Management of esophageal injuries secondary to trauma

G Model JINJ 7277 No. of Pages 8

Injury, Int. J. Care Injured xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Injury journal homepage: www.elsevier.com/locate/injury

Review

Management of esophageal injuries secondary to trauma Patrizio Petrone, MD MPH MSHSA FACSa,b,* , Kawthar Kassimi, MD1a , Marta Jiménez-Gómez, MD1a , Alejandro Betancourt, MD FACS3c , Alexander Axelrad, MD FACS FCCM1a , Corrado P. Marini, MD FACS1a a b c

Division of Trauma Surgery, Surgical Critical Care and Acute Care Surgery, Department of Surgery, NYU Winthrop Hospital, Mineola, New York, United States PhD Candidate, Universidad de Las Palmas de Gran Canaria, Spain Division of Trauma Surgery and Acute Care Surgery, Department of Surgery, Southside Hospital, Northwell Health, Bay Shore, New York, United States

A R T I C L E I N F O

A B S T R A C T

Keywords: Esophageal injuries Management Treatment Morbidity Mortality

Traumatic esophageal injuries occur less than 10% of the time in the setting of blunt or penetrating trauma. The purpose of this literature review is to provide an update on the most recent changes involving the diagnosis and treatment of esophageal injuries. A literature search was conducted using PubMed, to identify articles written in English language with the terms “non- iatrogenic”, “esophageal”, “trauma”, “diagnosis”, “management”, and “prognosis”. Case reports and articles involving nontraumatic esophageal perforations were excluded. Fifty pertinent articles in English language from 1947 to 2015 were selected for review. Based on the review of all articles, we designed a diagnostic and therapeutic algorithm to facilitate the diagnosis and management of the traumatic esophageal injury. © 2017 Elsevier Ltd. All rights reserved.

Contents Introduction . . . . . . . . . . . . . . . . . Anatomy . . . . . . . . . . . . . . . . . . . . Incidence . . . . . . . . . . . . . . . . . . . Classification of injuries . . . . . . . . Mechanism of injuries . . . . . . . . . Associated injuries . . . . . . . . . . . . Diagnosis . . . . . . . . . . . . . . . . . . . Physical exam . . . . . . . . . . . . . . . . Diagnostic studies . . . . . . . . . . . . Management . . . . . . . . . . . . . . . . Non-operative management . . . . Surgical management . . . . . . . . . Cervical esophageal injury . . Thoracic esophageal injury . . Abdominal esophageal injury Morbidity . . . . . . . . . . . . . . . . . . . Mortality . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

* Corresponding author at: Director of Surgical Research, Program Director, International Visiting Scholars & Research Fellowship, Division of Trauma Surgery, Surgical Critical Care and Acute Care Surgery, Department of Surgery - NYU Winthrop Hospital, Clinical Associate Professor of Surgery, New York Medical College, 222 Station Plaza North, Suite 603, Mineola, New York 11501, USA. E-mail addresses: [email protected], [email protected] (P. Petrone). http://dx.doi.org/10.1016/j.injury.2017.06.012 0020-1383/© 2017 Elsevier Ltd. All rights reserved.

Please cite this article in press as: P. Petrone, et al., Management of esophageal injuries secondary to trauma, Injury (2017), http://dx.doi.org/ 10.1016/j.injury.2017.06.012

00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

G Model JINJ 7277 No. of Pages 8

2

P. Petrone et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx

Conflict of interest Ethical approval . . Informed consent . References . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

Introduction Traumatic esophageal injury is a very rare entity that requires an individualized approach to its management [1]. The first report of an esophageal injury dates back to the surgical papyrus of Edwin Smith, approximately 2500 years BC. It described the injury as “a gaping wound of throat, penetrating the gullet” [2]. In 1941, Frink [3] reported the first case of penetrating esophageal injury treated successfully by surgical drainage. Subsequently, in 1947, Barrett [4] and Olsen and Clagett [5] published the first series of esophageal injuries treated with surgical repair. Due to the low incidence and the constant presence of associated injuries that may mask the diagnosis, treatment of esophageal injuries is often delayed, therefore, contributing to the increased morbidity and mortality associated with esophageal injuries [6]. Although articles have been published on this topic as a multicenter study of the American Association for the Surgery of Trauma (AAST) [7], as an analysis of the National Trauma Data Bank [8], or even as guidelines of the Western Trauma Association (WTA) [9]. This article aims to provide a more comprehensive literature review for an extended period of time that encompasses almost seven decades. It includes the most recent advances involving the diagnosis and treatment of traumatic perforations of the esophagus. Anatomy The anatomic peculiarity of the esophagus is that it crosses three anatomical regions: the neck, chest, and abdomen. Unlike the rest of the digestive tract, the esophagus has a segmental blood

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

00 00 00 00

supply without mesentery and serosal layer; therefore, its rupture is associated with rapid contamination of the neck, mediastinum, and/or abdomen depending on the location of the injury. As opposed to the aorta, carina, and/or main bronchi, which are at risk of injury following a horizontal deceleration mechanism of injury, the esophagus is rarely injured following a deceleration mechanism of injury [1]. Its size, the anatomical location protected within the chest wall, and the associated injuries that mask its diagnosis, account for the reported low incidence of traumatic esophageal injuries [10,11]. The difficulty of access, the poor network of collateral vessels, and the proximity to vital structures, all contribute to the high morbidity and mortality associated particularly to injuries of the thoracic esophagus [11,12]. The associated negative intrathoracic pressure favors rapid mediastinal and pleural contamination after esophageal rupture and, if not identified early and treated aggressively, causes a systemic inflammatory response, multi-organ failure, and frequently, death [11,12]. Incidence Approximately 60% of esophageal injuries are iatrogenic, typically the result of either diagnostic or therapeutic endoscopic procedures [13–15]. The real incidence of esophageal injuries by external trauma is difficult to estimate given the paucity of available studies [14], although some authors report an incidence of less than 10% [13,16–18]. Due to the low frequency of these injuries, guidelines regarding its management are based mostly on retrospective studies with limited sample size. Among the reviewed series of traumatic esophageal injuries, the highest

Table 1 Incidence by sex and average age in esophageal trauma. Author and year (n)

Duration of the study (years)

Men n (%)

Women n (%)

Average age (range)

Propovsky et al., 1974 [39] (11) Symbas et al., 1980 [27] (48) Propovsky, 1984 [40] (15) Weiman et al., 1995 [20] (19) Flowers et al., 1996 [35] (31) Asensio et al., 1997 [41] (43) Srinavasan et al., 2000 [23] (55) Hanpeter et al., 2000 [22] (24) Asensio et al., 2001 [7] (405) Fadoo et al., 2003 [21] (11) Vogel et al., 2005 [15] (47) Bell et al., 2007 [42] (120) Linden et al., 2007 [17] (43) Eroglu et al., 2009 [18] (44) Arantes et al., 2009 [43] (163) Ahmed et al., 2009 [38] (33) Onat et al., 2010 [14](12) Patel et al., 2013 [8] (227) Makhani et al., 2013 [12](327) TOTAL (1696/1352)

5

N/A

N/A

N/A

15

N/A

N/A

N/A (16–67)

9

N/A

N/A

N/A

5

15 (79)

4 (21)

22 (17–55)

3 6 8

28 (903) 36 (84) 52 (94)

3 (9.7) 7 (16) 3(6)

24 (16–54) N/A 29 (17–68)

14

21 (87.5)

3 (12.5)

26 (13–43)

10,5 N/A 12 5 14 20 7 4 28 2 6

355 (88) 8 (72.7) 31 (66) 89 (74.2) N/A 28 (63.6) 153 (93.9) 30 (91) 25 (83.3) N/A 267 (81.7) 1138 (84.2)

50 (12) 3 (27.3) 16 (34) 31 (25.8) N/A 16 (36.4) 10 (6.1) 3 (9) 5 (16.7) N/A 60 (18.3) 214 (15.8)

29 (2–87) 40 (16–84) 59 (18–90) 34 (4–92) 63 (32–94) 53 (5–81) 27,5 (7–78) 31 (22–40) 27,5 (2–71) 28 (21–38) 30 (9–50) 33 (2–94)

N/A: Non Available.

Please cite this article in press as: P. Petrone, et al., Management of esophageal injuries secondary to trauma, Injury (2017), http://dx.doi.org/ 10.1016/j.injury.2017.06.012

G Model JINJ 7277 No. of Pages 8

P. Petrone et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx

incidence of 13.6% was reported by Eroglu et al. [18] in a single center study that included traumatic esophageal injuries over a period of 20 years. Asensio et al. [7] published the largest retrospective multicenter study (34 centers) which included 405 patients with penetrating esophageal injuries during 10.5 years to assess the impact of time to diagnosis and treatment on the morbidity and mortality of patients with penetrating esophageal injuries. A more recent study [8] with data from the National Trauma Data Bank (NTDB) comprising 227 patients with penetrating esophageal injuries aimed to identify specific risk factors contributing to esophageal-related complications and mortality. While this review did not report the incidence of esophageal injuries, it showed that most injuries occurred in men [7,8,15] with an average age of 30 years and that most deaths occurred during the first 24 h from injury due to severe associated injuries (Table 1). Of note, this review failed to show that time to repair affects the outcome of patients with penetrating esophageal injuries as opposed to the multicenter study of Asensio et al. [7] that suggests that delayed operative repair increases morbidity and mortality.

3

compresses the esophagus against the thoracic spine [1]. Blunt esophageal injury can also result from a Heimlich maneuver, cardiopulmonary resuscitation [11], and barotrauma [12]. As opposed to penetrating injuries, blunt esophageal injuries are not always diagnosed early, therefore, the clinical outcome of patients is affected by delay in diagnosis. Penetrating, as well as blunt esophageal injuries associated with major vascular injuries have a very high mortality from exsanguination (Table 3) [8,10]. Associated injuries Isolated esophageal injuries are the exception more than the rule [1,20]. Associated injuries increase the complication rate and worsen the prognosis [21]. In the neck, severe associated injuries include injuries to the trachea and vascular structures. In the chest, associated injuries involve the lungs, heart and/or great vessels [22,23]. Abdominal injuries are less frequent, due to the short segment of esophagus in this location [7]. Abdominal esophageal injuries can be diagnosed promptly when there are associated injuries to the spleen, stomach, or pancreas; however, they may be missed if only the esophagus is injured.

Classification of injuries Diagnosis Esophageal injuries are classified according to the mechanism of injury in penetrating (firearms or stab wound) and blunt trauma (motor vehicle collisions, barotrauma), and according to its anatomic location in cervical (57%), thoracic (26%), which are more complex, and abdominal esophageal injury (17%). The severity of injury is graded using the Organ Injury Scale of the American Association of Trauma Surgery (AAST-OIS) [19] (Table 2).

Early diagnosis of traumatic esophageal injury requires a high level of suspicion [7] based on the mechanism of injury and suspected anatomic location of a potential injury. The evolution of these injuries is usually rapid and atypical, and a delay in diagnosis increases complications [7,11,23–25]. Physical exam

Mechanism of injuries The most frequent mechanism of injury is penetrating, with esophageal injuries from firearms, usually located in the neck [7,8,15,16]. The extent of the esophageal disruption after gunshot wounds depends on the type of weapon used and on the muzzle velocity of the bullet, namely, low (<1500 fps) versus high velocity gunshot wounds (>1500 fps). Low velocity bullets penetrate the tissues and the esophagus causing limited injury to the site of penetration without the massive disruption from the blast effect and cavitation associated with high velocity projectiles. Of note, the incidence of thoracic esophageal injury is only 1% even following trans-mediastinal gunshot wounds. Esophageal injury from blunt trauma is much less frequent. Usually, the mechanism of blunt injury involves the transmission of energy by a non-contiguous force that damages the esophagus indirectly [12,15,20]. Esophageal injuries from blunt trauma are more frequent in the neck and thorax, mostly from high speed motor vehicle crashes with horizontal deceleration that compresses the esophagus against the vertebral bodies. A typical mechanism of blunt injury involves perforation of the esophagus from the impact of the steering wheel on the sternum that

There are no specific symptoms or pathognomonic signs suggestive of esophageal injury [1,11]. Asensio et al. [7] did not find any relevant symptoms or signs associated with esophageal injury; in fact, only 7% presented with dysphagia and 19% with subcutaneous emphysema. The mechanism of injury was the most important factor leading to early diagnosis. In a series of 115 consecutive esophageal perforations, 65 iatrogenic, 28 traumatic, and 22 spontaneous, reported by Nesbitt and Sawyer [25], the most common symptoms included pain (71%), fever (51%), dyspnea (24%), and crepitus (22%). The classic presentation of esophageal perforation is that of acute dysphagia and inability to swallow saliva [1,26]. Pain, dysphagia, and subcutaneous emphysema are more commonly seen in patients with perforation of the cervical esophagus, whereas fever is a late sign, more frequently observed in patients with mediastinal contamination from perforation of the thoracic esophagus. Perforations of the abdominal esophagus are usually not contained, therefore, they present with an acute abdomen or when associated with splenic or vascular injuries with overt hemorrhagic shock [14]. Diagnostic studies

Table 2 Esophageal injury scale of the American Association for the Surgery of Trauma [19]. Grade I II III IV V 1

1

Description of Injury Contusion/Hematoma Partial thickness laceration Laceration < 50% circumference Laceration > 50% circumference Segmental loss or devascularization < 2 cm Segmental loss or devascularization > 2 cm

Advance one grade for multiple lesions up to grade III.

Laboratory studies are not useful for diagnosis during the early phase of esophageal injuries. They usually show non-specific alterations due to an inflammatory response such as leukocytosis, lactic acidosis, increased C-reactive protein, and increased procalcitonin [12]. If there is a high level of suspicion, cervical and thoracic plain radiographs can be performed immediately for the diagnosis [11,16], but these investigations are not specific [1,26]. Cervical subcutaneous emphysema may occur in 60% to 90% of patients; [9,27] in the thorax, pneumomediastinum, pneumothorax and/or pleural effusion [1] are detected, most commonly in the left hemithorax, in 80% of patients with thoracic esophageal

Please cite this article in press as: P. Petrone, et al., Management of esophageal injuries secondary to trauma, Injury (2017), http://dx.doi.org/ 10.1016/j.injury.2017.06.012

G Model JINJ 7277 No. of Pages 8

4

P. Petrone et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx

Table 3 Mechanism of injury and location. Author (year)

N patients

Mechanism of injury

Anatomic location

Penetrating n (%)

Propovsky et al., 1974 [39] Defore et al., 1977 [36] Symbas et al., 1980 [27] Propovsky, 1984 [40] Glatterer et al., 1985 [44] Nesbitt and Sawyers, 1987 [25] Hatzitheofilou et al., 1993 [45] Weiman et al., 1995 [20] Flowers et al., 1996 [35] Srinavasan et al., 2000 [23] Hanpeter et al., 2000 [22] Asensio et al., 2001 [7] Fadoo et al., 2003 [21] Vogel et al., 2005 [15] Bell et al., 2007 [42] Linden et al., 2007 [17] Eroglu et al., 2009 [18] Arantes et al., 2009 [43] Ahmed et al., 2009 [38] Onat et al., 2010 [14] Patel et al., 2013 [8] Makhani et al., 2013 [12] TOTAL

Blunt n (%)

Cervical n (%)

Thoracic n (%)

Abdominal n (%)

Others n (%)

Total

Firearm

Stab

Others

11

11 (100)

11 (100)

–

–

–

6 (54)

5 (45)

–

–

77

77 (100)

58 (75)

–

19 (15)

–

45

21

11

–

48

48 (100)

48 (100)

–

–

–

24 (50)

7 (14,6)

–

15

15 (100)

15 (100)

–

–

–

8 (53,3)

17 (35,4) 7(46,7)

–

–

262

17 (81)

–

–

–

4 (19)

21 (80,7)

5 (19,3)

–

–

115

N/A

N/A

N/A

N/A

N/A

27 (23,5)

69 (60)

19 (16,5)

–

51

50 (98)

17 (33,3)

33 (64,7)

–

1 (2)

N/A

N/A

N/A

N/A

19

18 (95)

18 (100)

–

–

1 (5)

12 (63,2)

2 (10,5)

5 (26,3)

–

31 55 24

241 (77,4) 50 (91) 24 (100)

– 46 (92) 24 (100)

– 4 (8) –

– – –

7 (22,6) 5 (9) –

5 (16,1) 20 (36) –

15 (48,4) 25 (46) 24 (100)

4 (12,9) 5 (9) –

– 5 (9) –

4053

405 (100)

307 (75,7)

75 (18,5)

24 (5,8)

–

229 (53)

121 (27,9)

69 (16)

14 (3,1)

11

N/A

N/A

N/A

N/A

N/A

N/A

47

N/A

N/A

10 (21,3)

37 (78,7)

–

–

120

120(100)

–

120 (100)

–

–

–

43 (1 trauma) 44 (6 trauma) 163

0

1 (2)

–

43 (100)

–

–

N/A

N/A

14 (31,8)

18 (40,9)

12 (27,3)

–

158 (97)

131 (80)

27 (17)

–

5 (3)

73 (45)

49 (30)

4 (2)

37 (23)

33

33 (100)

27 (81,8)

6 (18,2)

–

–

N/A

N/A

N/A

N/A

30 (21 trauma)

20 (95)

16 (76)

4 (19)

1 (5)

–

30 (100)

–

–

–

227

227 (100)

–

N/A

N/A

N/A

N/A

327

202 (61,8)

125 (38,2)

190 (58,1)

123 (37,6)

–

14 (4,3) T+C

4

31 (25)

156 (47,7)

63 (53)

26 (22)

46 (14,1)

–

1922

N/A: Non Available 1 The items are not measured together, only penetrating trauma was classified by location. 2 Only cervical lesions were classified by mechanism of injury. 3 The classification of the anatomical location is performed based on the lesions, not on the patients (n = 433 lesions). 4 Meaning by others: 8 patients (1.8%) abdominal and thoracic lesions; 5 (1.1%) cervical and thoracic lesions; 1 (0.2%) cervical and abdominal lesions.

injuries [16,20]. The presence of subdiaphragmatic air is a sign of possible abdominal esophageal perforation. The chemical analysis of pleural fluid obtained from either thoracentesis or from tube thoracostomy, while performed rarely, may aid in the diagnosis if food particles, a pH less than 6 or elevated salivary amylase are present in the fluid that is sampled [7,13]. The utilization of contrast material-enhanced CT scan to diagnose esophageal injuries after penetrating and blunt trauma in stable patients has been steadily increasing over the past years, due to its high sensitivity in addition to its ability to identify associated injuries.[9,12,21,28–32] In particular, CT esophagography has been shown to have a sensitivity and specificity of 95% and 91% respectively, for the diagnosis of penetrating upper digestive tract injuries [32]. Contrast material-enhanced CT scan is also useful in monitoring patients who do not improve after treatment, since it has a higher sensitivity to show paraesophageal manifestations such as collections, abscesses and spillages, it also

serves as a guide to the placement of percutaneous drains [11,15,24,28]. Data from Hanpeter [22] show that contrast material-enhanced CT scan is reliable in assessing the trajectory of the bullet in trans-mediastinal wounds. Furthermore, Wu et al. [24] showed that fluoroscopic esophagography is unnecessary in patients with pneumomediastinum on CT who do not have CT findings of esophageal perforation. The new recommendations of the Western Trauma Association (WTA) [9] advocate performing CT scan in penetrating injuries in hemodynamically stable patients with nonspecific signs or symptoms in cervical zones I and III. They also recommend CT scan in patients with suspected thoracic esophageal injuries, provided there are no obvious vascular or aerodigestive injuries. Contrast esophagogram has been the test of choice for years with a sensitivity of 100% and specificity of 96% [7,15,25]. It provides information about the injury, its location and the extent of the injury itself. It is considered positive if it shows contrast

Please cite this article in press as: P. Petrone, et al., Management of esophageal injuries secondary to trauma, Injury (2017), http://dx.doi.org/ 10.1016/j.injury.2017.06.012

G Model JINJ 7277 No. of Pages 8

P. Petrone et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx

extravasation, and this occurs in 60% to 80% of cervical injuries [1], and in 75% to 80% of thoracic perforations [13]. The standard technique for contrast esophagography is to first administer watersoluble contrast, namely, gastrografin. Gastrografin is easily absorbed by the mediastinum, therefore, its extravasation is unlikely to cause mediastinal fibrosis [25]. However, due to the rapidity of its transit time through the esophagus and its solubility, a negative gastrografin study requires a confirmatory study using thin barium, and it should be administered in a lateral decubitus position. Approximately 22% of missed injuries by aqueous contrast medium will be identified by esophagography with thin barium [33]. The performance of the esophagogram requires a stable and cooperative patient, conditions that are rarely encountered in trauma victims, where the patient is usually agitated, unconscious, and intubated [12,24,34]. Flexible endoscopy has many advantages over rigid esophagoscopy including its availability in most trauma centers [23], the ease of portability and the fact that it can be used in intubated patients in the emergency room or in the operating room. It has a sensitivity of 100% and a specificity of 83% [10,33]. Many authors advocate its use because it provides a direct visualization of the injury site [8,11,16,22,34,35]. It has been shown to be a useful confirmatory tool in patients with undetermined CT findings in the thoracic and cervical zones I and III [30,36,37]. As shown by Ahmed [38]flexible endoscopy is superior to esophagography for the diagnosis of injuries to the hypopharynx and cervical esophagus. Furthermore, Srinavasan [23] has documented that esophagoscopy may be used safely in unstable and intubated patients. He has also documented that in 69% of the patients endoscopy alters the surgical management. Those who do not favor the use of endoscopy as a confirmatory study argue that air insufflation during the procedure may increase the size of the perforation, promote mediastinal contamination, or even extend a contained leak [1]. Management The management of the esophageal injuries begins with the principles of the Advanced Trauma Life Support (ATLS1) [46] protocol of the American College of Surgeons, focusing on the patient’s life-threatening injuries. Thus, the first step is the securement of the airway followed by an aggressive resuscitation with intravenous isotonic fluids. The patient must remain NPO with a nasogastric tube placed ideally by endoscopy to prevent the extravasation of the gastric material within the paraesophageal space. In addition, it is mandatory to start immediately broad spectrum antibiotic coverage against anaerobic and aerobic microorganisms. For suspected esophageal injury, once the patient is stabilized, it is essential to define the location and the extent of the injury, as well as the possible associated injuries since they affect the management of the patient and the overall prognosis. Patients who are deemed to require surgical treatment should undergo surgery expeditiously since the interval elapsed from the time of the injury to the diagnosis and treatment of the injury may affect the postoperative morbidity. A multicenter study by Asensio et al. [7] concluded that delay in the diagnosis and treatment of the traumatic esophageal injury increases the risk of esophageal complications by 41%. Non-operative management Patients with esophageal injuries may be offered a nonoperative treatment if they meet the criteria first described by Mengoli et al. [47] in 1965 and then modified in 1997 by Altorjay et al. [48], namely:

5

- I. Intramural perforation. - II. Extramural perforation, in case of: 1. Perforation detected early, or when detected late, a circumscribed one 2. The lesion in the mediastinum or in the space between the mediastinum and visceral pleura creates a well encapsulated extravasation. 3. From the cavity surrounding the esophagus the contrast medium flows adequately back into the esophageal lumen. 4. The tissue defect is not in neoplastic tissue, not in the abdominal cavity, and is not accompanied by simultaneous obstructive esophageal disease 5. Symptoms are minimal 6. Symptoms and signs of septicemia are absent 7. Availability of adequate diagnostic techniques: swallowing radiological examination, eventually CT. 8. Adequate thoracic surgical experience and skills Patients with injuries from stab wounds and low caliber gunshot wounds limited to the hypopharynx just above arytenoid cartilages can be managed safely non-operatively since the low intraluminal pressure and the surrounding muscles facilitate healing. The non-operative management includes keeping the patient in nil per os (NPO) status, placement of a NG tube endoscopically, initiation of enteral feeding via a feeding tube, if feasible, or nutritional support with TPN, administration of broad spectrum antibiotic coverage, and drainage of potential collections. Additional measures may include:  Radiological percutaneous drainage.  Thoracic drainage by chest tube.  Video-thoracoscopic drainage since it provides a better view and a superior access to irrigation and removal of infected material [1], while at the same time it allows pulmonary decortication and an effective placement of the chest tube. The first 24 h after the injury are the most important from the standpoint of the possible clinical evolution of the patient’s condition. Therefore, the patient requires intense monitoring, ideally in an intensive care unit setting, to identify early signs of clinical deterioration requiring surgical intervention [48]. In a retrospective study published by Vogel et al. [15], 32 of 47 esophageal perforations were managed successfully nonoperatively showing that aggressive non-operative management provides adequate control of the sepsis secondary to the esophageal leakage. Decreased mortality and morbidity while avoiding unnecessary surgical interventions in the majority of the patients were observed. Surgical management The decision to proceed with an operative approach to the management of esophageal perforations depends on the patient's condition, the location, size, and the extent of the injuries. Patients who should be treated surgically include those with hemodynamic instability or sepsis, patients with obvious not-contained extravasation of contrast material and patients with contained perforations but with systemic signs of sepsis [1]. One of the remaining controversial issues surrounding the surgical treatment of patients with thoracic esophageal injuries is whether delayed treatment (>24 h from injury) affects postoperative morbidity and mortality. In the past, most surgeons embraced the dogma that delays to operative repair of thoracic esophageal perforations including the delays resulting from diagnostic investigations resulted in excessive morbidity [7]. However, more recently, the literature suggests

Please cite this article in press as: P. Petrone, et al., Management of esophageal injuries secondary to trauma, Injury (2017), http://dx.doi.org/ 10.1016/j.injury.2017.06.012

G Model JINJ 7277 No. of Pages 8

6

P. Petrone et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx

that while the rate of primary repair of thoracic esophageal perforations is lower, delayed diagnosis and treatment is not associated with worse outcome [32]. Cervical esophageal injury While cervical esophageal trauma can be generally managed non-operatively [48],due to the low operative morbidity of cervical exploration it is hard to justify the development of any complication from failure to proceed with exploration of penetrating injuries to zone II of the neck. Every patient with a cervical esophageal injury must be kept NPO, have a NG tube placed and must be treated with the intravenous administration of broad spectrum antibiotics against both aerobic and anaerobic bacteria. Clindamycin provides excellent coverage due to the typical flora of the mouth [1]. Due to the increased rate of colonization with Candida in patients with protracted use of proton pump inhibitors, the administration of fluconazole should be added to clindamycin in this subset of patients. The cervical esophagus is approached through an incision along the medial border of the left sternocleidomastoid muscle, or by a collar incision if bilateral cervical exploration is required. Intraoperative endoscopy can help to identify the site of the perforation in patients with large hematoma. The administration of methylene blue by the anesthesiologist can be of help in the case of small perforation/s from pellets from shotgun wounds. The authors favor the use of a 46 French Maloney bougie for the identification and the repair of perforations in this location and to prevent narrowing from the repair. The esophagus must be mobilized circumferentially avoiding excessive dissection to identify the injury and to facilitate repair. The surgical management includes appropriate debridement of the site of the perforation to healthy tissue followed by a single or double layer tension-free closure of the defect with either absorbable or nonabsorbable sutures and adequate drainage. To prevent a possible tracheoesophageal fistula, or disruption or blowout of carotid artery repair, it is recommended to buttress the repair with vascularized tissue, either with the strap muscle or alternatively with the sternocleidomastoid muscle [44]. While the strap muscle must be divided distally due its cephalad origin of its blood supply, the sternocleidomastoid muscle due to its tripartite blood supply can be divided either distally or proximally and then rotated to buttress the repair. The repair once completed can be tested with an air insufflation leak test using an endoscope. In the rare cases when a perforation cannot be identified intraoperatively, the ipsilateral site should be drained appropriately with an external drain. We prefer a closed drainage system such as a Jackson-Pratt 7 mm drain. Drainage alone in these cases has been shown to be adequate treatment [11]. A feeding tube placed at the time of the repair will allow the immediate initiation of adequate enteral nutritional support while healing occurs. In the case of esophageal injuries involving more than 50% of the circumference of the esophagus or in cases of massive destruction preventing primary repair, the surgeon is left with only two options: the performance of a side esophagostomy or an end cervical esophagostomy. To perform the side esophagostomy, the esophagus must be mobilized extensively avoiding injuries to the recurrent laryngeal nerves in order to place a plastic rod underneath it. The distal lumen can be left either open or closed with an absorbable suture. Nutritional support is provided via a feeding tube placed coaxially distal to the pylorus through a 28– 32 F decompressing gastrostomy tube done through a limited laparotomy. Avoiding an end cervical esophagostomy is preferable in order to avoid at a later time a complex reconstruction. A side esophagostomy may allow a simple transverse closure following resolution of the edema.

Thoracic esophageal injury Injuries to the thoracic esophagus are extremely rare even in the setting of trans-mediastinal gunshot wounds. Patients with very small contained perforations can be managed non-operatively and if necessary with placement of an esophageal stent and appropriate external drainage of the involved pleural cavity. Of note, while stents have been shown to be useful for the treatment of non-traumatic (due to neoplasms or after endoscopic procedures) esophageal perforations, there are no data comparing stents to surgical repair in the setting of esophageal perforations caused by external trauma. Occasionally, based on the local expertise available, small perforations can be approached thoracoscopically with debridement, repair and external drainage coupled with enteral nutrition via a feeding tube placed coaxially distal to the pylorus through a 28–32 F decompressing gastrostomy tube done through a limited laparotomy. Esophageal perforations diagnosed by obvious massive contrast extravasation should be ideally treated operatively with thoracotomy done as soon as feasible [1]. The negative intrathoracic pressure draws saliva (1013 bacteria/ ml), gastric juice and bacteria from the perforated esophagus into the mediastinum and pleural space causing marked fluid shift and shock. Perforation of the middle third of the esophagus are approached from the right side via thoracotomy through the fourth or through the sixth intercostal space depending on the location of the perforation, whereas injuries to the distal third of the esophagus require a left thoracotomy through the seventh intercostal space. The cornerstone of treatment is an immediate evacuation of the contaminated and necrotic material, debridement of the perforation to healthy tissue and a tension-free primary repair over a 46F Maloney bougie buttressed by an intercostal muscle flap [1,43].As mentioned for the treatment of cervical esophageal injuries, esophagoscopy and/or methylene blue may aid in detecting the site/s of the perforation. Enteral nutritional support as previously described is an essential component of the treatment plan. If primary repair is not possible because of hemodynamic instability, extreme contamination due to delayed diagnosis, and/ or the extent of the injury that prevents a safe primary repair, the surgeon is left with the option of repairing the esophagus over a large size T-tube in order to create a controlled esophagocutaneous fistula and in cases of extreme necrosis to proceed with a complete diversion that involves the creation of a sidecervical esophagostomy through a left cervical incision (spit fistula), a gastrostomy, a feeding jejunostomy, and debridement and drainage of the distal esophagus. In the latter case, if the patient survives, he/she can undergo either a gastric pull-up or a colonic interposition 6–12 months later to reconstruct the gastrointestinal tract. Abdominal esophageal injury Traumatic injury to the abdominal esophagus is less frequent due mostly to the shorter length of this segment and to the protection afforded by the surrounding structures. The surgical approach includes a standard midline trauma laparotomy, identification of the perforation, appropriate debridement to health tissue and primary repair over a 52F Maloney bougie buttressed with a gastroplasty using the fundus of the stomach after extensive abdominal washout with warm saline. Testing of the integrity of the repair with air insufflation through the NG tube while the stomach and abdominal esophagus are submerged under saline should be done at completion of the repair. External drainage with a 10F Jackson-Pratt drain of the left subphrenic space is part of the surgical plan. A feeding jejunostomy should be added

Please cite this article in press as: P. Petrone, et al., Management of esophageal injuries secondary to trauma, Injury (2017), http://dx.doi.org/ 10.1016/j.injury.2017.06.012

G Model JINJ 7277 No. of Pages 8

P. Petrone et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx

in those cases where the surgeon has concerns regarding the possibility of a leak due to the conditions of the esophagus at the time of the repair. Postoperative assessment of healing with esophagography done on postoperative day 5–7 is helpful following repair of all esophageal injuries. The modality of surgical treatment performed more frequently, according to the latest analysis made by Patel et al. [8], which included patients from the NTDB suffering esophageal perforation who survived for more than 24 h, was primary repair in 62% of the cases, followed by drainage in 13%, resection in 4%, and diversion only in 1% of the cases with 20% of procedures used to deal with the perforation not specified. Morbidity Makhani et al. [12] in a series of 327 cases, reported a morbidity of 46% with the highest morbidity rate in patients with thoracic esophageal injuries. Age and the abbreviate injury score (AIS) of the abdomen and pelvis greater than three are the most predictive incremental risk factors for increased morbidity [8]. Mortality The reported mortality rate of patients with traumatic esophageal perforations ranges from 29% to 44% [8,12]. Makhani et al. [12] reported a mortality of 20.5% in patients with thoracic esophageal injuries compared to a much lower mortality of patients with cervical esophageal injuries, whereas Patel et al. [8] reported a mortality rate of 44%, with 92% of deaths occurring in the first 24 h after the trauma. However, the time elapsed from the time of injury to the time of the intervention was not significantly different between patients with complications and deaths, in contrast to the results reported by Asensio et al. [7] showing increased morbidity in patients who had a delay in treatment as a result of diagnostic investigations. The only factor predictive of mortality was the ISS [8]. Conclusion While the incidence of traumatic esophageal injury is low, the rate of complications and mortality remains high especially in patients with associated injuries. Non-operative treatment and close monitoring is the current protocol in large and busy urban Level 1 trauma centers in stable patients with contained perforation without signs of systemic sepsis. Patients with a systemic sepsis and/or hemorrhagic instability require urgent surgical intervention. Our suggested surgical management includes:  Thorough irrigation, drainage and debridement of contaminated material.  Primary repair, whenever possible. Resection and diversion are option for patients with extreme destruction of the esophagus.  Interposition of autologous vascularized flaps such as the sternocleidomastoid muscle in case of neck injury or of intercostal muscle [44], diaphragm, and pericardium in patients with perforations of the thoracic esophagus.

Conflict of interest Patrizio Petrone, Kawthar Kassimi, Marta Jiménez-Gómez, Alejandro Betancourt, Alexander Axelrad and Corrado P. Marini declare that they have no conflict of interest.

7

Ethical approval The study was designed as a collective literature review. There were no patients involved in this study. Informed consent Not applicable. References [1] Bryant AS, Cerfolio RJ. Esophageal trauma. Thorac Surg Clin 2007;17:63–72. [2] Brewer 3rd LA. History of surgery of the esophagus. Am J Surg 1980;139 (6):730–43. [3] Frink NW. Spontaneous rupture of the esophagus; report of a case with recovery. J Thorac Surg 1947;16(3):291–7. [4] Barrett NR. Report of a case of spontaneous perforation of the oesophagus successfully treated by operation. Br J Surg 1947;35(138):216–8. [5] Olsen AM, Clagett OT. Spontaneous rupture of the esophagus; report of a case with immediate diagnosis and successful surgical repair. Postgrad Med 1947;2 (6):417–21. [6] Chirica M, Champaulta A, Drayb X, Sulpicec L, Munoz-Bongranda N, Sarfati E, et al. Esophageal perforations. J Visc Surg 2010;147:117–28. [7] Asensio JA, Chahwan S, Forno W, MacKersie R, Wall M, Lake J, et al. Penetrating esophageal injuries: multicenter study of the American Association for the Surgery of Trauma. J Trauma 2001;50:289–96. [8] Patel MS, Malinoski DJ, Zhou L, Neal ML, Hoyt DB. Penetrating oesophageal injury: a contemporary analysis of the National Trauma Data Bank. Injury 2013;44:48–55. [9] Biff WL, Morre EE, Feliciano DV, Albrecht RA, Croce M, Karmy-Jones R, et al. Western Trauma Association critical decisions in trauma: diagnosis and management of esophageal injuries. J Trauma Acute Care Surg 2015;79 (6):1089–95. [10] Brinster CJ, Singhal SS, Lee L, Marshall MB, Kaiser L, Kucharczuk J. Evolving options in the management of esophageal perforation. Ann Thorac Surg 2004;77:1475–83. [11] Kumar B, Shenfine J. Oesophageal trauma. Surgery 2011;29(11):563–7. [12] Makhani M, Midani D, Goldberg A, Friedenberg FK. Pathogenesis and outcomes of traumatic injuries of the esophagus. Dis Esophagus 2013;27:630–6. [13] Scott B, Johnson MD. Esophageal trauma. Semin Thorac Cardiovasc Surg 2008;20:46–51. [14] Onat S, Ulku R, Cigdem KM, Avci A, Ozcelik C. Factors affecting the outcome of surgically treated non-iatraogenic traumatic cervical esophageal perforation: 28 years experience at a single center. J Cardiothoracic Surg 2010;5:46. [15] Vogel S, Rout R, Martin T, Abbitt P. Esophageal perforation in adults Aggressive, conservative treatment lowers morbidity and mortality. Ann Surg 2005;241:1016–23. [16] Gupta NM, Kaman L. Personal management of 57 consecutive patients with esophageal perforations. Am J Surg 2004;187(1):58–63. [17] Linden PA, Bueno R, Mentzer SJ, Zellos L, Lebenthal A, Colson YL, et al. Modified T-tube repair of delayed esophageal perforation results in a low mortality rate similar to that seen with acute perforations. Ann Thorac Surg 2007;83 (3):1129–33. [18] Eroglu A, Turkyilmaz A, Aydin Y, Yekeler E, Karaoglanoglu N. Current management of esophageal perforation: 20 years’ experience. Dis Esophagus 2009;22:374–80. [19] Moore EE, Jurkovich GJ, Knudson MM, Cogbill TH, Malangoni MA, Champion HR, et al. Organ injury scaling VI: Extrahepatic biliary, esophagus, stomach, vulva, vagina, uterus (nonpregnant), uterus (pregnant), fallopian tube, and ovary. J Trauma 1995;39(6):1069–70. [20] Weiman DS, William WA, Katheleen BM, James WP, Timothy CF. Noniatrogenic esophageal trauma. Ann Thorac Surg 1995;59:845–50. [21] Fadoo F, Ruiz D, Dawn S, Webb WR, Gotway MB. Helical CT esophagography for the evaluation of suspected esophageal perforation rupture. AJR 2004;18:1177–9. [22] Hanpeter D, Demetriades D, Asensio JA, Berne TV, Velmahos GV, Murray J. Helical computed tomographic scan in the evaluation of mediastinal gunshot wounds. J Trauma 2000;49:689–95. [23] Srinavasan R, Haywood T, Horwitz B, Buckman R, Fisher R, Krevsky B. Role of flexible endoscopy in the evaluation of possible esophageal trauma after penetrating injuries. Am J Gastroenterol 2000;95:1725–9. [24] Wu CH, Chen CM, Chen CC, Wong YC, Wang CJ, Lo WC, et al. Esophagography after pneumomediastinum without CT findings of esophageal perforation: is it necessary. AJR 2013;201:977–84. [25] Nesbitt JC, Sawyers JL. Surgical management of esophageal perforation. Am Surg 1987;53(4):183–91. [26] Han SY, McElvein RB, Aldrete JS, Tishler JM. Perforation of the esophagus: correlation of site and cause with plain film findings. AJR 1985;145:537–40. [27] Symbas PN, Hatcher Jr. CR, Vlasis SE. Esophageal gunshot injuries. Ann Surg 1980;191(6):703–7. [28] White CS, Templeton PA, Attar S. Esophageal perforation: CT findings. AJR 1993;160:767–70.

Please cite this article in press as: P. Petrone, et al., Management of esophageal injuries secondary to trauma, Injury (2017), http://dx.doi.org/ 10.1016/j.injury.2017.06.012

G Model JINJ 7277 No. of Pages 8

8

P. Petrone et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx

[29] Backer CL, Lo Cicero J, Hartz RS, Donaldson JS, Chields T. Computed tomography in patients with esophageal perforation. Chest 1990;98:1078–80. [30] Ibirogba S, Nicol AJ, Navsaria PH. Screening helical computed tomographic scanning in haemodynamic stable patients with transmediastinal gunshot wounds. Injury 2007;38:48–52. [31] Soreide JA, Viste A. Esophageal perforation: diagnostic work-up clinical decision-making in the first 24 hours. SJTREM 2011;19:66. [32] Ivatury RR, Moore FA, Biffl W, Leppaniemi A, Ansaloni L, Catena F, et al. Oesophageal injuries: position paper, WSES, 2013. World J Em Surg 2014;9 (1):9. [33] Conradie WJ, Gebremariam FA. Can computed tomography esophagography reliably diagnose traumatic penetrating upper digestive tract injuries? Clin Imaging 2015;39(6):1039–45. [34] Horwitz B, Krevsky B, Buckman Jr. RF, Fisher RS, Dabezies MA. Endoscopic evaluation of penetrating esophageal injuries. Am J Gastroenterol 1993;88:1249–53. [35] Flowers JL, Scott G, Ugarte M, Sartor W, Rodriguez A, Gens D, et al. Flexible endoscopy for the diagnosis of esophageal trauma. J Trauma 1996;40:261–6. [36] Defore Jr. WW, Mattox KL, Hansen HA, Garcia-Rinaldi R, Beall Jr AC, BeBakey ME. Surgical management of penetrating injuries of the esophagus. Am J Surg 1977;134:734–8. [37] Sperry JL, Moore EE, Coimbra R, Croce M, Davis JW, Karmy-Jones R, et al. Western Trauma Association critical decisions in trauma: penetrating neck trauma. J Trauma Acute Care Surg 2013;75(6):936–40. [38] Ahmed N, Massier C, Tassie J, Whalen J, Chung R. Diagnosis of penetrating injuries of the pharynx and esophagus in the severly injured patient. J Trauma 2009;67:152–4.

[39] Propovsky J, Lee YC, Berk JL. Gunshot wounds of the esophagus. J Thorac Cardiovasc Surg 1974;72(4):609–12. [40] Propovsky J. Perforations of the esophagus from gunshot wounds. J Trauma 1984;24(4):337–9. [41] Asensio JA, Berne J, Demetriades D, Murray J, Gomez H, Falabella A, et al. Penetrating esophageal injuries: time interval of safety for preoperative evaluation. How long is safe? J Trauma 1997;43:319–24. [42] Bell RB, Osborn T, Dierks EJ, Potter BE, Long WB. Management of penetrating neck injuries: a new paradigm for civilian trauma. J Oral Maxillofac Surg 2007;65:691–705. [43] Arantes V, Campolina C, Valerio SH, Nogueira de Sa R, Toledo C, Abreu Ferrari T, et al. Flexible esophagoscopy as a diagnostic tool for traumatic esophageal injuries. J Trauma 2009;66:1677–82. [44] Glatterer Jr. MS, Toon RS, Ellestad C, McFee AS, Rogers W, Mack JW, et al. Management of blunt and penetrating external esophageal trauma. J Trauma 1985;25(8):784–92. [45] Hatzitheofilou C, Strahlendorf C, Kakoyiannis S, Charalambides D, Demetriades D. Penetrating external injuries of the oesophagus and pharynx. Br J Surg 1993;80(9):1147–9. [46] American College of Surgeons. Advanced Trauma Life Support for Doctors, Student Course Manual. Chicago, IL: ACS (Ed); 2004. [47] Mengoli LR, Klassen KP. Conservative management of esophageal perforation. Arch Surg 1965;91(2):238–40. [48] Altorjay A, Kiss J, Vörös A, Bohák A. Nonoperative management of esophageal perforations. Is it justified? Ann Surg 1997;225(4):415–21.

Please cite this article in press as: P. Petrone, et al., Management of esophageal injuries secondary to trauma, Injury (2017), http://dx.doi.org/ 10.1016/j.injury.2017.06.012