Indication of whole body computed tomography in pediatric polytrauma patients—Diagnostic potential of the Glasgow Coma Scale, the mechanism of injury and clinical examination

Accepted Manuscript Title: Indication of whole body computed tomography in pediatric polytrauma patients—diagnostic potential of the Glasgow Coma Scale, the mechanism of injury and clinical examination Authors: Claudia Frellesen, Daniel Klein, Patricia Tischendorf, Julian L. Wichmann, Sebastian Wutzler, Johannes Frank, Hanns Ackermann, Thomas J. Vogl, Moritz Albrecht, Katrin Eichler PII: DOI: Reference:

S0720-048X(18)30184-0 https://doi.org/10.1016/j.ejrad.2018.05.022 EURR 8198

To appear in:

European Journal of Radiology

Received date: Revised date: Accepted date:

25-2-2018 13-5-2018 20-5-2018

Please cite this article as: Frellesen C, Klein D, Tischendorf P, Wichmann JL, Wutzler S, Frank J, Ackermann H, Vogl TJ, Albrecht M, Eichler K, Indication of whole body computed tomography in pediatric polytrauma patients—diagnostic potential of the Glasgow Coma Scale, the mechanism of injury and clinical examination, European Journal of Radiology (2018), https://doi.org/10.1016/j.ejrad.2018.05.022 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Original Research

Indication of whole body computed tomography in pediatric polytrauma patients – diagnostic potential of the Glasgow Coma Scale, the mechanism of

IP T

injury and clinical examination

Claudia Frellesen MD1,*, Daniel Klein MD1, Patricia Tischendorf MD1, Julian L. Wichmann MD1,

SC R

Sebastian Wutzler MD2, Johannes Frank MD2, Hanns Ackermann PhD3, Thomas J. Vogl MD1, Moritz Albrecht MD1, Katrin Eichler MD1

1

U

Department of Diagnostic and Interventional Radiology, Clinic of the Goethe University, Theodor-

N

Stern-Kai 7, 60590 Frankfurt / Germany 2

A

Department of Trauma, Hand and Reconstructive Surgery, Clinic of the Goethe University, Theodor-

M

Stern-Kai 7, 60590 Frankfurt / Germany 3

Department of Biostatistics and Mathematical Modelling, Clinic of the Goethe University, Theodor-

Correspondence:

PT

*

ED

Stern-Kai 7, 60590 Frankfurt / Germany

CC E

Dr. med. Claudia Frellesen

Klinikum der Goethe-Universitaet Institut fuer Diagnostische und Interventionelle Radiologie

A

Haus 23C UG

Theodor-Stern-Kai 7 60590 Frankfurt am Main / Germany Email: [email protected] Tel: +49-69-6301-80407

1

Fax: +49-69-6301-7258

Highlights 

There is no clinical variable, which can be used as sole indication for WBCT in pediatric polytrauma patients. The indication to undergo WBCT in pediatric polytrauma patients should be based on

IP T



the synopsis of all clinical variables available. 

In combination without a suspicious finding in the clinical examination, major trauma



SC R

is unlikely after mild trauma.

Dedicated CT of individual body regions based on combinations of severity of MOI and clinical examination may be preferable in order to reduce radiation exposure. Cranial CT is recommended at GCS ≤13.

N

U



A

Abstract

Objective: To evaluate the diagnostic potential of the Glasgow Coma Scale (GCS), the

M

mechanism of injury (MOI) and clinical examination (CE) for the indication of whole body computed tomography (WBCT) in pediatric polytrauma patients.

ED

Materials & Methods: 100 pediatric polytrauma patients with WBCT were analysed in terms of age, gender, (MOI), GCS, detected injury, FAST, CE and Injury Severity Score (ISS).

PT

Correlations between all clinical variables and patient groups with (p+) and without (p-) injury were assessed.

CC E

Results: Mean age was 9.13±4.4 years (28% female patients). Injury was detected in 71% of the patients, most commonly of the head (43%). There was no significant correlation between type or severity of MOI and ISS (p>0.1). None of the clinical variables had a significant predictive effect on p+. The optimum discrimination threshold of GCS was at 12.5 relating to

A

craniocerebral injuries. Severity of MOI and FAST showed best predictive effects on thoracic and abdominal pathologies, respectively, but with only low sensitivities (<20%). Conclusion: There is no clinical variable, which can be used as sole indication for WBCT in pediatric polytrauma patients. GCS had a significant predictive value for craniocerbral injuries and CCT is recommended at GCS≤13.

2

Keywords: Pediatric polytrauma patients; Whole body CT; indication; GCS; mechanism of injury

Introduction

IP T

Pediatric polytrauma patients are less frequently admitted to the trauma room than adults. Even in supra-regional trauma centers they represent merely 7% of the whole polytrauma

SC R

patient collective [1, 2]. However, polytrauma of children is accompanied with a high

mortality rate [3]. Thus, an optimal trauma room management conducted by a constantly trained professional trauma team is required. Due to differences regarding behavior and

U

physiological conditions, children often show other injury patterns when compared to adults,

N

which can additionally impede patient care [4]. Aside from the personell resources in terms

A

of healthcare providers, imaging techniques are a crucial part in the treatment of polytrauma

M

patients. It has been demonstrated that the use of whole body computed tomography (WBCT) in severely injured adult patients positively impacts on the survival [5]. Nevertheless, this

ED

benefit has to be weighed against the associated significant increased exposure to ionising radiation, especially in pediatric patients [6]. It is known that the radiation sensitivity of

PT

children is primarily based on a higher cellular proliferation rate and the different anatomical

CC E

distribution of radiation sensitive organs [7]. Several studies have concluded a small but definite risk of long-term cancer as a consequence of CT examinations in early life [8]–[11]. Therefore, the use of X-rays in pediatric patients has to be considered carefully, even in

A

potentially severe injured children. Prescribed factors based on the mechanism of injury (MOI), the injury pattern and the vital parameters, including the Glasgow Coma Scale (GCS), help to identify possibly life-threatening injuries and lead to routine WBCT in adult patients. However, it remains uncertain, if the benefit overweighs the radiation risk, in particular for children. The indication for WBCT is further complicated, when the parameters deviate from 3

each other, e.g. severe MOI, but no obvious injury or restrictions of the vital parameters. With regard to the MOI it has been shown that it cannot generally be relied upon to be the sole indication for the use of computed tomography (CT) in blunt pediatric trauma [12]. Garcia et al. suggested that combining the mechanism of injury with vital signs, historical data and physical examination to build a clinical suspicion to direct focused imaging may be a safe and

IP T

effective method of assessing children with blunt trauma [13]. In this study, we retrospectively analyzed pediatric polytrauma patients that had undergone WBCT in our

SC R

trauma room between 2007 and 2016. We aimed to refine the indication for WBCT and to compare previously published data. Specific focus was put on the GCS, which is considered to indicate traumatic brain injury and helps to justify cranial CT [14], [15]. The aim of this

U

study was furthermore to evaluate whether the GCS may also be an appropriate scoring index

Materials & Methods

M

A

N

for whole-body injury or whether other clinical variables may be more precise.

ED

CT parameters and WBCT protocol

PT

Between 2007 and 2012, WBCT was performed on a 16-slice CT device (Somatom Sensation

CC E

16, Siemens Healthcare, Forchheim, Germany). Since 2012, patients have been examined on a 64-slice sliding gantry CT (Somatom Definition AS, Siemens). WBCT starts with a noncontrast CT of the head and the cervical spine including the seventh cervical vertebra with

A

100 or 120 kV and 300 ref.mAs using automated exposure control (AEC). During this scan, the arms of the patient are positioned besides the body. Afterwards, a monophasic contrastenhanced thoraco-abdominal CT is performed [16], with the arms being elevated above the head in order to avoid artifacts and additional radiation exposure with 100 or 120 kV and 190 ref.mAs using AEC. The amount of the contrast medium is dependent on body weight 4

((Imeron 400, Bracco, Konstanz, Germany with 0.5 mg iodine/kg body weight), if known. The length of the scan extents from the seventh cervical spine to the symphysis. Dose parameters are individually selected according to the recommendations of the Federal Office for Radiation Protection (BfS) for pediatric X-ray imaging, see table 1.

IP T

Patient population and data gathering The study complies with the Guidelines of Helsinki in its current version and was approved

SC R

by the ethics committee of our hospital. The need for informed patient consent was waived.

Patients were identified by a search in the RIS/PACS system of our department. The data was

U

retrospectively acquired from pediatric polytrauma patients under the age of 18 years that

N

underwent WBCT in our trauma room due to suspicion of life-threatening injuries between

A

July 2007 and November 2016. Patients with focused CT examinations of individual body

M

regions or incomplete data documentation were excluded. Demographics including gender and age were listed. Detected injuries were extracted from radiological reports and the patient

ED

collective was split into two groups, the first with (p+) and the second without pathology (p-). Furthermore, subgroups were structured according to injured body regions and detected

PT

pathology: craniocerebrum, thorax, abdomen and skeleton. Only injuries that were found in

CC E

the WBCT report were taken into account and findings in additional examinations were ignored in the statistical analysis. Further clinical information according to the pre-clinical phase and the trauma room were acquired from the documentation system TraumaRegister

A

DGU®. Two GCS, the first at the initial scene of the accident (GCSi) and the second at the time of admission to the trauma room (GCSa), the ISS and the MOI were collected for each patient. The GCS is a neurological scale, which is used to assess the level of consciousness in both adults and paediatrics with a minimum of 3 and a maximum of 15. Values less than or equal to 8 indicate a severe head injury [14]. As it is based on verbal communication, a 5

modified scoring system, the Pediatric Glasgow Coma Scale (PGCS) has been developed for preverbal pediatrics [17], see table 2. With regard to the degree of severity, the MOI has been rated as either severe or mild (see severe cause of trauma in table 3). Additionally, the MOI has been divided into four groups

< 3m and other MOI. Road accidents also include cyclists and pedestrians.

IP T

regarding the type of accident: road accidents, fall from a height of > 3m, fall from a height of

The ISS is an anatomical scoring system that provides an overall score for patients with

SC R

multiple injuries. Each injury is assigned an Abbreviated Injury Scale (AIS) score and is allocated to one of six body regions (Head, Face, Chest, Abdomen, Extremities (including

U

Pelvis), External). Only the highest AIS score in each body region is used. The three most

N

severely injured body regions have their score squared and added together to produce the ISS

A

score. The ISS score takes values from 0-75. Injuries assigned with an AIS (Abbreviated

M

Injury Scale) of 6 (unsurvivable injury), automatically result in an ISS of 75 [18]. A major trauma is defined with an ISS of at least 16 [19]. Table 4 shows an example of the calculation

ED

of the ISS.

PT

Furthermore, findings from the clinical examination (CE) in the trauma room, such as blood pressure, heart rate, oxygen saturation, auscultation, external signs of injury (e.g. hematoma,

CC E

abrasions or bruises), abdominal muscular defence, stability of the thorax and pelvis as well as the results from the focussed assessment with sonography for trauma (FAST) were recorded and rated as either normal or suspicious. Age groups have been defined according to

A

the classification mentioned in the "Paediatric trauma protocols" guideline by the Royal College of Radiologists [20]: <1 year, 1-5 years, 6-11 years, 12-15 years and 16-18 years. Volume CT dose index (CTDIvol) and Dose length product (DLP) were taken from the individually documented examination protocol for every patient. CTDIvol values of the

6

head/neck scans were referring to a 16 cm body phantom and those of the thoracoabdominal scans to a 32 cm body phantom, according to the manufacturer’s standard.

Statistical analysis Statistical analysis was performed using dedicated software (BiAS 11.06 for Windows,

IP T

Epsilon Verlag, Frankfurt, Germany). The Mann-Whitney-U test was used to assess the

correlation between the two groups with (p+) and without (p-) detected pathology and CE,

SC R

FAST, type and severity of the MOI as well as the GCS values (GCSi, GCSa). Similarly, the body region subgroups were evaluated. Additionally Mann-Whitney estimator with a

U

confidence interval was performed to determine the probability of consistency of the GCS

N

values in the groups. Receiver-operation-characteristic curve (ROC) analysis was performed

A

to assess the diagnostic ability of both GCS values by varying their discrimination thresholds.

M

Chi-squared test and Fisher`s exact test were used to calculate the specificity, the sensitivity and the predictive values of CE, FAST and severity of MOI with regard to the individual body

ED

regions. Multiple regression analysis was performed to compare these variables and to determine the strongest predictive effect on pathologies regarding whole-body and individual

PT

body regions. The Kruskal-Wallis-test with post-hoc correction according to Bonferroni-Holm

CC E

was performed to compare the age groups regarding dose parameters, detected pathologies, intubation, the ISS and both GSC values. To compare the dose parameters of the two different CT scanners the Mann-Whitney-U test

A

was used. The significance level was defined at p < 0.05 for all tests that were performed.

Results Patients and history of injury 7

On average, there were 12 pediatric polytrauma patients per year that underwent WBCT during the investigation period. In total, 100 pediatric polytrauma patients that met the inclusion criteria (average age 9.1±4.4 years) were included, among them 72 male and 28 female patients (average age 9.1±4.2 and 9.2±5 years, respectively). The age group between 6-11years was the largest (n=50, p<0.001).

IP T

63 patients were status post road accidents, which was significantly the most frequent type of MOI (p>0.001). 14 patients sustained a fall from a height > 3m, 10 patients a fall from a

SC R

height <3 m and 13 patients had other types of MOI: 6 post accident in equestrian sports, 3 post brawl, 2 post railway accident and 2 post abuse. The average ISS was 15±11.5 with a

U

median value of 13.6 altogether, 14.3±11 for males and 16.8±12.6 for female patients. 43

N

patients had an ISS of at least 16, among them 30 male and 13 female patients. There was no

A

significant correlation between the type of MOI and the ISS (p>0.1). With regard to the

M

severity, there were 23 patients with a mild and 77 with a severe MOI (post accident: 53, fall from >3m: 14 and 10 others). Among the patients with a mild MOI, there were 14 in group p+

ED

(39%; 5 with ISS>16) and among the patients with a severe MOI, there were 57 in group p+ (26%; 38 with ISS>16). There was no significant correlation between p+ and the severity of

PT

MOI (p=0.61). According to the Chi-squared test and Fisher`s exact test, there were no

CC E

significant correlations between the severity of the MOI and the individual body regions: craniocerebrum (p=1.0), thorax (p=0.17), abdomen (p=0.21) and skeleton (p=0.09). However, the Mann-Whitney U test revealed significant correlations with thoracic, thoraco-abdominal

A

and skeletal pathologies (p=0.03, 0.02 and <0.001, respectively), see table 5. The GCS was lower at admission than at initial trauma scene. The average GCSi was 11.56±4.34 and the average GCSa was 10.27±5.36 with a median value of 14 for both. 31 patients were intubated between the evaluations of both GCS, automatically resulting in a GCSa of 3. Among them there were 23 patients with GCSi of less than/or equal to 8. The GCSi 8

values of the other intubated patients were 9 (n=2), 12 (n=3), 14 (n=2) and 15 (n=1). There was no statistical frequency of intubations regarding the age groups (p=0.06). Furthermore, no significant differences between the age groups and both GCS values were found (p>0.28), see table 6.

IP T

Trauma injuries 71 patients showed at least one traumatic pathology in the WBCT (p+). The subgroup analysis

SC R

regarding pathologies according the body regions revealed the craniocerebrum to be the most

common with 43% of all patients, followed by the thorax with 37%, the skeleton with 30% and the abdomen with 20%. Table 7 displays the details of trauma injuries by body regions.

U

Four patients died from the trauma injuries in the intensive-care unit. GCSa was 3 for all of

N

them, GCSi values were 3, 3, 5 and 7 and the ISS values were 34, 38, 26 and 38, respectively.

A

All of these patients had craniocerebral injuries combined with thorax injuries, but no

M

abdominal injury. One patient showed skeletal injury.

ED

Even though the age group between 6-11 years was the largest in the whole study population, there were no significant differences between the age groups regarding the number of patients

PT

with detected pathologies, neither whole-body nor the individual body regions (p>0.9 for all groups). Furthermore, no significant difference was found between the age groups regarding

CC E

the ISS (p=1), see table 6.

A

Predictive value of GCS Regarding the GCS there was a significant difference between patient groups p+ and p(p<0.01 for GCSi, and p<0.001 for GCSa ). The Mann-Whitney estimator revealed a probability of 73% that both GCS values were lower in p+. There were significant differences between p+ and p- according both GCS values regarding the subgroups craniocerebrum and 9

thorax (p<0.001). Probabilities of 77% and 83% were calculated, that GCS i and GCSa were lower for patients with craniocerebral injuries in comparison to those without. In this context, there was only a 68% probability for thorax injuries. There were no significant differences between p+ and p- regarding abdominal and skeletal injuries and the GCS (p=0.19 and 0.62 for GCSi and p=0.08 and 0.66 for GCSa), see table 8.

IP T

Regarding the whole body pathologies ROC analysis showed a discrimination threshold for

the GCSi at 12.5 with an inaccuracy rate of 30% and the GCSa at 11.5 with an inaccuracy rate

SC R

of 29%. Sensitivities were 54% and 56%, respectively and specificities were 86% for both, see figure 1. Regarding the craniocerebral subgroup the discrimination threshold for GCSi

U

and GCSa was at 12.5 with inaccuracy rates of 24% and 21%, sensitivities of 72% and 79%

N

and specificities of 81% and 79%. The inaccuracy rate increased for the other subgroups, see

M

Clinical examination and FAST

A

table 9 and figure 2.

ED

None of the patients in our study group had relevant restrictions of the oxygen saturation, blood pressure and heart rate at the time of being admitted to the trauma room. Thus, these

PT

parameters have been neglected in the statistical analysis.

CC E

There was a significant correlation between external signs of injury and craniocerebral injuries (p<0.001) with good sensitivity (79%). Thoracic pathologies and pelvic fractures also showed significant correlations with external signs of injury (p=0.03 and 0.04, respectively).

A

Regarding abdominal injuries there was no significant correlation (p=0.41). There is a slight, but significant correlation between pathological findings in the auscultation and thoracic pathologies (p=0.05). Abdominal pathologies correlated significantly with muscular defense (p=0.05) and free fluid detected in the FAST (p<0.001), both with 100% specificity.

10

Regarding free fluid, FAST correlated significantly with the findings in the WBCT (p<0.0001), see table 5. According the multiple regression model GCSa has the strongest predictive effect on craniocerebral pathologies (p<0.001), followed by external signs of injury (p=0.02). Regarding thoracic and skeletal pathologies the severity of MOI was the most predictive

IP T

variable (p=0.05 and 0.002, respectively) and FAST for abdominal pathologies (p<0.001).

None of the variables had a predictive significant effect on whole-body pathologies, see table

SC R

5.

Six patients underwent WBCT before April 2013 without severe MOI or any suspicious

U

finding in the clinical examination, in which no pathology was detected. The other 94 children

N

had at least one suspicious finding in CE (CE+) and/or a severe MOI: 17 patients with mild

A

MOI and CE+ and detected pathology; 54 patients with severe MOI and CE+, among them 44

M

with detected pathology; 23 with severe MOI but without CE+, among them 15 with detected

Dose parameters

ED

pathology.

PT

CTDI and DLP significantly decreased in the second time period between January 2012 and

CC E

November 2016 using the 64-slice CT scanner regarding both individual scans, head/neck as well as the thoracoabdominal scan (p< 0.0001, respectively). In this regard, values of the total mAs and the total DLP were lower, but non-significant (p=0.19 and 0.6, respectively).

A

Dose parameters have generally increased within the age groups, except regarding the head/neck scans of the age groups 1-5 years and 6-11 years. Significant differences were calculated between the age groups 6-11 and 16-18 years regarding the CTDI and the DLP of the head/neck scan (p=0.024 and 0.02, respectively). A slight difference has been shown between the age groups 1-5 and 16-18 years regarding the DLP of the thoracoabdominal scan 11

(p=0.05). However, all other comparisons of the dose parameters between the age groups were non-significant, see table 6.

Discussion

IP T

The results of our study underline the complexity of the management of diagnostic imaging in

pediatric polytrauma patients, especially with regard to the justification of WBCT. Even

SC R

though the GCS significantly correlated with whole-body injury, we did not find a reliable

threshold for the indication of WBCT. Furthermore, none of the other clinical variables,

U

including the severity of MOI, had a significant predictive effect on whole-body pathologies.

N

Thus, the indication for WBCT should be based on the synopsis of all clinical variables

A

available. In a prospective study Garcia and Cunningham reported that clinical suspicion in

M

addition to severe MOI is able to identify clinically significant injury (CSI) [13]. Only children with severe MOI were included and CSI was defined as injury requiring intervention

ED

or hospital admission ≥24 hours. In our study, 39% of the patients with a mild MOI showed pathologies in the WBCT, among them five with an ISS>16, indicating major trauma. Thus,

PT

even though it has been reported that only few injuries detected in CT scans required interventions [20, 21], patients with a mild MOI should not be excluded a priori from further

CC E

diagnostic imaging. As neither the type nor the severity of MOI significantly correlated with the ISS in our study, we agree that it is not recommendable as a predictor for injury severity

A

and should not be used as the sole justification for WBCT [12, 23–25]. Based on the results of the subgroup analyses, we conclude that dedicated CT scans of individual body regions might be preferable in order to reduce radiation dose. There are already several studies that provide decision rules on CT examinations of specific body regions in blunt-injured children [15, 26–29]. In this matter, our results indicate that the 12

severity of MOI seems to be helpful to identify thoracic injury as it showed the strongest predictive effect. The combination with external signs of injury and/or a pathological auscultation is even more suspicious. Thorax injuries are known to be a relevant factor in the prognosis of injured children. Similar to our results it has been reported to be the second leading pathology and cause of death in pediatric polytrauma [30, 31]. McNamara et al.

IP T

named other predictors of thorax injury, such as hypoxia, syncope/dizziness, cervical spine

tenderness, thoraco-lumbar-sacral spine tenderness, and abdominal/pelvic tenderness [26].

SC R

Regarding abdominal pathologies, FAST had the best predictive effect, which was further enhanced in combination with muscular defense and external signs of injury. It has been reported that the number of abdominal CT scans can significantly be reduced if FAST is

U

initially performed in suspected intra-abdominal injury [32, 33]. However, the sensitivities of

N

the clinical variables with significant predictive effect on thoracic and abdominal pathologies

A

were low. Thus, the risk of missed injuries remains, if the CT scan is renounced based on

M

these variables. It is uncertain if missed injuries would have a negative impact on the patient`s

ED

outcome. Prospective studies should be conducted in this matter. Regarding craniocerebral injuries, the GCS at the time of admission to the trauma room had

PT

the strongest predictive effect, followed by external signs of injury. Craniocerebral injuries are

CC E

known to be the most frequent injury in pediatric polytrauma and accompanied with high lethality and irreversible damages [34, 35]. The GCS is designated as decision support for the indication of CCT in trauma guidelines. Our statistical analysis revealed an optimum

A

threshold of the GCS at 12.5. As half values do not exist on the GCS, the indication for CCT should be determined by values of less than or equal to 13. Melo et al. reported that children with a GCS 13/14 were frequently associated with abnormalities in the CCT [15]. Biberthaler et al. also recommended CCT for children with GCS score of less than or equal to 13 [34]. However, comparable to other body regions, the relevance of detected pathologies is 13

uncertain. It should be discussed individually, if the benefit overweighs the radiation risk or if clinical observation is to be preferred. Mannix et al. as well as Pandit et al. have shown that the use of WBCT in pediatric trauma patients was significantly lower in designated pediatric centers than in adult centers, with similar mortality rates [6, 36]. In our study population, six children underwent WBCT without severe MOI or any suspicious findings in the clinical

IP T

examination. No pathology was detected. Interestingly, these cases occurred during a period up to April 2013. In the following investigation period, no child was examined without at least

SC R

one suspicious finding in the clinical examination or severe MOI. This can most probably be

explained by an increased awareness of radiation risk over the last years. In August 2012 a retrospective study of Pearce et al. was published in the Lancet, estimating an occurrence of

U

one excess case of leukemia and one excess case of brain tumor per 10 000 head CT scans in

N

the 10 years after the first scan for patients younger than 10 years [9]. The coincidence in time

A

is conspicuous, but a correlation is not evident. In the following years numerous further

M

studies have been published, informing about the risk of CT scans in early life being

ED

associated with a later risk of cancer [8, 10, 11, 37–39]. In this context, Boutis et al. reported about a considerable increase regarding the radiation dose and risk awareness among

PT

emergency physicians in Canada over the past decade [40]. This development is gratifying, as a correct indication for CT still remains the simplest way to reduce radiation dose and its

CC E

accompanying risks. However, we observed a significant decrease of the radiation dose of WBCT in the second investigation period, which is most likely explained by technical

A

advances.

This study has limitations that need to be acknowledged beyond the retrospective design. First, the small patient number, which might restrain the validity of the statistical analysis regarding the subgroups. Second, all clinical examinations including FAST and the determination of the GCS are subjective assessments performed by different investigators, 14

which might reduce the general objectivity. Third, there was no comparison group. A further comparative study with pediatric trauma patients that were admitted to the trauma room, but did not underwent WBCT would be reasonable. In conclusion, the indication of WBCT in pediatric polytrauma patients should not solely be based on one parameter, but on the synopsis of all clinical variables available. The severity of

IP T

MOI does not correlate with the severity of injury. However, in combination without suspicious findings in the clinical examination, major trauma is unlikely after mild MOI.

SC R

Dedicated CT of individual body regions might be preferable in order to reduce radiation exposure. In this regard, GCS showed good predictive diagnostic potential regarding

U

craniocerebral injuries and we recommend CCT at a GCS of less than or equal to 13. With

N

respect to injuries of other body regions, the GCS is not the appropriate index. The

A

combination of severe MOI, external signs of injury and pathological auscultation seemed to

M

be the best indicator for thoracic pathologies. Regarding abdominal injuries, FAST in combination with muscular defense and external signs of injury showed the strongest

ED

predictive effect. Further studies are necessary to assess the impact of detected pathologies regarding emergency management, interventional requirements and patient`s outcome and the

CC E

PT

benefit in relation to the radiation risk.

Declaration of interest

A

Julian L. Wichmann received speakers’ fees from GE Healthcare and Siemens Healthcare. Moritz H. Albrecht received speakers’ fees from Siemens Healthcare. The other authors have no potential conflict of interest to declare.

15

References [1] B. Auner und I. Marzi, „[Pediatric multiple trauma]“, Chir. Z. Alle Geb. Oper. Medizen, Bd. 85, Nr. 5, S. 451–461; quiz 462–463, Mai 2014. [2] I. Reichmann, M. Aufmkolk, F. Neudeck, M. Bardenheuer, K. P. Schmit-Neuerburg, und U. Obertacke, „[Comparison of severe multiple injuries in childhood and adulthood]“, Unfallchirurg, Bd. 101, Nr. 12, S. 919–927, Dez. 1998.

SC R

IP T

[3] C. Gatzka, P. G. C. Begemann, A. Wolff, J. Zörb, J. M. Rueger, und J. Windolf, „[Injury pattern and clinical course of children with multiple injuries in comparison to adults, Ab 11-year analysis at a clinic of maximum utilization]“, Unfallchirurg, Bd. 108, Nr. 6, S. 470–480, Juni 2005. [4] J. Zwingmann, H. Schmal, N. P. Südkamp, und P. C. Strohm, „[Injury severity and localisations seen in polytraumatised children compared to adults and the relevance for emergency room management]“, Zentralbl. Chir., Bd. 133, Nr. 1, S. 68–75, Feb. 2008.

N

U

[5] S. Huber-Wagner u. a., „Whole-body CT in haemodynamically unstable severely injured patients--a retrospective, multicentre study“, PloS One, Bd. 8, Nr. 7, S. e68880, 2013.

M

A

[6] V. Pandit u. a., „The use of whole body computed tomography scans in pediatric trauma patients: Are there differences among adults and pediatric centers?“, J. Pediatr. Surg., Bd. 51, Nr. 4, S. 649–653, Apr. 2016.

ED

[7] United Nations Scientific Committee on the Effects of Atomic Radiation, „Report of the United Nations Scientific Committee on the Effects of Atomic Radiation to the General Assembly“, in Sources, Effects and Risks of Ionizing Radiation, UNSCEAR 2012 Report, UN, 2015, S. 5–15.

PT

[8] J. D. Mathews u. a., „Cancer risk in 680,000 people exposed to computed tomography scans in childhood or adolescence: data linkage study of 11 million Australians“, BMJ, Bd. 346, S. f2360, Mai 2013.

CC E

[9] M. S. Pearce u. a., „Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study“, Lancet Lond. Engl., Bd. 380, Nr. 9840, S. 499–505, Aug. 2012.

A

[10] D. L. Miglioretti u. a., „The use of computed tomography in pediatrics and the associated radiation exposure and estimated cancer risk“, JAMA Pediatr., Bd. 167, Nr. 8, S. 700–707, Aug. 2013. [11] A. Berrington de Gonzalez u. a., „Relationship between paediatric CT scans and subsequent risk of leukaemia and brain tumours: assessment of the impact of underlying conditions“, Br. J. Cancer, Bd. 114, Nr. 4, S. 388–394, Feb. 2016.

16

[12] H. B. Moore u. a., „Mechanism of injury alone is not justified as the sole indication for computed tomographic imaging in blunt pediatric trauma“, J. Trauma Acute Care Surg., Bd. 75, Nr. 6, S. 995–1001, Dez. 2013. [13] C. M. Garcia und S. J. Cunningham, „Role of clinical suspicion in pediatric blunt trauma patients with severe mechanisms of injury“, Am. J. Emerg. Med., Juli 2017. [14] G. Teasdale und B. Jennett, „Assessment of coma and impaired consciousness. A practical scale“, Lancet Lond. Engl., Bd. 2, Nr. 7872, S. 81–84, Juli 1974.

IP T

[15] J. R. T. Melo u. a., „Do children with Glasgow 13/14 could be identified as mild traumatic brain injury?“, Arq. Neuropsiquiatr., Bd. 68, Nr. 3, S. 381–384, Juni 2010.

SC R

[16] K. Eichler, I. Marzi, H. Wyen, S. Zangos, M. G. Mack, und T. J. Vogl, „Multidetector computed tomography (MDCT): simple CT protocol for trauma patient“, Clin. Imaging, Bd. 39, Nr. 1, S. 110–115, Feb. 2015.

U

[17] J. F. Holmes, M. J. Palchak, T. MacFarlane, und N. Kuppermann, „Performance of the pediatric glasgow coma scale in children with blunt head trauma“, Acad. Emerg. Med. Off. J. Soc. Acad. Emerg. Med., Bd. 12, Nr. 9, S. 814–819, Sep. 2005.

A

N

[18] S. P. Baker, B. O’Neill, W. Haddon, und W. B. Long, „The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care“, J. Trauma, Bd. 14, Nr. 3, S. 187–196, März 1974.

M

[19] W. S. Copes, H. R. Champion, W. J. Sacco, M. M. Lawnick, S. L. Keast, und L. W. Bain, „The Injury Severity Score revisited“, J. Trauma, Bd. 28, Nr. 1, S. 69–77, Jan. 1988.

ED

[20] The Royal College of Radiologists. Paediatric trauma protocols. London: The Royal College of Radiologists; 2014. Ref No. BFCR (14)8. C The Royal College of Radiologists, August 2014

CC E

PT

[21] A. K. Exadaktylos, G. Sclabas, S. W. Schmid, B. Schaller, und H. Zimmermann, „Do we really need routine computed tomographic scanning in the primary evaluation of blunt chest trauma in patients with ‚normal‘ chest radiograph?“, J. Trauma, Bd. 51, Nr. 6, S. 1173–1176, Dez. 2001. [22] C. M. Smith, L. Woolrich-Burt, R. Wellings, und M. L. Costa, „Major trauma CT scanning: the experience of a regional trauma centre in the UK“, Emerg. Med. J. EMJ, Bd. 28, Nr. 5, S. 378–382, Mai 2011.

A

[23] B. D. Tojuola, X. Gu, N. R. Littlejohn, J. P. Sharpe, M. A. Williams, und D. W. Giel, „Does the mechanism of injury in pediatric blunt trauma patients correlate with the severity of genitourinary organ injury?“, Can. J. Urol., Bd. 21, Nr. 6, S. 7570–7573, Dez. 2014. [24] K. Qazi, M. S. Wright, und C. Kippes, „Stable pediatric blunt trauma patients: is trauma team activation always necessary?“, J. Trauma, Bd. 45, Nr. 3, S. 562–564, Sep. 1998.

17

[25] K. Mukherjee, M. Rimer, M. D. McConnell, R. S. Miller, und S. E. Morrow, „Physiologically focused triage criteria improve utilization of pediatric surgeon-directed trauma teams and reduce costs“, J. Pediatr. Surg., Bd. 45, Nr. 6, S. 1315–1323, Juni 2010. [26] C. McNamara, I. Mironova, E. Lehman, und R. P. Olympia, „Predictors of Intrathoracic Injury after Blunt Torso Trauma in Children Presenting to an Emergency Department as Trauma Activations“, J. Emerg. Med., Bd. 52, Nr. 6, S. 793–800, Juni 2017.

IP T

[27] N. H. Hynick, M. Brennan, P. Schmit, S. Noseworthy, und N. L. Yanchar, „Identification of blunt abdominal injuries in children“, J. Trauma Acute Care Surg., Bd. 76, Nr. 1, S. 95–100, Jan. 2014.

SC R

[28] S. N. Acker, C. L. Stewart, G. E. Roosevelt, D. A. Partrick, E. E. Moore, und D. D. Bensard, „When is it safe to forgo abdominal CT in blunt-injured children?“, Surgery, Bd. 158, Nr. 2, S. 408–412, Aug. 2015. [29] J. F. Holmes u. a., „Identifying children at very low risk of clinically important blunt abdominal injuries“, Ann. Emerg. Med., Bd. 62, Nr. 2, S. 107-116.e2, Aug. 2013.

A

N

U

[30] P. Störmann, J. N. Weber, H. Jakob, I. Marzi, und D. Schneidmueller, „[Thoracic injuries in severely injured children : Association with increased injury severity and a higher number of complications]“, Unfallchirurg, Jan. 2017.

M

[31] C. Schöneberg, B. Schweiger, M. Metzelder, D. Müller, E. Tschiedel, und S. Lendemans, „[The injured child--diagnostic work-up in the emergency room]“, Unfallchirurg, Bd. 117, Nr. 9, S. 829–841, Sep. 2014.

ED

[32] J. E. Sola u. a., „Pediatric FAST and elevated liver transaminases: An effective screening tool in blunt abdominal trauma“, J. Surg. Res., Bd. 157, Nr. 1, S. 103–107, Nov. 2009.

PT

[33] J. Menaker u. a., „Use of the focused assessment with sonography for trauma (FAST) examination and its impact on abdominal computed tomography use in hemodynamically stable children with blunt torso trauma“, J. Trauma Acute Care Surg., Bd. 77, Nr. 3, S. 427–432, Sep. 2014.

CC E

[34] P. Biberthaler und W. Mutschler, „[Diagnostic management of children with craniocerebral trauma]“, Unfallchirurg, Bd. 110, Nr. 3, S. 233–234, März 2007.

A

[35] K. M. Orzechowski, E. A. Edgerton, D. I. Bulas, P. M. McLaughlin, und M. R. Eichelberger, „Patterns of injury to restrained children in side impact motor vehicle crashes: the side impact syndrome“, J. Trauma, Bd. 54, Nr. 6, S. 1094–1101, Juni 2003. [36] R. Mannix u. a., „Factors associated with the use of cervical spine computed tomography imaging in pediatric trauma patients“, Acad. Emerg. Med. Off. J. Soc. Acad. Emerg. Med., Bd. 18, Nr. 9, S. 905–911, Sep. 2011. [37] R. W. Harbron u. a., „Cancer incidence among children and young adults who have undergone x-ray guided cardiac catheterization procedures“, Eur. J. Epidemiol., Jan. 2018. 18

[38] N. Journy u. a., „Predicted cancer risks induced by computed tomography examinations during childhood, by a quantitative risk assessment approach“, Radiat. Environ. Biophys., Bd. 53, Nr. 1, S. 39–54, März 2014. [39] E. Samei, X. Tian, W. Paul Segars, und D. P. Frush, „Radiation risk index for pediatric CT: a patient-derived metric“, Pediatr. Radiol., Bd. 47, Nr. 13, S. 1737–1744, Dez. 2017.

A

CC E

PT

ED

M

A

N

U

SC R

IP T

[40] K. Boutis und K. E. Thomas, „Radiation dose awareness and disclosure practice in paediatric emergency medicine: how far have we come?“, Br. J. Radiol., Bd. 89, Nr. 1061, S. 20160022, 2016.

19

Figures: Figure 1: ROC analysis of whole body pathologies. The discrimination threshold (cut-off) for patients with pathologies (p+) was 12.5 for GCSi (see image a) and 11.5 for GCSa (see image b). The specificities were good for both values (86%), but sensitivities were low (54

A

CC E

PT

ED

M

A

N

U

SC R

IP T

and 56%). Inaccuracy rates were also only moderate (30% and 29%).

20

IP T SC R

U

Figure 2: ROC analysis of craniocerebral pathologies. The optimal discrimination

N

threshold (cut-off) for the subgroup with craniocerebral pathologies (craniocerebrum+) was

A

12.5 for both, GCSi (see image a) and GCSa (see image b). Sensitivities (72 and 79%) and

A

CC E

PT

ED

M

specificities were good (81 and 79%) with acceptable inaccuracy rates (24% and 21 %).

21

22

A ED

PT

CC E

IP T

SC R

U

N

A

M

Tables Table 1: Scan and dose parameters. Basic scanning parameters have not been changed (-c indicates scans without, + c indicates scans with contrast medium). Dose parameters were significantly lower since January 2012 using the 64-slice CT scanner (p < 0.0001). Juli 2007- January 2012

January 2012- November 2016

Sensation 16 (Siemens)

Somatom Definition AS 64 (Siemens) Body region Head/Neck (-c) Head/Neck (-c) Chest/Abdomen (+c) Chest/Abdomen (+c) Contrast medium Weight adapted (0.5 mg Weight adapted (0.5 mg iodine/ (c) iodine/ kg) Xenetix (Guerbet) kg) Xenetix (Guerbet) 350 + 350 + 50ml NaCL chaser 50ml NaCL chaser Delay 80 seconds 80 seconds Flow 2ml/second 2ml/second Tube voltage (kV) Head/Neck

120

120

Chest/abdomen

100/120

100/120

N

U

SC R

IP T

CT-Device

A

Quality reference

M

tube current

300

Chest/abdomen

190

PT

Head/Neck

ED

(ref. mAs)

300 190

On

On

Rotation time

0.5 s

0.5 s

Collimation

16 x 1.5 mm

64 x 0.6 mm

Pitch

1.2

1.2

Data acquisition

craniocaudal

craniocaudal

inspiration in chest/abdomen

inspiration in chest/abdomen

Automated

CC E

exposure control

A

(CAREDose4D)

23

Patient number

47

53

Total mAs

10023.76 ± 4755.82

8917.04 ± 2534.78

Total

1957.95 ± 474.44

1812.15 ± 612.84 (p = 0.6)

Head/neck

1658.46 ± 348,34

1322.61 ± 440,87 (p < 0.0001)

Chest/Abdomen

505.54 ± 211,74

255.09 ± 134,36

Head/neck

58.55+-9,28

46.58 ± 10.37

(p < 0.0001)

Chest/Abdomen

7.97+-2.87

5.04 ± 2.48

U

(p < 0.0001)

Average

(p = 0.19)

DLP

CTDI

(p < 0.0001)

SC R

Average

IP T

(mGy*cm)

A

CC E

PT

ED

M

A

N

(mGy)

24

Table 2: Determination of the Glasgow Coma Scale (GCS) and the Pediatric Glasgow Coma Scale (PGCS).

Score

Spontaneously

Spontaneously

4

To verbal command

To shout

3

To pain

To pain

2

No response

No response

1

Obeys

Spontaneous

Localizes pain

Localizes pain

Flexion-withdrawal

Flexion-withdrawal

4

Flexion-abnormal (decorticate rigidity)

Flexion-abnormal

3

IP T

Motor response

< 1 year

6 5

SC R

Eye opening

> 1 year

(decorticate rigidity) Extension (decerebrate rigidity)

Extension

2

U

(decerebrate rigidity)

N

No response

Oriented

Appropriate

M

Verbal response

2-5 years

A

> 5 years

ED

Disoriented/confused

PT

Inappropriate words

CC E

Incomprehensible

1

0-23 months Smiles/Coos

5

words/phrases

appropriately

Inappropriate

Cries and is consolable

4

words Persistent cries Persistent and screams

inappropriate

3 crying

and/or screaming Grunts

sounds No response

No response

Grunts, agitated and 2 restless

No response

No response

1

A

Total GCS/PGCS

25

Table 3: Identification of a polytrauma patient. These parameters help to identify patients with potentially life-threatening injuries, which justify whole body CT (WBCT).

Severe cause of trauma

Injury pattern

Vital parameter

Fall from >3m

Unstable thorax

Systolic blood-pressure < 90 mmHg (adapted to age in children)

Open

or

penetrating Altered consciousness

injury of the thorax

or

coma

IP T

Ejection from a vehicle

(Glascow

Coma Scale < 10) Dyspnoea

cyclist

combination

SC R

Crashes as pedestrian or Unstable pelvic fracture

in

with

concomitant injury

U

High-speed accident with Fractures of at least two motorbike or car (>30-50 long bones

N

km/h)

amputation of proximal

Exposure to high energy

M

extremities

A

Entrapment or burial

Serial rib fracture

with

ED

combination

in

concomitant injury

A

CC E

PT

Death of passenger

26

Table 4: Example of the calculation of the ISS: 6 year old boy post severe road accident with critical craniocerebral and thoracic injuries, but no other injuries, resulting in an ISS of 50.

Injury description

AIS

Square Top Three

Head & Neck

Open head injury

5

25

Face

No injury

0

-

Chest

Pneumothorax

5

25

No injury

Extremity

No injury

0

External

No injury

0

-

N

U

Abdomen

SC R

Pulmonary Contusion

IP T

Region

50

A

CC E

PT

ED

M

A

Injury Severity Score

27

I N U SC R

Table 5: Clinical examination, FAST and severity of MOI: Chi-squared Test, Fisher`s exact Test, specificity, sensitivity and predictive values of clinical examinations, FAST and severity of MOI regarding the individual body regions (+ indicates with pathology). Multiple regression revealed the variable with the

External injury (n=53)

M

PT

n= number of p-values patients Cranio-cerebrum+ (n=43):

p-value Sensitivity Chi-squared (%) test *= Fisher`s exact Test

Specificity (%)

Pos. pred. value

Neg. pred. value

< 0.0001

< 0.0001

79

66.7

64.15

80.85

External injury (n=20)

0.018

0.033

32.4

87.3

60

68.75

Auscultate (n=3)

0.02

0.09/*0.05

8

100

100

64.94

0.19

0.79/*0.37

97.3

/

36.36

/

CC E

Multiple regression model

ED

MannWhitney-U test

A

strongest predictive effect on injuries regarding the individual body regions. No significant effect was found regarding whole-body injury (p+).

A

Thorax+ (n=37):

Stability (n=1)

28

I External (n=8)

injury

93.75

37.5

81.52

100

100

81.63

20

100

100

83.33

0.69/*0.042

29.4

90

38.46

86.2

0.01

0.27/*0.13

92.3

/

12.12

/

P+ (n=71)

0.22

0.61/*0.54

62.5

27.4

14.1

79.31

/ n.s.

Head/neck+ (n=43)

0.96

1.0

43.75

42.85

16.28

84.21

GCSa <0.0001

Thorax+ (n= 37)

0.03

0.17

18.75

59.52

8.1

79.36

Severity of

0.41/*0.2

15

0.004

0.05/*0.04

10

< 0.0001

0.0006/*0.001

injury

PT

Pelvis+ (n=13):

CC E

0.03

Stability (n=1)

ED

FAST (n= 4)

External (n=17)

A

defense

M

Muscular (n=2)

0.2

N U SC R

Abdomen+ (n=20):

A

Severity of MOI (n= 84 of severe)

29

I 0.12

0.27/*0.18

6.25

0.02

0.09

25

<0.001

0.09/*0.07

77.38

5

81.25

FAST <0.0001

48.8

8.51

77.35

Severity of MOI 0.04

78.57

/

80.5

M

A

Thorax/Abdomen+ (n=47)

N U SC R

Abdomen+ (n=20)

/

Severity of MOI 0.002

A

CC E

PT

ED

Skeleton+ (n=18)

MOI 0.05

30

I N U SC R

Table 6: Age groups: Statistical results grouped by patient`s age groups regarding GCSi and GCSa, intubation, pathologies, ISS as well as dose parameters. *The age group between 6-11 years was significantly the largest (p>0.001).

N

GCSi

total

100

<1 1-5

1 21

5 10.67± 4.23

6-11

50

12-15

18

12.68± 3.76 10.22± 5.37

16-18

10

p-value

*

GCSa

N of in.

P+

31

71

3 8.81± 5.54

1 9

1 17

11.08± 4.96 10.83± 5.61

10

ISS

Total mAs

Total DLP (mGycm)

A

10.9± 4.51 >0.28

26 16.52 ±12.8 4 13.74 ±9.86 13.61 ±11.5 4 19.5± 16.53 =1

12

9±6.32

5

8

>0.98

> 0.5 6

=1

CTDI head/ Neck (mGy)

DLP Head/ Neck (mGycm)

CTDI chest/ Abdomen (mGy)

DLP chest/ Abdomen (mGycm)

373.99± 215.04 139 313.31± 222.9

9431.27± 3797.51 3119 9056.43± 1720.08

1877.46± 562.57 722 1853.14± 501.96

52.14± 11.51 39.15 52.0± 11.9

1478.66±4 32.66 578 1444.53±4 11.73

6.43±3.04

8904.16± 3574.52 10977.28 ±6152.35

1804.33± 619.46 1941.81± 523.61

49.58± 11.7 53.85± 10

1397.41±4 65.33 1551.85±3 49.9

6.31±2.95

10206.1± 1769.96 >0.37

2194.8± 408.64 >0.27

61.27± 8.25 3 vs.5 = 0.024

1818.1± 313.26 3 vs. 5 = 0.02

8.56±3.48

Others >0.3

Others >0.13

ED 33

PT

CC E

6

M

A

Age group (years)

3.46 5.37±2.99

6.59±2.64

>0.08

350.91± 197.47 402.06± 193.51 547.45± 256.42 2 vs. 5 = 0.05 Others >0.15

31

I N U SC R

Table 7: Detected pathologies in the WBCT. Number of trauma injuries by body regions.

Injuries #

Injuries #

A

Total 334

M

Abdomen

Head/neck

Open 16

skull-brain

Parenchymal 12

CC E

Edema 5

Subarachnoid 16

A

Subdural 17 Epidural 3

Intraventricular 12 Skull

of

ED

sign

Small trauma 0

PT

External 53

External injury 20

sign

of

bowel

injury injury

Colon contusion 0

injury

Liver 4

laceration

Splenic hemorrhage 7

laceration

Kidney hemorrhage 1

laceration

Active hemorrhage 8

bleeding

Pelvis hemorrhage fracture

External 13

sign

of

injury

Efferent

urinary

tract

injury

32

I 1

Facial 13

fracture Sacral 4 fractures Pelvic 6

A

Vertebral 2

sign

injury

fractures

fracture

Lower 1

extremity

fracture

ED

extremity

laceration

Pericardial 3

effusion

A

Pulmonary 3

6 (single)/3

Clavicle 6

fracture

Rib fracture(s) (serial)

spine

contusion Upper 1

PT

CC E

Hemothorax 2

vertebral

fracture(s)

Proximal Extremities

Pulmonary 26 Pneumothorax 14

of

Thoracolumbar 14

ramus

fracture

M

Thorax External 20

N U SC R

22

33

I N U SC R

Table 8: Mann-Whitney-U test and Mann-Whitney estimator. The correlation between detected pathologies and both GCS values (GCSi, GCSa) was calculated for the whole body group and the subgroups according to the body regions. *It was most significant for the craniocerebrum with the highest consistencies according to the Mann-Whitney-estimator, that both GCS values were lower for patients with pathologies.

GCSi

A

Body region

ED

M

p-value

Mann-Whitneyestimator for P {(X|group1)<(Y|g roup2)}

WBCT

0.010153

0.663672

0.000304

0.726566

*Craniocerebrum

0.000002

0.770502

< 10-6

0.826805

Thorax

0.004450

0.669670

0.002328

0.681467

Abdomen

0.199267

0.593750

0.075033

0.629375

Skeleton

0.618772

0.531905

0.661616

0.528333

PT CC E A

Mann-Whitney- GCSa estimator for P p-value {(X|group1)<(Y |group2)}

34

I N U SC R

Table 9: ROC analysis: The GCSi and GCSa with an optimal ratio of sensitivity and specificity, thus indicating the best prognostic value of the injury risk, were calculated. Due to the low sensitivities and moderate inaccuracy rates, results were considered not reliable for whole body pathologies (p+), even though the specificities were good. *Regarding the craniocerebrum the discrimination threshold at 12.5 for both GCS values showed good sensitivities and specificities

A

with acceptable inaccuracy rates. + indicates that at least one traumatic pathology has been detected in the respective groups.

Optimal discrimination treshold

Sensitivity

GCS values

GCSi

GCSi

p+

12.5

*Craniocerebrum+

12.5

GCSi

GCSa

GCSi

GCSa

ED

53.5%

56.3%

86.2%

86.2%

30.1%

28.7%

12.5

72.1%

79.1%

80.7%

78.9%

23.6%

21.0%

14.5

54.1%

75.7%

79.4%

55.6%

33.3%

34.4%

10.5

10.5

55.0%

65.0%

72.5%

65.0%

36.2%

35.0%

11.5

11.5

50.0%

53.3%

68.6%

60.0%

40.7%

43.3%

A

Skeleton+

GCSa

10.5

CC E

Abdomen+

Inaccuracy rate

11.5

PT

Thorax+

GCSa

Specificity

M

Body region

35