a b s t r a c t

Background: Trauma level designation and verification are examples of healthcare regionalization aiming at im- proving patient outcomes. This study examines impact of Trauma Levels on survival of patients arriving with “no signs of life” to US trauma centers.

Methods: This retrospective study used the US National Trauma Data Bank 2015 dataset. A descriptive followed by a bivariate analysis was done comparing variables by the Trauma designation levels. A multivariate analysis assessed the effect of the trauma designation on survival to hospital discharge after controlling for poten- tial confounding factors.

Results: 6160 patients without signs of life were included. The average age was 40.66 years (+-19.96) with male predominance (77.3%). Most patients were transported using ground ambulance (83.5%) and were taken to Level I (57%) and Level II (32.4%) centers. Blunt injuries were the most common (56.9%). motor vehicle collision (38.5%) and firearm (33.8%) were the most common mechanisms of injury. Survival to hospital discharge among patients with no signs of life ranged from 13.7% at Level I to 27.9% at Level III. After adjusting for confounders, in- cluding Injury Severity Score (ISS), higher survival was noted at level II trauma centers compared to Level I. Conclusions: Patients presenting without signs of life to Level II trauma centers had higher survival to hospital dis- charge compared to Level I and Level III centers. These findings can guide future prehospital triage criteria of trauma patients in organized Emergency Medical Services (EMS) systems and highlight the need for more out- come research on Trauma systems.

(C) 2019

Introduction

Healthcare regionalization aims at improving outcomes of patients and minimizing cost of care [1]. More specifically, trauma systems in the United States are often “the Most frequently cited example of suc- cessful regionalization of healthcare” [2]. Elements of this system in- clude state designation of a trauma center and verification of level by organizations such as the American College of Surgeons (ACS). Both designation and verification use similar criteria to categorize trauma centers. The process by which verification takes place consists of consul- tation visits and multi-disciplinary review, which can take up to one year [3]. After meeting all of the criteria and correcting deficiencies, a hospital can be verified as an ACS Trauma Center [3].

Impact of trauma center verification on outcomes has been previ-

ously examined in specific patient populations with conflicting results.

* Corresponding author at: Department of Emergency Medicine, American University of Beirut Medical Center, P.O. Box - 11-0236, Riad El Solh, Beirut 1107 2020, Lebanon.

E-mail addresses: [email protected] (A. Dakessian), [email protected] (R. Bachir), [email protected] (M. El Sayed).

One study found lower mortality and lower severe disability at dis- charge for patients with high Injury Severity Score (ISS N 15) when treated at Level 1 trauma centers compared to lower level centers [4]. Another study compared outcomes (mortality and complications) for trauma patients treated at ACS vs non-ACS trauma centers and found no differences in mortality but less complications in ACS centers [5]. Im- proved survival and lower progression of neurologic insult were also re- ported for patients with severe traumatic brain injury treated at ACS- designated level I trauma centers compared to Level II centers [6]. While one study reported improved survival of patients with severe trauma (ISS >= 25) when treated at a trauma center [7], another study that examined outcomes in patients who underwent early thoracotomy reported that “ACS trauma center designation did not significantly im- pact mortality” [8].

Patients who present with no signs of life from a traumatic injury are expected to benefit the most from care at designated trauma centers be- cause of availability of resources and of experienced staff. These pa- tients, mainly those with traumatic arrests, are often excluded from studies examining outcomes in trauma patients.

This study evaluates outcomes of trauma systems by examining the impact of trauma center designation and level verification on survival to

https://doi.org/10.1016/j.ajem.2019.158390

0735-6757/(C) 2019

hospital discharge of trauma patients presenting with no signs of life to US trauma centers.

Methods

The National Trauma Data Bank is the largest registry in the United States for trauma patients [9]. Data is gathered and released on an an- nual basis and the 2015 data set contains records from over 900 facilities in the United States. The data bank has compiled over 6 million records since its establishment [10]. Patients are included in the dataset when they are injured as a result of trauma and when they meet the criteria of International Classification of Diseases (ICD) codes pertaining to spe- cific injuries (excluding superficial injuries). The databank includes pre- hospital information, patient demographics, injury details and severity, and ED and hospital information, diagnosis, outcome and payment method [11].

This retrospective cohort study utilized the 2015 dataset of the NTDB. Patients who had presented to the Emergency Department (ED) with “no signs of life” were considered eligible for inclusion in the study. “Signs of life” is coded as a separate variable in the trauma dataset and is defined as the lack of “organized EKG activity, pupillary responses, spontaneous respiratory attempts or movement and unas- sisted blood pressure” on arrival to the ED [11]. Patients were consid- ered adults if their age was 16 years and above similar to other trauma studies [12]. Patients with undetermined age (value = -99) were excluded. Patients with unknown outcomes including disposition from the ED (not known, not applicable, against medical advice), discharged home directly from ED (with or without services), unknown hospital disposition and transfers (inter-hospital facility, institutional care, mental health facility and jail) were also excluded. Patients discharged against medical advice from the hospital were not excluded and were considered alive at hospital discharge. Trauma Level designa- tion combined both ACS (adult or pediatric) and State Designation. The higher level from that combination was taken as the Trauma Center Level [13]. Patients taken to unverified ACS and unverified State- Designated facilities were excluded. Fig. 1 shows the numbers of in- cluded and excluded patients. The Institutional Review Board (IRB) at

the American University of Beirut provided an exemption for the use of a de-identified dataset.

Statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS, version 24). Descriptive analysis was con- ducted to present the frequencies and percentages of the categorical variables and to display the mean +- SD, median, and interquartile range (IQR) of the continuous ones. Depending on the size sample, Pearson’s Chi-Square or Fishers’ exact tests were used to compare the proportions of all categorical variables according to the trauma designa- tion levels. Homogeneity of the variances of the continuous variables was estimated using the Levene’s test. To evaluate the difference of the continuous variables among the trauma designation levels, the one-way analysis of variance (ANOVA) with LSD post hoc test was used where the variances were homogenous, whereas Welch ANOVA was preferred with Games-Howell post hoc test where the assumption of homogeneity of variances was violated. A multivariate logistic regres- sion using a backward selection procedure was conducted to determine the impact of trauma designation levels on patient survival. P-value of

<=0.05 was used to denote statistical significance.

Results

The total number of patients in the dataset was 917,865 and those who presented to the ED with no signs of life were 8026. A total of 6160 patients with no signs of life on arrival to ED were included in this study. The majority were adults (94.9%) with a mean age of

40.66 years (+-19.96) and of male gender (77.3%). More than half of the patients who presented with no signs of life were White (54.5%), followed by Black/African American (31.5%). They were also predomi- nantly healthy, with only 28.7% having co-morbidities. Most of these pa- tients were transported to the hospital using ground ambulance (83.5%). Regarding Trauma Level designation, most of the patients were taken to Level I (57.0%) and Level II (32.4%), with the rest being 9.9% and 0.7% for Levels III and IV respectively. Only 11.5% of patients survived to hospital discharge (Table 1).

Blunt trauma was more common than penetrating injury (56.9% vs 38.5%). Injuries were also mostly unintentional (58.9%) or a result of

NTDB 2015

Dataset

(n=917,865)

Excluded (n=909,839)*

-Patients presenting with signs of life (n=909,839)

Included in Univariate Analysis

(n=6,160)

-Patients with unknown age (n=443)

-Patients with unknown dispositions (n=1,459)

Included in Bivariate and Multivariate Analyses

(n=6,114)

-Patients with unverified trauma designation (n=165)

Fig. 1. Inclusion and exclusion flowchart.

Table 1 General characteristics.
	Trauma designation level				p-Value
	All levels	I	II	III
	(N = 6160)	(N = 3509)	(N = 1995)	(N = 610)
Agea (Mean +- SD)	40.66 +- 19.96	39.04 +- 18.73	41.71 +- 21.02	45.33 +- 21.44	b0.001
Gender
Male	4760 (77.3%)	2765 (78.8%)	1529 (76.7%)	439 (72.0%)	0.001
Female	1398 (22.7%)	743 (21.2%)	465 (23.3%)	171 (28.0%)
Race White	3357 (54.5%)	1727 (51.3%)	1177 (62.2%)	419 (71.6%)	b0.001
African-American	1942 (31.5%)	1314 (39.0%)	508 (26.8%)	113 (19.3%)
Asian	104 (1.7%)	59 (1.8%)	40 (2.1%)	4 (0.7%)
Otherb	487 (7.9%)	266 (7.9%)	168 (8.9%)	49 (8.4%)
Co-morbidities
Yes	1768 (28.7%)	918 (26.2%)	634 (31.8%)	197 (32.3%)	b0.001
No	4392 (71.3%)	2591 (73.8%)	1361 (68.2%)	413 (67.7%)
Mode of transportation
Ground ambulance	5141 (83.5%)	2907 (83.0%)	1686 (84.9%)	513 (84.4%)	b0.001
Helicopter/fixed-wing	565 (9.2%)	346 (9.9%)	193 (9.7%)	26 (4.3%)
Public/private vehicle	289 (4.7%)	120 (3.4%)	92 (4.6%)	66 (10.9%)
Police & other	149 (2.4%)	130 (3.7%)	16 (0.8%)	3 (0.5%)
Patient disposition (ED)
Deceased/expired	4610 (74.8%)	2700 (76.9%)	1493 (74.8%)	406 (66.6%)	b0.001
Admitted to the hospital	1044 (16.9%)	507 (14.4%)	355 (17.8%)	148 (24.3%)
Operating room	506 (8.2%)	302 (8.6%)	147 (7.4%)	56 (9.2%)
Patient disposition (Hospital) Discharged home	707 (11.5%)	353 (10.1%)	233 (11.7%)	102 (16.7%)	b0.001
Deceased/expired	521 (8.5%)	330 (9.4%)	157 (7.9%)	34 (5.6%)
Transferred to other destination	313 (5.1%)	121 (3.4%)	111 (5.6%)	65 (10.7%)
Left against medical advice	9 (0.1%)	5 (0.1%)	1 (0.1%)	3 (0.5%)
Not applicable (died in ED)	4610 (74.8%)	2700 (76.9%)	1493 (74.8%)	406 (66.6%)
Died in ED/hospital Yes	5131 (83.3%)	3030 (86.3%)	1650 (82.7%)	440 (72.1%)	b0.001
No	1029 (16.7%)	479 (13.7%)	345 (17.3%)	170 (27.9%)

^a Games-Howell post hoc test was used to identify the significant groups: 1 b 2 b 3.

^b American Indian, Native Hawaiian or other Pacific islander, other race.

assault (28.8%). Common injury mechanisms included MVC (38.5%), firearm (33.8%) and fall (12.0%). Injuries occurred mainly in Public Build- ing/Street/Recreation (57.4%) or at Home/Residential Institution (25.6%). These injuries resulted in fractures in 58.9% of patients, internal organ damage in 56.4% and open wounds in 50.0%. The most common injured site was reported to be the torso (58.9%), head and neck (53.2%), and ex- tremities (43.4%). The median amount of time taken from the dispatch of EMS to arrival of patient in the ED was 34 min (IQR 25-51) and the me- dian Injury Severity Score was 19 (IQR 9-30) (Table 2).

In the bivariate analysis, patients taken to Level I centers were youn- ger, healthier and more likely to be African American. Survival to hospi- tal discharge was higher in Level II (17.3%) and III (27.9%) compared to Level I centers (13.7%) (Table 1).

Patients taken to Level I centers had more Penetrating injuries, with more assault, and sustained injuries from firearm mechanism. Their in- juries were more likely to occur in Public places and they were more likely to sustain fractures, internal Organ injuries, open wounds and blood vessels injuries. They also had more torso and extremities inju- ries. Patients taken to Level I centers had also higher ISS (Table 2).

After adjusting for important confounders including patient demo- graphics, injury details including mechanism, type of injury and ISS, mode of transportation, time taken to arrive to ED and hospital details, patients with no signs of life who were treated at a Level II trauma cen- ter were more likely to survive to hospital discharge compared to those treated at Level I centers (OR = 1.279, 95% CI: 1.018-1.608). There was no significant difference in survival between Level I and Level III centers in this category of patients (Table 3).

Discussion

This is the first study that evaluates the impact of trauma level des- ignation on the survival of patients who presented to US trauma centers

without signs of life. Survival rates among patients with no signs of life ranged from 13.7% at Level I trauma centers to 27.9% at Level III centers. After adjusting for confounders, including clinical severity (ISS), higher survival was noted at Level II trauma centers compared to Level I.

This finding of increased survival for trauma patients treated at Level II compared to Level I centers has been previously reported in the liter- ature. In a study conducted using data from the Los Angeles County Trauma and Emergency Medical System lower overall mortality was found in Level II compared to Level I in both urban and suburban cen- ters. In that study, all adult trauma patients regardless of injury severity were included [14]. Potential factors contributing to this observed sur- vival benefit at Level II centers may not be related to criteria used for designation of trauma centers. A previous study examining components of regionalized trauma care in Canada identified prehospital notification and presence of performance improvement programs as important pre- dictors of survival [15]. Both of these components were not directly measured in our study. Similarly residency training programs which are more available in Level I centers have been previously associated with poorer outcomes in trauma patients [14]. These factors should be examined more closely in future studies identifying significant associa- tions with outcomes in regionalized trauma systems.

Other studies however reported outcome benefits in Level I centers when compared to Level II. In a previous study that used data from the Ohio Trauma Registry, patients taken to Level I centers had im- proved survival compared to their Level II counterparts [16]. That study was however representative of one state, whereas the NTDB used in our study contains information from several states. The study also used ACS verification alone, while our study additionally incorpo- rated State Designation Level. Moreover, it included all patients sustain- ing regardless of injury severity. Another study examined outcomes in patients with specific severe injuries and identified decrease in mortal- ity in Level I centers compared to Level II in severe traumatic injuries

Table 2 Event & injury characteristics.
	Trauma designation level				p-Value
	All Levels	I	II	III
	(N = 6160)	(N = 3509)	(N = 1995)	(N = 610)
Nature of trauma
Blunt	3321 (56.9%)	1766 (53.3%)	1155 (60.7%)	364 (63.1%)	b0.001
Penetrating	2247 (38.5%)	1402 (42.3%)	668 (35.1%)	172 (29.8%)
Burn	72 (1.2%)	45 (1.4%)	20 (1.1%)	6 (1.0%)
Other & unspecified	196 (3.4%)	100 (3.0%)	59 (3.1%)	35 (6.1%)
Missing 324 (5.3%)
Injury intentionality
Unintentional	3439 (58.9%)	1827 (55.1%)	1172 (61.6%)	400 (69.3%)	b0.001
Assault	1681 (28.8%)	1102 (33.3%)	474 (24.9%)	104 (18.0%)
Self-inflicted	548 (9.4%)	296 (8.9%)	192 (10.1%)	58 (10.1%)
Undetermined & other	168 (2.9%)	88 (2.7%)	64 (3.4%)	15 (2.6%)
Missing 324 (5.3%) Mechanism of injury MVCa	2246 (38.5%)	1285 (38.8%)	755 (39.7%)	203 (35.2%)	b0.001
Firearm	1971 (33.8%)	1232 (37.2%)	585 (30.8%)	149 (25.8%)
Fall	703 (12.0%)	314 (9.5%)	251 (13.2%)	108 (18.7%)
Cut/pierce	276 (4.7%)	170 (5.1%)	83 (4.4%)	23 (4.0%)
Struck by, against	127 (2.2%)	55 (1.7%)	48 (2.5%)	23 (4.0%)
Transport, other & unspecified	187 (3.2%)	81 (2.4%)	79 (4.2%)	26 (4.5%)
Other specified classifiable & other	90 (1.5%)	46 (1.4%)	27 (1.4%)	16 (2.8%)
Specified not elsewhere classifiable Otherb	236 (4.0%)	130 (3.9%)	74 (3.9%)	29 (5.0%)
Missing 324 (5.3%) Nature of injury
Fractures	3628 (58.9%)	2062 (58.8%)	1208 (60.6%)	325 (53.3%)	0.006
Internal organ	3476 (56.4%)	2076 (59.2%)	1121 (56.2%)	270 (44.3%)	b0.001
Open wounds	3080 (50.0%)	1835 (52.3%)	1006 (50.4%)	233 (38.2%)	b0.001
Blood vessels	1003 (16.3%)	635 (18.1%)	308 (15.4%)	60 (9.8%)	b0.001
System wide & late effects	92 (1.5%)	67 (1.9%)	19 (1.0%)	6 (1.0%)	0.011
Otherc	1102 (17.9%)	626 (17.8%)	356 (17.8%)	113 (18.5%)	0.917
Region of injury

Torso	3627 (58.9%)	2151 (61.3%)	1181 (59.2%)	284 (46.6%)	b0.001
Head & neck	3279 (53.2%)	1870 (53.3%)	1107 (55.5%)	293 (48.0%)	0.005
Extremities	2673 (43.4%)	1546 (44.1%)	864 (43.3%)	235 (38.5%)	0.039
Spine & back	899 (14.6%)	530 (15.1%)	301 (15.1%)	66 (10.8%)	0.018
Forearm and elbow	1 (0%)	1 (0%)	0 (0%)	0 (0%)	1.000
Unclassifiable Injury severity scored	490 (8.0%)	300 (8.5%)	149 (7.5%)	38 (6.2%)	0.090
Mean +- SD	23.76 +- 20.81	24.98 +- 20.93	23.52 +- 20.64	18.44 +- 19.70	b0.001

^a MVC motorcyclist & MVC occupant & MVC Other & MVC Pedal cyclist & MVC pedestrian & MVC unspecified.

^b Drowning/submersion & fire/flame & hot object/substance & machinery & pedal cyclist, other & pedestrian, other & natural/environmental, bites and stings & natural/environmental, other & overexertion & poisoning & suffocation.

^c Amputations, burns, crush, dislocation, nerves, sprains & strains and unspecified.

^d LSD post hoc test was used to identify the significant groups: 1 N 2 N 3.

pertaining to vascular, cardiac, liver, pelvic and neurologic damage [4]. This study however excluded patients who were dead on arrival to the ED. Similar findings of higher mortality in Level II compared to Level I (ACS verification only) were also reported in patients in Trau- matic Brain Injury (TBI) who ever alive on admission (OR = 1.57) [6]. Inclusion criteria were therefore different in all these studies that re- ported improved in outcomes in Level I compared to Level II.

In the United States, trauma centers are classified into levels I, II, II, IV or V depending on resources and patient load. Referral networks are usually established to transfer patients from lower to higher level cen- ters. Different elements are also considered for classification including

availability of specialist coverage, educational capabilities for trauma team members, quality assessment and community EDucation. The main difference between Level I and II centers is that Level I centers must receive a minimum annual volume of patients who are severely injured. Other differences include the presence of lower number of in- house specialists in Level II centers, in addition to lack of some tertiary care needs [17]. The presence of in-house physicians in trauma centers was previously shown not to impact clinical outcomes [18]. Further- more, the average total readiness cost for Level I trauma centers is N3 times of that of Level II centers [19]. Higher level centers are therefore expected to receive higher volumes of patients which should translate

Table 3

Logistic regression model of survival.

Trauma level (I)	OR (95% CI)	p-Value	Adjusted OR (95% CI)	Adjusteda p-value
II	1.323 (1.137-1.538)	b0.001	1.279 (1.018-1.608)	0.035
III	2.444 (1.998-2.990)	b0.001	1.046 (0.740-1.478)	0.800

^a Variables that were included in the model are: age, gender, race, trauma designation level, geographic region for the hospital, ICD-9-CM Mechanism of Injury E-Code, indication of the type (nature) of trauma produced by an injury, Injury Intentionality as defined by the CDC Injury Intentionality Matrix, location where injury occurred, the patient’s primary method of payment, mode of transportation, the Injury Severity Score reflecting the patient’s injuries directly submitted by the facility regardless of the method of calculation, total elapsed time from dispatch of the EMS transporting unit to hospital arrival of the EMS transporting unit, ICD-9 body region as defined by the Barell injury diagnosis Matrix (blood vessels, fractures, internal organ, open wounds), Nature of injury as defined by the Barell Injury Diagnosis Matrix (extremities, head & neck, spine & back, torso).

into increased providers’ experience and improved outcomes. This ad- vantage or increase in survival was not however identified in this study when examining patients with no signs of life which highlights the need for more outcome research to guide resource allocation in or- ganized trauma systems.

Examining outcomes in specific trauma population can also help support more evidence-based field triage guidelines. In the United States, the decision regarding transferring patients to trauma centers is taken by Emergency Medical Services (EMS), using guidelines for field triage published by the Centers for Disease Control and Prevention (CDC). It is a multistep process that is based on physiologic and ana- tomic criteria. It also takes into account the mechanism by which the pa- tient was injured and special considerations for underlying and co- morbid conditions. severely injured patients who do not have any signs of life i.e. GCS <=13 or systolic blood pressure b90 mmHg or respi- ratory rate b10/min are taken to centers with the highest level of care within the trauma system, which are usually Level I centers [20]. Our study revealed that survival rates of patients with no signs of life who were taken to a Level II center are actually higher than those of Level I which would question the need to transfer such patients to the highest level of care center (Level I), when a Level II center is more proximal geographically. The survival rate reaching up to 27.9% in some centers is in line with specific guidelines recommending aggressive resuscita- tion of patients with traumatic arrest or near arrest in the absence of in- juries or wounds that are obviously incompatible with life [21].

Our study identified differences in survival between Patient transported to other designated trauma centers and did not examine the benefits of trauma center designation itself. Other studies have shown conflicting results when trauma centers are compared with non-trauma ones. One study concluded that there was no difference in mortality of adults between ACS and non-ACS verified centers when they compared adults with minor and major trauma. Moreover, adult patients who sustained major trauma and were taken to an ACS verified center were actually more likely to die compared to non-ACS ones [5]. Whereas another study concluded that patients sustaining severe inju- ries were less likely to die when in Level I centers compared to NTC (nontrauma centers), and also in Level II centers compared to NTC [7]. More evidence is therefore needed for the benefit of not only assigning levels of trauma level designation but also of designation itself.

Potential limitations to our study are related to the dataset used and to its retrospective nature. Firstly, the NTDB excludes patients who died on scene and were not taken to the ED which might overestimate the survival rate in all centers. Moreover, some prehospital traumatic ar- rests could have been excluded depending on which US State they oc- curred in. Details were lacking with regards to the exact time of cardiac arrest, the study analysis however adjusted for total elapsed time from EMS dispatch until arrival at the hospital. Other limitations are related to differences in documentation from one hospital to an- other in the NTDB. However, the data is continuously monitored and cleaned in order to ensure quality with “outlier hospitals” constantly identified and thoroughly reviewed [10]. This study also examined sur- vival only as an outcome since additional morbidity outcomes of pa- tients who were discharged alive were lacking (disability, cerebral performance and functional status).

Despite these limitations, this study examined the effect of Trauma Level designation on survival using the population with highest risk of mortality, those presenting with no signs of life. It also used the largest trauma registry in the Unites States and its findings can be generalized to settings with similar advanced trauma systems.

Conclusion

Trauma systems are designed to improve patient outcomes. Survival rates of patients who sustain life threatening injuries and who present

to designated trauma centers with no signs of life can reach up to 27.9%. Higher survival was noted in this population for patients treated at Level II trauma centers compared to Level I centers. The findings of this study can help guide future prehospital triage criteria of trauma pa- tients in organized EMS systems and highlight the need for more out- come research when examining trauma systems.

Funding statement

This research did not receive any specific grant from funding agen- cies in the public, commercial, or not-for-profit sectors.

References

1. Luft HS. Regionalization of medical care. Am J Public Health 1985;75:125-6.
2. Carr BG, Asplin BR. Regionalization and emergency care: the Institute of Medicine Reports and a Federal Government Update. Acad Emerg Med 2010;17:1351-3. https://doi.org/10.1111/j.1553-2712.2010.00944.x.
3. Verification, review, and consultation (VRC) program [Internet]. American College of Surgeons Available from: https://www.facs.org/quality-programs/trauma/vrc.
4. Demetriades D, Martin M, Salim A, Rhee P, Brown C, Chan L. The effect of trauma center designation and trauma volume on outcome in specific severe injuries. Ann Surg 2005;242:512-7. https://doi.org/10.1097/01.sla.0000184169.73614.09.
5. Grossman MD, Yelon JA, Szydiak L. Effect of American College of Surgeons trauma center designation on outcomes: measurable benefit at the extremes of age and in- jury. J Am Coll Surg 2017;225:194-9. https://doi.org/10.1016/j.jamcollsurg.2017.04. 034.
6. Dubose JJ, Browder T, Inaba K, Teixeira PG, Chan LS, Demetriades D. Effect of trauma center designation on outcome in patients with severe traumatic brain injury. Arch Surg 2008;143:1213-7. https://doi.org/10.1001/archsurg.143.12.1213.
7. Vickers BP, Shi J, Lu B. Comparative study of ED mortality risk of US trauma patients treated at level I and level II vs nontrauma centers. Am J Emerg Med 2015;33: 1158-65. https://doi.org/10.1016/j.ajem.2015.05.010.
8. Bukur M, Castelo Branco B, Inaba K, Cestero R, Kobayashi L, Tang A, et al. The impact of American College of Surgeons trauma center designation and outcomes after early thoracotomy: a National Trauma Databank analysis. Am Surg 2012;78:36-41.
9. Annual call for data: National Trauma Data Bank (NTDB) [Internet]. American Col- lege of Surgeons Available from https://www.facs.org/quality-programs/trauma/ tqp/center-programs/ntdb.
10. National Trauma Data Bank: NTDB research data set user manual and variable de- scription list [Internet]. American College of Surgeons. Available from: https:// www.facs.org/~/media/files/quality%20programs/trauma/ntdb/ntdb%20rds%20user

    %20manual%20all%20years.ashx

    National Trauma Data Standard. Data dictionary [Internet]. American College of Sur- geons Available from: https://www.facs.org/~/media/files/quality%20programs/ trauma/ntdb/ntds/data%20dictionaries/ntds%20data%20dictionary%202015.ashx.

11. Hicks CW, Hashmi ZG, Velopulos C. Association between race and age in survival after trauma. JAMA Surg 2014;149:642. https://doi.org/10.1001/jamasurg.2014.166.
12. National Trauma Data Bank. Annual report [Internet]. American College of Surgeons; 2016 Available from: https://www.facs.org/~/media/files/quality%20programs/ trauma/ntdb/ntdb%20annual%20report%202016.ashx.
13. Kaji AH, Bosson N, Gausche-Hill M. Patient outcomes at urban and suburban level I versus level II trauma centers. Ann Emerg Med 2017;70:161-8. https://doi.org/10. 1016/j.annemergmed.2017.01.040.
14. Liberman M, Mulder DS, Jurkovich GJ, Sampalis JS. The association between trauma system and trauma center components and outcome in a mature regionalized trauma system. Surgery 2005;137:647-58. https://doi.org/10.1016/j.surg.2005.03. 011.
15. Cudnik MT, Newgard CD, Sayre MR, Steinberg SM. Level I versus level II trauma cen- ters: an outcomes-based assessment. J Trauma 2009;66:1321-6. https://doi.org/10. 1097/TA.0b013e3181929e2b.
16. Trauma levels explained [internet]. American Trauma Society. Available from: https://www.amtrauma.org/page/traumalevels
17. Demarest GB, Scannell G, Sanchez K, Dziwulski A, Qualls C, Schermer CR, et al. In- house versus on-call attending trauma surgeons at comparable level I trauma cen- ters: a prospective study. J Trauma 1999;46:535-40.
18. Ashley DW, Mullins RF, Dente CJ, Garlow L, Medeiros RS, Atkins EV, et al. What are the costs of trauma center readiness? Defining and standardizing readiness costs for trauma centers statewide. Am Surg 2017;83:979-90.
19. Sasser SM, Hunt RC, Faul M, et al. Guidelines for field triage of injured patients: rec- ommendations of the National Expert Panel on Field Triage. MMWR; 2011.
20. ANZCOR guideline 11.10.1 management of cardiac arrest due to trauma [Internet]. Australian Resuscitation Council Available from: https://resus.org.au/download/ section_11/anzcor-guideline-11-10-1-als-traumatic-arrest-27apr16.pdf.