Traumatic diaphragmatic hernia: 28-years analysis at a Brazilian university hospital

Background The objective of the study is evaluate the approach to patients with acute traumatic diaphragmatic hernia at a Brazilian university hospital during a 28-year period. Traumatic diaphragmatic hernia is an uncommon injury, however its real incidence may be higher than expected. Sometimes is missed in trauma patients, and is usually associated with signicant morbidity and mortality, this analysis may improve outcomes for the trauma patient care. Methods Retrospective study of time series using and analisys database records of trauma patients at HC- Unicamp was performed to investigate the incidence, trauma mechanism, diagnosis, herniated organs, associated injuries, trauma score, morbidity, and mortality of this injury. Results Fifty-ve cases were analyzed. Blunt trauma was two-fold frequent than penetrating trauma, are associated with high grade injury and motor vehicle collision was the most common mechanism. Left side hernia was four-fold frequent than right side. Diagnose was mostly performed by chest radiography (31 cases; 56%). Associated intra-abdominal injuries were found in 37 patients (67.3%) and extra-abdominal injuries in 35 cases (63.6%). The mortality was 20% (11 cases). Conclusions Computed tomography scan identied associated injuries and had high specicity. Isolated injuries were rare, and the presence of associated injuries increased morbidity and mortality. Despite advances in imaging methods, chest radiography in the trauma bay is useful as an initial approach. Despite the use of laparoscopy in a few cases, laparotomy was the most frequent approach for repair.


Background
A traumatic diaphragmatic hernia (TDH) is de ned as the protrusion of abdominal structures through an injured diaphragm into the thoracic cavity. Traumatic diaphragmatic injuries (TDI) without diaphragmatic hernias are present in 3% of all abdominal injuries. Data suggest that TDI are present in 1-7% of patients with blunt trauma (BT), and in 10-15% of those with penetrating wounds [1]. Gunshot wounds (GSW) are commonly (66%) involved in penetrating trauma, and motor vehicle collisions (MVC) (63%) in BT [2]. A diagnosis of TDI can be initially missed by computed tomography (CT-scan) in up to 60% of cases [3].
TDH is an uncommon injury; however, its real incidence may be slightly higher than expected [4]. Some collective reviews suggest that the incidence of TDH due to BT is as high as 75% because it involves more energy dissipation [5].
The clinical diagnosis of TDI is challenging for trauma and acute care surgeons, who depend on a high index of suspicion, mechanism of trauma, and interpretation of radiological images for the diagnosis.
Although uncommon, TDH are sometimes missed in trauma injuries and are associated with signi cant morbidity and mortality. It is essential to diagnose in the acute phase of trauma, avoiding missed injuries and delayed diagnosis. The increased number of trauma patient is a challenge for public health Page 3/16 authorities. We performed a retrospective analysis of the records of our trauma department patients to investigate the incidence, trauma mechanism, diagnosis, herniated organs, associated injuries, complications, morbidity, and mortality of this injury, evaluating the diagnosis, treatment and outcomes.

Methods
Descriptive retrospective analysis of the medical records between January 1990 and December 2017 of patients from the trauma database of the Division of Trauma Surgery at University (DCT) of Campinas.
Since 1987, the DCT has a team of specialists in trauma and acute care surgery actively working on training residents, teaching trauma league students, and promoting trauma prevention programs. The metropolitan area of Campinas has a population of 3.8 million, encompassing 20 cities. HC-Unicamp is one of the reference hospitals for the population using the public health system in this region, is a level 1 trauma center, has about 500 beds and a speci c Trauma-ICU for patients with high complexity injuries.
The medical records of 3.003 patients treated by exploratory laparotomy for penetrating or blunt abdominal trauma were reviewed. The records of patients with TDI were analyzed for demographic data, trauma mechanism, hemodynamic status at admission, grade and location of the diaphragmatic injury, herniated organs, diagnostic methods, interval between surgery, length of stay, trauma scores, associated injuries, morbidity, and mortality. The trauma scores were calculated using the Revised Trauma Score (RTS), Injury Severity Score (ISS), and the Trauma Injury Severity Score (TRISS).
TDI were graded according to the American Association for the Surgery of Trauma (AAST) injury scale [6] as follows: grade I, diaphragm contusion; grade II, laceration less than or equal to 2 cm; grade III, laceration greater than 2 cm and less than 10 cm; grade IV, laceration greater than 10 cm, with tissue loss less than or equal to 25 cm 2 ; and grade V, laceration with tissue loss greater than 25 cm 2 . Grade I injuries and chronic hernias (> 180 days) were excluded.
Our institutional protocol for treatment of TDI involves repair with a nonabsorbable suture, such as polypropylene, placement of an ipsilateral chest tube for drainage, and a nasogastric tube. The research protocol of this study was approved without restrictions, all methods were committee complies with all

Results
Of the 3.003 reviewed patients, 425 had TDI (14.1%). A diagnosis of TDH was established in 55 cases (1.8%). All further analyses were performed in theses 55 cases. TDH was predominant in males with 46 cases (83%).
The administration of a water-soluble contrast by nasogastric tube was performed in 2 cases, supporting the diagnosis of TDH.
Since 2008, at HC-Unicamp, hemodynamically stable trauma patients had access to a multi-slice CT scan, which was used to diagnose 7 cases (12%) (Figure 2 and 3). TDH was observed intraoperatively in 17 cases (32%); 14 of these were diagnosed by laparotomy and had associated intra-abdominal injury or hemodynamically unstable ( Figure 4); and 3 by laparoscopy.
Diagnostic peritoneal lavage (DPL) was performed in 5 cases, 3 were positive for intra-abdominal injury, 2 negative, and all cases of DPL were subjected to exploratory laparotomy. In the 2 negative cases of DPL, laparotomy was performed due to the suggestive herniation image in CXR and worsening of the patient's hemodynamics. Surgical procedures were performed within 24 hours in 42 cases (76%), between 24-72 h in 8 (14%), and 72-96 h in 3 cases (5%) Two cases were diagnosed and treated 60 and 180 days after the trauma, respectively.
Diaphragmatic grade III injury occurred in 21 patients (38%), grade IV in 19 (34%), grade V in 9 (17%), and grade II in 6 (11%) ( Figure 5). In SW, grade III injury was the most common, followed by grade II. In GSW, grades II, III, and IV had proportionally equal cases. In BT, grade IV was most common followed by grade III.
Associated extra-abdominal injuries were found in 35 cases (63.6%). The most common was pelvic fracture (14 cases), and others included long bone fractures (12 cases), traumatic brain injury (TBI) and face trauma (5 cases each), and spinal cord trauma (3 cases). Associated intrathoracic injuries were identi ed in 13 cases (23.6%). The most common injury was hemopneumothorax and fracture of the ribs (6 cases), followed by lung contusion (4 cases), cardiac injury (2 cases), and thoracic aorta injury (1 case).
The length of stay included a maximum of 61 days, and the average was 14 days.

Discussion
A recent review by the National Bank Trauma Date [2] showed that a majority of diaphragmatic injuries in BT cases were due to MVC (63%) and pedestrian-hit-by-car accidents (11%). In penetrating trauma, diaphragmatic injuries were due to GSW (66%). Penetrating wounds lead to minor lacerations, and take time to present any complications; however, large linear ruptures are common in BT, leading to immediate diaphragmatic herniation [7,8] BT injuries on the left side were more common than on the right side. In the literature, the incidence of left diaphragmatic injury was reported in 75% of cases, as opposed to right in 25% and bilateral in less than 2%, corroborating the ndings of our analysis [9,10] Explanation for this would be muscle weakness in the posterolateral area of the left diaphragm [11,12,13]. The lower incidence of injury to the right side is attributed to the protective effect of the liver [13]. Patients with rupture of the right hemidiaphragm had a higher prehospital mortality as a result of greater trauma severity and high grade of injury, typically associated with signi cant vascular tears in the inferior or retro-hepatic vena cava [8].
Nowadays, diagnosing a right-side hernia in severe trauma cases is possible due to e cient prehospital teams, which stabilize and transport the patient to a hospital timeously, due to the technology used in obtaining quality CT-scan images. Aun et al. [14] compared 97 cases of TDHs treated surgically with 146 cases seen in 12.276 consecutive autopsies of patients who died because they did not receive medical care. There was a higher incidence of diaphragmatic lacerations on the right side in the autopsy group (49.6%) than in the hospitalized group (14.4%) and hemorrhagic shock (80%) being the main cause of death in the autopsy group.
The diaphragm is closely related to the intrathoracic and intra-abdominal organs. At least 50% of patients with diaphragmatic injury also experience associated injuries [15]. Our series presented results similar to the literature [25] when evaluate the associated injuries.
Associated injuries include pelvic and long bone fractures, and injuries to intra-abdominal structures, such as the spleen and liver. [16] reported an 8% incidence of traumatic aortic injury and 32% traumatic brain injury in 68 patients with diaphragmatic rupture and recognized the importance of TDH as a marker of serious injury. There are few signs and symptoms for the immediate clinical diagnosis of an acute diaphragmatic rupture.

Meyers and McCabe
Hirano et al. [17] concluded that a diaphragmatic hernia should be suspected on the basis of the trauma mechanism and physical examination. Symptoms and signs attributable to diaphragmatic injury include respiratory distress, shock, shoulder pain, and the presence of bowel sounds in the chest, which can be unnoticed by environmental noise. The diagnosis of diaphragmatic injuries through imaging presents a challenge for trauma surgeons and radiologists [17]. A delayed diagnosis of a traumatic diaphragmatic injury resulting in diaphragmatic hernia is problematic due to the associated morbidity and mortality [18]. This mandates aggressive diagnostic evaluation of patients at risk for diaphragmatic injury to avoid long-term sequelae [19]. Despite advances in imaging methods, CXR is still useful in diagnosing or suspecting traumatic diaphragmatic hernia, and is the initial step for evaluation [17,20]. CXR allows the diagnosis in 30-60% of cases with left-sided lesions, against 17% with right-sided lesions [21,22]. A review of the literature reported that diagnosis on the basis of CXR was made in only 40.7% of cases [5].
Currently, for a broad assessment of associated injuries and the possibility of non-operative treatment, most hemodynamically stable patients should undergo a CT-scan. A multidetector CT-scan with multiple coronal and sagittal reforms contributes to the accuracy of the diagnosis [18,[23][24][25]. Image ndings such as diaphragmatic discontinuity, collar and dependent viscera sign, and intrathoracic herniation of the abdominal structures have a good sensitivity and high speci city for the diagnosis [3,20,26]. Magnetic resonance imaging, despite offering excellent resolution images for viewing the diaphragm, is not suitable for trauma patients in the acute phase, is not used in our hospital.
Focused assessment with sonography for trauma (FAST) is an essential propaedeutic tool in the initial assessment of a traumatized patient. Currently, extended FAST (e-FAST) is considered important in the diagnosis of thoracic cavity injuries, such as pneumothorax, pulmonary contusion, hemothorax, and abdominal or pericardial free uid; however, there is insu cient evidence of its use in the diagnosis of diaphragmatic injuries.
Ivatury et al. [27,28] reported that the diagnostic accuracy of laparoscopy was excellent for hemoperitoneum and diaphragmatic lacerations, validating laparoscopy as an excellent tool to evaluate the diaphragm in penetrating injuries. Thus, expanding the spectrum of minimally invasive surgery can play an important role in the diagnosis and treatment of patients with diaphragmatic herniation, but minimally invasive approach is highly dependent on the surgeon's expertise to be safe and have better outcomes. DPL was propaedeutic for the evaluation of blunt abdominal trauma in the past. However, the literature demonstrates the inadequacy of DPL in the diagnosis of acute diaphragmatic rupture [29]. Although 5 cases underwent this method, it has not been used in our hospital since 2010 due to low speci city, large number of false negative ndings in isolated diaphragmatic injuries, and high rates of non-therapeutic laparotomies.
Treatment of TDH involves surgical repair, either laparoscopy or laparotomy. Acutely, laparotomy is the preferred approach because many cases have associated intra-abdominal injuries. Herniated viscera must be reduced into the abdominal position and the thoracic cavity must be inspected. Lacerations are sutured with non-absorbable mono laments, followed by the placement of an ipsilateral chest tube for drainage, and a nasogastric tube for stomach decompression. For large defects, the use of prosthetic materials such as mesh may be necessary.
Most complications are related to associated injuries, not attributed to the diaphragmatic injury itself. Patients with hemodynamic instability during admission and BT had higher complication rates [13].
This collective review re ects a mortality rate of 20%, being higher in the group with BT and ISS > 25. In the literature, mortality rates range from 17% to 51% [5,16,30,31]. High morbidity and mortality were mainly related to associated injuries, characterizing diaphragmatic injury as a marker of potentially serious and lethal injuries.
This study has limitations related to its retrospective design. The potential biases may occur because the analysis of established data shows the change in perspectives and tendencies of patient assessment. Data collection also relied on chart review at the participating institution which may contain errors or omissions.

Conclusion
Traumatic diaphragmatic hernias are uncommon but signi cant in trauma patients. Isolated injuries were rare, and the presence of associated injuries increased morbidity and mortality. Despite advances in imaging methods, chest radiography in the trauma bay is useful as an initial approach. CT identi ed associated injuries and had high speci city. Despite the use of laparoscopy, laparotomy was the most frequent approach for repair. Consent for publication: Our study is retrospective design (from 1990 to 2017) with long collection data, there is no participant in current follow-up to which the informed consent form can be applied or any subsequent intervention. Was a request for release of informed consent term. After analyzing the Institutional Research Board (research ethics committee -University of Campinas), this release of the term was approved.
For used images (intraoperative and CT-scan) the consent informed term is attached in the supplementary documents.
Attachments also include the director of HC-Clinics UNICAMP consent to carry out the evaluation of medical records and research with the medical records studied. All methods were committee complies with all regulations, following the Helsinki statement.
Available data or materials: All data generated or analysed during this study are included in this published article and its supplementary information les.
Competing interests: The authors declare that they have no competing interests Funding: The author(s) received no speci c funding for this work Authors contributions: VFK participated in the acquisition, analysis and interpretation of data, participated in the study's design and drafted the manuscript. TAC: contributed to the study's design and the revision of the manuscript. RBC: revision of the manuscript. ESH: interpretation of data, contributed to the study's design and the revision of the manuscript. GPF conceived the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the nal manuscript.   Diaphragmatic grade vs trauma mechanism SW: stab wound; GSW: gunshot wound Coronal section: sign of the collar, loss of diaphragmatic contour, and pulmonary contusion.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.