICD-9), diagnosis codes present in the trauma registry. These included intracranial hemorrhage (e.g., epidural hematoma, subdural hematoma, subarachnoid

Differences between PH-GCS and TC-GCS were defined a priori as large (3 points) or small (<3 points). Multivariable linear regression was performed to assess the effect of age, TC-GCS, injury mechanism, and type of TBI on the difference between PH-GCS and TC-GCS. Agreement between PH-GCS and TC-GCS was evaluated by calculation of weighted kappa coefficients in which PH-GCS and TC-GCS were considered the two raters.17 Kappa ranges from 1 (worse than chance) to 1 (perfect), with 0 indicating agreement no better than that expected by chance alone. Weighted kappa coefficients are preferred for ordinal variables.18 We used CicchettieAllison weights to calculate the kappa coefficient.19 A similar analysis of weighted kappa coefficients was performed for the GCS subscores (eye, verbal, and motor). A sensitivity analysis was also performed to assess effect of transport time on agreement.

Table 1 e Adult and pediatric GCS.8 Points

Adult

Pediatric (for use in nonverbal patients)

Eyes

Verbal

Motor

Eyes

Verbal

Motor

1

Does not open

No response

No response

Does not open

No response

No response

2

Opens to pain

Incomprehensible sounds

Abnormal extension posturing

Opens to pain

Moans

Abnormal extension posturing

3

Opens to voice

Inappropriate words

Abnormal flexure posturing

Opens to voice

Cries to pain

Abnormal flexure posturing

4

Opens spontaneously

Confused

Withdraws to pain

Opens spontaneously

Irritable, cries

Withdraws to pain

5

N/A

Oriented

Localizes to pain

N/A

Coos, babbles

Withdraws to touch

6

N/A

N/A

Follows commands

N/A

N/A

Moves spontaneously

N/A ¼ not applicable.

114

j o u r n a l o f s u r g i c a l r e s e a r c h  s e p t e m b e r 2 0 1 9 ( 2 4 1 ) 1 1 2 e1 1 8

T-tests were used to assess for differences between kappa values. The threshold of statistical significance was defined as P < 0.05. All analyses were conducted using SAS Enterprise Guide, Version 7.11 HF3 (SAS Institute Inc, Cary, NC).

Results Demographics and raw GCS scores A total of 1711 patients met the inclusion criteria, 263 (15.4%) of whom were aged <3 y (Table 2). Of these, 621 (36.3%) were critically ill and were admitted to the intensive care unit or taken directly to the operating room from the trauma bay. The most common mechanisms of injury were motor vehicle

Table 2 e Baseline patient characteristics. Characteristic Total number of patients Male Median age, y (IQR)

Number of patients (%) 1711 1090 (63.7)

50 (35-75)

Median injury severity score (IQR)

9 (5-16)

Median head Abbreviated Injury Scale (IQR)

2 (2-3)

Mechanism Motor vehicle collision

498 (29.1)

Fall

391 (22.9)

Struck as pedestrian

276 (16.1)

Bicycle collision

210 (12.3)

Motorcycle collision

100 (5.8)

Sports injury

67 (3.9)

Assault

36 (2.1)

Gunshot wound

17 (1.0) 116 (6.8)

Type of traumatic brain injury Concussion

633 (37.0)

Skull fracture

609 (35.6)

Subdural hematoma

174 (10.2)

Subarachnoid hemorrhage

127 (7.4)

Epidural hematoma Multiple or indeterminate injuries

51 (3.0) 723 (42.3)

Emergency department disposition Floor

993 (60.9)

Intensive care unit

520 (31.9)

Operating room

101 (6.2)

Other

18 (1.1)

Hospital disposition Home

1259 (73.6)

Rehab

115 (6.7)

Other or unknown

337 (19.7)

Mortality IQR ¼ interquartile range.

Multivariable regression of difference in scores

9 (5-13)

Median transport time, min (IQR)

Other

collision (n ¼ 498, 29.1%), fall (n ¼ 391, 22.9%), victim struck as pedestrian (n ¼ 276, 16.1%), and bicycle collision (n ¼ 210, 12.3%), with the vast majority classified as blunt trauma (n ¼ 1685, 98.5%). The most common TBI-related diagnoses were concussion (n ¼ 633, 37.0%) and skull fracture (n ¼ 609, 35.6%). Raw GCS total scores and subscores are provided in Figure 1. Examining overall GCS discrepancies, PH-GCS, and TC-GCS were identical in 945 patients (55.2%); PH-GCS was lower than TC-GCS for 473 patients (27.6%), and PH-GCS was higher than TC-GCS for 293 patients (17.1%). Overall agreement was less for children aged <3 y (46.8% of whom had identical PH-GCS and TC-GCS) compared with older patients (56.8%). PH-GCS was lower than TC-GCS in 33.8% of patients aged <3 y (versus 26.5% of children aged 3-18 y). PH-GCS was higher than TC-GCS in 19.4% of younger children (versus 16.7% of older children). A total of 3.7% of patients overall (and 6.8% of patients aged <3 y) with a PH-GCS of
8 were determined to have a TC-GCS of 9. A total of 3.2% of all patients (and 3.0% of patients aged < 3 y) with a PH-GCS of 9 scored
8 in the trauma center.

40 (2.3)

On multivariable linear regression (Supplementary Table 2), increasing age (P < 0.001) and TC-GCS (P < 0.001) were associated with significantly closer PH-GCS and TC-GCS scores when also controlling for mechanism and diagnosis. Being struck as a pedestrian was associated with closer PH-GCS and TC-GCS scores (P ¼ 0.05). A diagnosis of epidural hematoma was associated with a 0.6-point larger difference between PHGCS and TC-GCS scores (P ¼ 0.03), whereas a diagnosis of subarachnoid hemorrhage was associated with mean PH-GCS and TC-GCS scores being 0.5 points closer (P ¼ 0.01).

Interrater GCS agreement by kappa coefficient All patients To account for the possibility of randomly occurring agreement between PH-GCS and TC-GCS, weighted kappa coefficients were used to assess interrater reliability (Fig. 2). Agreement was moderate for all patients (k ¼ 0.61, 95% confidence interval [CI] 0.57-0.64), and 13.6% had large differences in scores of at least three points. The level of agreement was similar for the eye, verbal, and motor subscores (eye: k ¼ 0.56, 95% CI 0.52-0.60; verbal: k ¼ 0.58, 95% CI 0.55-0.62; motor: k ¼ 0.58, 95% CI 0.54-0.63).

Patients with moderate TBI Agreement was slightly better than chance alone in patients with TC-GCS between 9 and 12 (k ¼ 0.09, 95% CI 0.03-0.15). Of these 143 patients, 89.5% had different prehospital and trauma center GCS scores, with a mean difference of 2.7 points; 46.2% of patients had a difference in scores of at least 3 points. PH-GCS both overestimated (50.3% of patients) and underestimated (39.2%) TC-GCS in this subgroup. Kappa coefficients were low for all three subscores (eye: k ¼ 0.18, 95% CI 0.05-0.30; verbal: k ¼ 0.19, 95% CI 0.07-0.31; motor: k ¼ 0.11, 95% CI 0.01 to 0.22).

115

drews et al  pediatric gcs agreement

A

Eye

TBI PATIENTS < 3 YEARS OLD

TC-GCS

PH-GCS

Total Score

1

2

3

4

1

18

4

0

7

2

3

4

3

22

3

1

2

2

19

4

1

7

9

158

Verbal

PH-GCS

TC-GCS 1

2

3

4

1

15

3

5

4

5 8

2

2

1

1

2

4

3

2

1

2

2

10

4

0

1

11

15

21

5

2

1

2

27

118

Motor

PH-GCS

TC-GCS 2

3

4

5

6

11

1

0

2

2

3

2

1

0

0

1

0

2

3

1

0

0

0

1

1

4

3

1

0

5

6

16

5

0

1

1

4

7

25

6

1

0

0

4

4

156

Eye

TBI PATIENTS ≥ 3 YEARS OLD

TC-GCS

Total Score

PH-GCS

B

1 1

1

2

3

4

1

83

10

4

21

2

31

17

15

22

3

8

9

34

135

4

16

14

50

955

Verbal

PH-GCS

TC-GCS 1

2

3

4

5

1

85

11

6

10

10

2

10

22

10

9

4

3

7

13

11

13

19

4

7

6

11

98

154

5

15

9

15

79

790

Motor

Fig. 1 e Raw GCS scores and subscores ([A] TBI patients <3 y [B] TBI patients ‡ 3y). Yellow boxes indicate instances in which PH-GCS and TC-GCS agreed perfectly. Blue represents cases in which PH-GCS overestimated TC-GCS, and green indicates PH-GCS underestimating TC-GCS. (Color version of figure is available online.)

Children aged <3 y Agreement between PH-GCS and TC-GCS was lower for children aged <3 y than for those aged between 3 and 18 y (k ¼ 0.51

versus 0.63, P ¼ 0.02). PH-GCS both overestimated (33.8%) and underestimated (19.4%) TC-GCS in this age group. Younger children were more likely to have score differences of at least

116

j o u r n a l o f s u r g i c a l r e s e a r c h  s e p t e m b e r 2 0 1 9 ( 2 4 1 ) 1 1 2 e1 1 8

GCS agreement by traumatic brain injury severity AGREEMENT

NONE

MINIMAL

WEAK

MODERATE

All patients

All GCS

STRONG ALMOST PERFECT

TC-GCS 13-15 TC-GCS 9-12 TC-GCS 3-8

Age 0-2 yrs

Age 3-18 yrs

AGREEMENT

NONE

0.0

MINIMAL

0.2

WEAK

MODERATE

0.4

0.6

ALMOST STRONG PERFECT

0.8

1.0

Kappa

Fig. 2 e PH-GCS and TC-GCS agreement by age and severity of traumatic brain injury. Interrater agreement was weak for all ages and TC-GCS scores but was worse for children aged <3 y and those with moderate TC-GCS scores of 9-12 (open squares).

3 points (21.3% of children aged 0-2 y versus 13.6% of children aged 3-18 y, P < 0.001). Among children aged <3 y, agreement was similar to chance alone for those with TC-GCS of 13-15 (k ¼ 0.07, 95% CI 0.01-0.13) and TC-GCS of 9-12 (k ¼ 0.07, 95% CI 0.05 to 0.19). In young children with moderate TC-GCS scores, the mean difference between PH-GCS and TC-GCS was 2.97 points (standard deviation 2.34). Fifty percent of young children with moderate TC-GCS of 9-12 had large differences in PH-GCS and TC-GCS scores. Patients with severe TC-GCS scores of 3-8 fared better (k ¼ 0.41, 95% CI 0.24-0.59).

transport time (Fig. 3). Agreement between PH-GCS and TCGCS was similar regardless of whether patients arrived within 30 min (k ¼ 0.61, 95% CI 0.56-0.66), 31-60 min (k ¼ 0.62, 95% CI 0.56-0.69), or 61-120 min (k ¼ 0.54, 95% CI 0.44-0.63). Children aged <3 y arriving within 30 min had weak interrater agreement (k ¼ 0.47, 95% CI 0.36-0.58), and 60.5% had different PH-GCS and TC-GCS scores with a mean difference of 1.85 points. Agreement was higher, but still only moderate, for patients aged 3-18 y arriving to the trauma center within 30 min (k ¼ 0.65, 95% CI 0.60-0.70).

Sensitivity analysis: stratification by transport time

Discussion To attempt to exclude patients with potentially true neurologic changes during prolonged transports, a sensitivity analysis was performed in which patients were stratified by

The GCS plays an important role in the evaluation, triage, and management of patients with TBIs,8,20,21 but it is only as useful

GCS agreement by transport time AGREEMENT

NONE

MINIMAL

WEAK

MODERATE

All patients

STRONG ALMOST PERFECT

< 30 min 31-60 min 61-120 min

Age 0-2 yrs

Age 3-18 yrs

AGREEMENT

NONE

0.0

MINIMAL

0.2

WEAK

0.4

MODERATE

0.6

ALMOST STRONG PERFECT

0.8

1.0

Kappa

Fig. 3 e PH-GCS and TC-GCS agreement by transport time. Interrater agreement was weak even for patients who arrived at the trauma center within 30 min. Agreement decreased slightly with increased transport times.

drews et al  pediatric gcs agreement

as its reproducibility. This study indicates that agreement between prehospital and trauma center GCS scores is relatively low in pediatric trauma patients with TBIs, which is consistent with other reports.10,22 Agreement was lower in children aged <3 y, and more than 20% of these younger children had large differences in PH-GCS and TC-GCS. Agreement in scores was slightly better than chance alone in pediatric patients with moderate TC-GCS of 9-12. Inaccurate GCS calculation in the prehospital setting can result in both under- and over-triage on arrival at the trauma center. Although overtriage (mobilization of greater resources that are required to assess and treat the injury complex) may result in increased costs and resource utilization without benefit, undertriage (wherein the initial response is insufficient to optimally manage the patient’s injuries) may delay critical interventions and increase morbidity and mortality associated with TBI. The present study demonstrates poor agreement and an overall bias toward lower GCS values in the field versus in the trauma center (27% of field-determined values were lower than that reported in the trauma center), indicating the potential for significant overtriage. Scoring the GCS is challenging when patients are in the moderate range,10 but accurate scores in this span are critically important because they affect the decision to intubate along with activation of trauma alerts. In our institution, a PHGCS score of 9 is sufficient to activate a level 1 trauma alert (requiring the greatest resource mobilization). Significant overtriage across the threshold GCS of 9 can thus prompt unneeded institutional resource expenditure along with patient cost. Children aged <3 y were particularly susceptible to this potential misclassification. This may be because prehospital providers encounter more adult and verbal pediatric trauma patients than nonverbal pediatric trauma patients, thus making them more prone to miscalculation of the pediatric GCS or inappropriate usage of the adult GCS. The pediatric GCS, and particularly the verbal subscore, is challenging to calculate even for experienced providers.22-24 Although challenges with tabulating the verbal subscore in a nonverbal population are intuitive, our data demonstrate similarly low agreement for the eye and motor subscores. Alternatively, younger patients could have greater fluctuations in their neurologic status between the time of injury and arrival at the trauma center. There are several important limitations of the present study. It is retrospective and is based on the GCS scores and diagnoses in our institutional trauma database. The database does not specify if patients were scored on the adult or pediatric GCS. The pediatric GCS is used in nonverbal patients. Our practice is to use it in patients aged <3 y; therefore, we dichotomized patients at that cutoff with the presumption that younger patients were likely scored on the pediatric GCS and older patients on the adult GCS. However, it is possible that there was overlap that we could not determine retrospectively or that prehospital and in-hospital providers used different scales (adult and pediatric) for the same patient. The most significant limitation is that the GCS is a dynamic score, which changes with the patient’s condition. Patients could have had true changes in their GCS scores between their prehospital and trauma center evaluations. With different providers scoring the patient at different time points, it is

117

impossible to determine if differences in GCS scores were due to inaccurate scoring on the part of the prehospital providers and/or trauma center team or if the patient’s neurologic status changed. We tested this to the best of our ability by stratifying by transport time, and agreement was weak even for patients who arrived at the trauma center within 30 min. Regardless of the reason for the disagreement, however, these results suggest that PH-GCS may not be the best criterion by which to triage trauma patients arriving from the field, particularly those aged <3 y.

Conclusions Prehospital and trauma center GCS scores frequently disagree in pediatric patients. Agreement is low for all patients but is especially poor in children aged <3 y and in those with moderate TBI (GCS 9-12). Future directions should include a prospective evaluation of PH-GCS and TC-GCS agreement in which an independent expert scores at both evaluation points and evaluates the frequency and accuracy with which prehospital and in-hospital providers use the pediatric GCS. In the meantime, the inconsistency of prehospital and trauma center GCS scores should be considered when triaging pediatric TBI patients.

Acknowledgment This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Authors’ contributions: J.D.D conceptualized and designed the study, acquired, analyzed, and interpreted the data, drafted the initial article, and critically reviewed and revised the manuscript for important intellectual content. J.S. and K.K.W. conducted the statistical analysis and interpreted the data as well as critically reviewed and revised the article for important intellectual content. D.P. analyzed and interpreted the data, drafted the initial article, and critically reviewed and revised the article for important intellectual content. E.A.S. conceptualized and designed the study, interpreted the data, as well as critically reviewed and revised the article for important intellectual content. R.K.T. conceptualized and designed the study, acquired, analyzed, and interpreted the data, as well as critically reviewed and revised the article for important intellectual content. All authors approved the final article as submitted and agree to be accountable for all aspects of the work.

Disclosure The authors report no proprietary or commercial interest in any product mentioned or concept discussed in this article.

Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.jss.2019.03.038.

118

j o u r n a l o f s u r g i c a l r e s e a r c h  s e p t e m b e r 2 0 1 9 ( 2 4 1 ) 1 1 2 e1 1 8

references

1. National Center for Injury Prevention and Control. 10 Leading causes of Death by age Group, United States e 2016. Centers for Disease Control and Prevention; 2018. Available at: https://www. cdc.gov/injury/wisqars/LeadingCauses.html. Accessed November 20, 2018. 2. Taylor CA, Bell JM, Breiding MJ, Xu L. Traumatic brain injuryerelated emergency department visits, hospitalizations, and deaths d United States, 2007 and 2013. MMWR Surveill Summ. 2017;66:1e16. 3. Dunning J, Daly JP, Lomas J-P, et al. Derivation of the children’s head injury algorithm for the prediction of important clinical events decision rule for head injury in children. Arch Dis Child. 2006;91:885e891. 4. Oman JA, Cooper RJ, Holmes JF, et al. Performance of a decision rule to predict need for computed tomography among children with blunt head trauma. Pediatrics. 2006;117:e238ee246. 5. Kuppermann N, Holmes JF, Dayan PS, et al. Identification of children at very low risk of clinically-important brain injuries after head trauma: a prospective cohort study. Lancet. 2009;374:1160e1170. 6. Osmond MH, Klassen TP, Wells GA, et al. CATCH: a clinical decision rule for the use of computed tomography in children with minor head injury. Can Med Assoc J. 2010;182:341e348. 7. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale. Lancet. 1974;2:81e84. 8. Borgialli DA, Mahajan P, Hoyle JD, et al. Performance of the pediatric Glasgow Coma Scale score in the evaluation of children with blunt head trauma. Acad Emerg Med. 2016;23:878e884. 9. Reilly PL, Simpson DA, Sprod R, Thomas L. Assessing the conscious level in infants and young children: a paediatric version of the Glasgow Coma Scale. Childs Nerv Syst. 1988;4:30e33. 10. Kerby JD, MacLennan PA, Burton JN, McGwin G, Rue LW. Agreement between prehospital and emergency department Glasgow Coma Scores. J Trauma. 2007;63:1026e1031. 11. Bazarian JJ, Eirich MA, Salhanick SD. The relationship between pre-hospital and emergency department Glasgow Coma Scale scores. Brain Inj. 2003;17:553e560.

12. Rimel RW, Giordani B, Barth JT, Boll TJ, Jane JA. Disability caused by minor head injury. Neurosurgery. 1981;9:221e228. 13. Rimel RW, Giordani B, Barth JT, Jane JA. Moderate head injury: completing the clinical spectrum of brain trauma. Neurosurgery. 1982;11:344e351. 14. Scholten AC, Haagsma JA, Andriessen TMJC, et al. Healthrelated quality of life after mild, moderate and severe traumatic brain injury: patterns and predictors of suboptimal functioning during the first year after injury. Injury. 2015;46:616e624. 15. Kraus JF, Fife D, Cox P, Ramstein K, Conroy C. Incidence, severity, and external causes of pediatric brain injury. Am J Dis Child. 1986;140:687e693. 16. Pal J, Brown R, Fleiszer D. The value of the Glasgow Coma Scale and injury severity score: predicting outcome in multiple trauma patients with head injury. J Trauma. 1989;29:746e748. 17. McHugh ML. Interrater reliability: the kappa statistic. Biochem Med. 2012;22:276e282. 18. Cohen J. Weighted kappa: nominal scale agreement provision for scaled disagreement or partial credit. Psychol Bull. 1968;70:213e220. 19. Cicchetti DV, Allison T. A new procedure for assessing reliability of scoring EEG sleep recordings. Am J EEG Technol. 1971;11:101e110. 20. Marcin JP, Pollack MM. Triage scoring systems, severity of illness measures, and mortality prediction models in pediatric trauma. Crit Care Med. 2002;30(11 Suppl l):S457eS467. 21. Holmes JF, Palchak MJ, MacFarlane T, Kuppermann N. Performance of the pediatric Glasgow Coma Scale in children with blunt head trauma. Acad Emerg Med. 2005;12:814e819. 22. DiBrito SR, Cerullo M, Goldstein SD, Ziegfeld S, Stewart D, Nasr IW. Reliability of Glasgow Coma Score in pediatric trauma patients. J Pediatr Surg. 2018;53:1789e1794. 23. Lieh-Lai MW, Theodorou AA, Sarnaik AP, Meert KL, Moylan PM, Canady AI. Limitations of the Glasgow Coma Scale in predicting outcome in children with traumatic brain injury. J Pediatr. 1992;120(2 PART 1):195e199. 24. Acker SN, Ross JT, Partrick DA, Nadlonek NA, Bronsert M, Bensard DD. Glasgow motor scale alone is equivalent to Glasgow Coma Scale at identifying children at risk for serious traumatic brain injury. J Trauma Acute Care Surg. 2014;77(2):304e309.