ATS
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Ann Thorac Surg 2008;85:1044-1048. doi:10.1016/j.athoracsur.2007.10.084
© 2008 The Society of Thoracic Surgeons
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Author home page(s):
Lorenzo E. Ferri
David S. Mulder
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hanna, W. C.
Right arrow Articles by Mulder, D. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hanna, W. C.
Right arrow Articles by Mulder, D. S.
Related Collections
Right arrow Diaphragm

Original Articles: General Thoracic

The Current Status of Traumatic Diaphragmatic Injury: Lessons Learned From 105 Patients Over 13 Years

Waël C. Hanna, MDa, Lorenzo E. Ferri, MDa,*, Paola Fata, MDb, Tarek Razek, MDb, David S. Mulder, MDa

a Division of Thoracic Surgery, McGill University Health Centre, Montréal, Québec, Canada
b Division of Trauma, McGill University Health Centre, Montréal, Québec, Canada

Accepted for publication October 24, 2007.

* Address correspondence to Dr Ferri, Montréal General Hospital, 1650 Cedar Ave, L9-112, Montréal, Québec, H3G 1A4, Canada (Email: lorenzo.ferri{at}muhc.mcgill.ca).


   Abstract
Top
 Abstract
Introduction
 Patients and Methods
 Results
 Comment
 References
 
Background: Our understanding of traumatic diaphragmatic injury (TDI) is based primarily on outdated retrospective series. We sought to reexamine present day patterns of diagnosis, associated injuries, predictors of mortality, and long-term outcomes of this condition.

Methods: A prospectively entered trauma database from the Montréal General Hospital was reviewed for patients admitted with a TDI from 1993 to 2006. Hospital charts were reviewed, and patient characteristics, mechanism of injury, associated injuries, operative management, and postoperative outcomes were recorded. Logistic regression was used to identify predictors for mortality.

Results: Identified were 105 patients with TDI consisting of blunt in 37% and penetrating in 63%. Only 23% of TDI were diagnosed on initial chest roentgenogram. External wounds in penetrating TDI cases were found in the abdomen alone in 19%, in the chest alone in 46%, and in both in 35%, which was associated with intraabdominal organ injury in 83%, 55%, and 87%, respectively. Less than half of patients had a diaphragmatic hernia. Lung, chest wall, and thoracic organ injuries were more common in blunt trauma, but there was no significant difference between abdominal injuries in both mechanisms. Overall mortality from TDI was 18%, and there was no difference between blunt and penetrating injury. In blunt trauma, brain injury and an Injury Severity Score (ISS) exceeding 15 were independently associated with increased death. In penetrating trauma, only an ISS exceeding 15 predicted death.

Conclusions: Traumatic diaphragmatic injury remains a challenge to diagnose and treat, primarily due to the presence of associated injuries. The high incidence of intraabdominal organ injury, irrespective of the site of penetrating wound, dictates a transabdominal approach for exploration and repair. Severity of associated injuries (ISS) predicts death.


   Introduction
Top
 Abstract
 Introduction
Patients and Methods
 Results
 Comment
 References
 
Traumatic diaphragmatic injury (TDI) is an under-recognized sequela of thoracoabdominal trauma and represents a diagnostic challenge to the surgeon confronted with the multiply injured patient. Diaphragmatic injury has been reported in 0.8% to 1.6% of patients hospitalized for blunt trauma [1], and up to 8% of all patients undergoing surgical exploration for trauma will have an incidental finding of diaphragmatic injury [2]. By virtue of its location between the abdomen and chest, an injured diaphragm is almost never the sole injury and has an associated injury rate approaching 100% [3].

Death as a result of TDI has largely been ascribed to associated injuries rather than to the diaphragmatic injury itself [4]. Rupture of the diaphragm due to blunt trauma represents the presence of a significant crush or deceleration force dissipated in the abdominopelvic cavity, and in prior studies has been reported to impart a higher mortality rate than TDI due to penetrating trauma [2, 4].

Our current knowledge on the characteristics, treatment, and clinical sequelae of TDI is based on small retrospective series, the most recent of which is at least 20 years old [5]. A prospective study published in 1997 examined the incidence of diaphragmatic injury and the value of thoracoscopy only in patients with penetrating trauma to the left lower thorax [6]. Advances in prehospital care, trauma management, and critical care during the last decade mandate a new look at this condition. Given the absence of current literature, we sought to reexamine the present day patterns of diagnosis, the incidence of associated injuries, and the predictors of death in patients with TDI.


   Patients and Methods
Top
 Abstract
 Introduction
 Patients and Methods
Results
 Comment
 References
 
Patients
We reviewed a prospectively entered trauma registry at the Montréal General Hospital, a busy level 1 trauma center, for all patients with TDI from 1993 to 2006. Permission from the Director of Professional Services and the Chairman of the Department of Surgery was obtained for the study, thereby waiving the requirement for Institutional Review Board approval. Informed consent was obtained from patients who presented for long-term follow-up.

During this 13-year period, trauma admissions averaged 1900 per year, 31% of which were for an Injury Severity Score (ISS) exceeding 15 (signifying severe injury). The mechanism of injury managed at our institution is primarily blunt (approximately 85%). Patients were selected if they had a diagnosis of diaphragmatic injury at any time during the hospitalization. Initially, 128 patients were identified and their charts individually reviewed for patient characteristics, mechanism of trauma, associated injuries, and mode and timing of diagnosis and repair. Of these, 102 patients were included in the study after 26 were eliminated owing to uncertainty of the diagnosis or lack of relevant information. Efforts were made to identify patients with initially missed diaphragmatic injuries who presented with a hernia at a later date by reviewing the databases of the two thoracic surgeons at our institution. Three patients were found and included in this study, bringing the total number of patients to 105.

Results of initial chest roentgenograms were entered as interpreted by the trauma team leader (surgical senior resident, emergency physician, or trauma surgeon) in the trauma bay and compared with final radiologists’ readings and the intraoperative findings. Traumatic diaphragmatic injury characteristics were noted and the method and approach of repair recorded. Postoperative data were collected for the length of stay, complications, and mortality. The data are presented as median (range) or total (percentage).

Statistical Analysis
Results of blunt vs penetrating trauma were compared using the Student t test or {chi}2, and differences with a p < 0.05 were considered significant. Survivors and nonsurvivors were compared with logistic regression. The dependant variable studied was death in the context of both blunt and penetrating TDI. For blunt TDI, the variables entered in the univariate analysis were ISS (p = 0.044), head trauma (p < 0.001), and rib fractures (p = 0.023). For penetrating TDI, the variables used were hollow viscus injury (p = 0.044), ISS (p = 0.032), and age (p = 0.016). The cutoff for the multivariate analysis was p = 0.05. Long-term follow-up was attempted for all living patients. Those contacted were requested to undergo a chest roentgenogram and complete a structured 7-point questionnaire, scaled on a Likert (0 to 5) score, aimed at symptoms thought to be associated with diaphragmatic hernia.


   Results
Top
 Abstract
 Introduction
 Patients and Methods
 Results
Comment
 References
 
A total of 105 patients (91 men [86.7%]) with traumatic diaphragmatic hernia were identified from a prospectively entered database from 1993 to 2006. The median age was 34 years old (range, 16 to 79). Penetrating TDI occurred in 66 of 105 (62.9%) compared with 39 blunt TDIs (37.1%). Stab wounds caused 49 of 66 (74.2%) of the penetrating injuries, and gunshot wounds caused the rest. Motor vehicle crashes resulted in 31 of 39 blunt injuries (79.5%), whereas the rest were divided equally between fall and crush injuries.

Traumatic diaphragmatic hernia was present in 46 of 105 patients (43.8%) and was localized to the left hemidiaphragm in 35 (76.1%). The incidence of hernia in blunt trauma (24 of 39, 61.5%) was significantly higher than in penetrating trauma (22 of 66, 33.3%; p < 0.001). Also, the mean diaphragmatic defect size in blunt traumatic hernia was significantly larger than that encountered in hernias of penetrating trauma (10.6 ± 5.0 vs 3.1 ± 2.8 cm, p < 0.001). The stomach was the organ most frequently found herniating into the chest (22 of 46, 47.8%), followed by the spleen (12 of 46, 26.1%), and small bowel (6 of 46, 13.0%).

The TDI was diagnosed within 6 hours of injury in 95 of 105 patients (90.5%), diagnosed and repaired later within the same hospital stay (delayed phase) in 7 (6.7%), and was missed and treated 4 months to 3 years later, after the diagnosis of diaphragmatic hernia, in 3 (2.7%). Only 22.8% (24 of 105) of diaphragmatic injuries were diagnosed by the initial trauma bay chest roentgenogram as read by the trauma team leader. However, attending radiologists who read the same roentgenograms were more able to correctly identify a diaphragmatic injury (46 of 105, 43.8%; Fig 1). Although computed tomography scan identified an additional 10.5% of TDI (11 of 105) preoperatively, 57.1% (60 of 105) were discovered only at the time of surgical exploration for associated injuries.


Figure 1
View larger version (10K):
[in this window]
[in a new window]

  		Fig 1. Methods of diagnosis for traumatic diaphragmatic injury. (CXR = chest x-ray; CT = computed tomography scan; Exp Lap = exploratory laparotomy.)

 
Most patients with TDI had severe associated injuries as evidenced by a median ISS of 22 (range, 9 to 75). The ISS in blunt TDI patients was significantly greater at 36 (range, 10 to 75) than the ISS of 21(range, 9 to 75) in patients with penetrating TDI patients (p < 0.001). Head injuries and pelvic fractures were associated only with blunt TDI, whereas cardiac injuries were only associated with penetrating TDI. Lung, chest wall, or thoracic organ injury was more common in blunt trauma, but there was no significant difference in abdominal hollow viscus or solid organ injuries between both mechanisms (Table 1).


View this table:
[in this window]
[in a new window]

  		Table 1 Injuries Associated With Traumatic Diaphragmatic Injury
 
To determine the most appropriate approach to repair the diaphragm in patients with penetrating injuries, we explored the relationship between the site of penetrating wound and location (intrathoracic vs intraabdominal) of associated injury requiring surgical attention. When the site of penetration was in the abdomen, there was an associated intraabdominal organ injury in 83.3% (10 of 12). When the penetrating wound was in the chest, associated organ injury in 54.8% (17 of 31) was again primarily within the abdomen. In cases of penetration of both the chest and abdomen, there was an equally high 86.9% rate (20 of 23) of intraabdominal injury (Fig 2).


Figure 2
View larger version (27K):
[in this window]
[in a new window]

  		Fig 2. Relationship between the site of traumatic diaphragmatic injury—abdomen (gray bar), chest (black bar), or both—and the association with major organ injury. Abdominal injuries were predominant, regardless of site of external penetrating wound. The clear bars designate no organ injury.

 
In 95 patients, the diagnosis and repair occurred acutely at presentation to the trauma bay. Most defects were repaired using nonabsorbable suture in a simple interrupted, continuous, or figure-of-eight fashion. Two patients were repaired with absorbable suture. Given the high rate of hollow viscus injury, none of the defects were repaired using mesh in the acute setting.

The approach to the repair (laparotomy vs thoracotomy) in the acute setting was dictated by the need to explore for associated life-threatening injuries. All patients underwent an exploratory laparotomy, and 6 patients underwent both a laparotomy and a thoracotomy. The major indication for thoracotomy was hemodynamic instability caused by injury to the heart or great vessels. In the delayed phase, either during the same admission (posttrauma day 2 to 11) or after discharge (4 months to 3 years), the approach to repair was determined by the preference of the operating surgeon (Table 2).


View this table:
[in this window]
[in a new window]

  		Table 2 Management of Traumatic Diaphragmatic Injury Within the Acute and the Delayed Settings
 
We found the rate of pulmonary complications and pneumonia was significantly more prevalent in blunt TDI (12 of 39, 30.7%) than in penetrating (3 of 66, 5.1%). Empyema was a rare complication, occurring in only 4 patients; however, three of these were associated with a hollow viscus injury.

Overall mortality from TDI was 18.1% (19 of 105). There was no significant difference between the mortality rates of blunt TDI at 17.9% (7 of 39) and penetrating TDI at 18.5% (12 of 66; p = 0.776). The 19 patients who died had associated injuries requiring operative intervention, whereas only 74.4% (64 of 86) of survivors had associated injuries requiring operative intervention (p = 0.044). The major causes of death were traumatic brain injury in 31.6% (6 of 19) and hemorrhagic shock in 68.4% (13 of 19).

Univariate analysis was used to identify possible predictors of death after blunt and penetrating TDI. The variables found to be associated with increased risk of death were ISS (p = 0.044), head trauma (p < 0.001), and rib fractures (p = 0.023) in blunt trauma; and hollow viscus injury (p = 0.044), ISS (p = 0.032), and age (p = 0.016) for penetrating trauma. Those same variables were tested in a multivariate logistic regression analysis, with a cutoff value of p = 0.05, to define the predictors for death. In blunt TDI, death was predicted by traumatic brain injury and ISS exceeding 15. Only ISS exceeding 15 predicted death in patients with penetrating TDI (Table 3).


View this table:
[in this window]
[in a new window]

  		Table 3 Independent Predictors of Mortality in Traumatic Diaphragmatic Injury on Multivariate Regression Analysis
 
Of the 76 patients alive at discharge, we were able to identify and locate only 13 (17.1%) to assess long-term outcomes. All patients consented to follow-up (mean, 48.5 months; range, 1 to 120 months). A standard 7-point questionnaire was administered to screen for symptoms of diaphragmatic hernia (chest pain, regurgitation, heartburn, dysphagia, shortness of breath, nausea/vomiting, abdominal pain) using a 0 to 5 Likert scale. Only 1 patient had significant symptoms manifested as heartburn (score 4 of 5).

All 13 patients subsequently underwent an anteroposterior and lateral chest roentgenogram to assess the anatomic integrity of the repair. The chest roentgenograms, as interpreted by a radiologist, were normal in 6 patients, another 6 had mild pleural thickening, and 1 had a recurrence of a left diaphragmatic hernia. Overall, 2 patients had recurrence: 1 acutely on the first postoperative day and another at 14 months, as mentioned. The repair in both had been done with absorbable sutures.


   Comment
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Comment
References
 
In this large, single-institution series, we reviewed the characteristics and outcomes of patients with TDI during a period of 13 years. Because most of the data on TDI are more than a decade old, and although advances in critical care have dramatically changed the way we manage trauma patients, this series sheds new light on the patterns of injury, treatment, and outcomes of TDI.

In previously published series, the incidence of blunt injuries to the diaphragm was almost threefold that of penetrating injuries [2, 4, 7]. However, two studies looking specifically at penetrating left lower thoracic injuries identified a high incidence of diaphragmatic rupture (42% and 24%, respectively), but no comparison was made with blunt trauma at that time [6, 8]. In the present study, we have shown the exact opposite, with penetrating trauma accounting for 63% of TDI vs 37% for blunt. This predominance of penetrating-mediated TDI is even more impressive because only a small proportion of traumas managed by our center are penetrating (15%). This is likely due to an increased awareness of the treating physicians to the association of diaphragmatic injuries in patients with penetrating wounds to the left upper quadrant or left lower chest, as has been demonstrated by several authors [6, 8, 9].

It is important to distinguish between isolated injury to the diaphragm and the occurrence of a diaphragmatic hernia secondary to this injury. Almost half of the patients we reviewed had developed a traumatic diaphragmatic hernia. We demonstrated that diaphragmatic hernia occurs more frequently in blunt TDI, and that the size of the diaphragmatic defect is also significantly larger in blunt trauma patients. This finding is likely due to the increased pressure gradient across the diaphragm associated with blunt crush injuries and is consistent with that reported by prior studies [2, 4, 5].

The plain chest roentgenogram has been reported to be useful for the diagnosis of diaphragmatic injury, with sensitivities ranging of 30% to 62% [10] in the absence of a hernia, and up to 94% in the presence of a hernia [11]. Most of these studies used chest roentgenogram interpretations by a radiologist. On the other hand, we found that chest roentgenogram is much less accurate in diagnosing a TDI (22.8%) when interpreted by the trauma team leader on the spot in the trauma bay. Although the trauma team leader was able to identify all cases of diaphragmatic hernia, many cases of injury without an obvious hernia were missed. When those same images were submitted to interpretation by attending radiologists, the sensitivity of the radiograph increased to 43.8%. Although no effort was made to communicate the diagnosis to the radiologist, this is still a retrospective study in which total blinding cannot be ensured. This figure may well be an overestimate of the radiologist’s ability to identify missed diaphragmatic injuries. Nevertheless, an attending radiologist in the trauma bay is a rare sight indeed; it is thus imperative for surgeons and emergency physicians to improve their roentgenogram interpretation skills and maintain a high index of suspicion for TDI whenever the mechanism of injury is suggestive.

Intraoperative identification remains the gold standard for the diagnosis and treatment of TDI. Exploration of the abdomen, by laparotomy or laparoscopy, has traditionally been advocated because it allows concurrent examination of the often-injured abdominal organs [8, 12–14]. The data we have provided support this approach. For penetrating TDI, we found that intraabdominal organs were injured in most cases, irrespective of the location (thoracic vs abdominal) of the penetrating wound. In rare cases, a thoracotomy was required to repair select intrathoracic organs, primarily the heart or great vessels.

Long-term follow up of any trauma patient is an arduous task. This patient population is young, mobile, and particularly those on the receiving end of penetrating trauma, do not want to be found. Our attempts to obtain long-term clinical and anatomic follow-up bear this true: we were able to locate only a small minority of evaluable patients. We were nonetheless able to gain some information: 2 patients with recurrences were identified, both of which had undergone repair with absorbable suture. On the other hand, all patients whose repair was done with nonabsorbable suture remained without recurrence. These findings, although not definitive, provide a strong argument against the use of absorbable suture in the management of TDI.

Despite the greater incidence of life-threatening injuries (pelvic fractures and traumatic brain injury) and a higher ISS in blunt TDI, we did not identify an increased mortality rate for this subset compared with penetrating TDI. This differs significantly from the results of previous studies that almost uniformly cite an increased mortality rate for blunt TDI [2–5, 7, 12, 14]. This finding is likely due to the significant advances in trauma, critical care, and interventional radiology, allowing an improved survival in the polytrauma patient. We identified ISS as an independent predictor for mortality in both penetrating and blunt TDI, and traumatic brain injury as a predictor in blunt TDI alone.

In conclusion, we have found that although many of the long-established beliefs about TDI are still true, changes in trauma and critical care have altered the outlook for these patients. A preoperative diagnosis of TDI can be made in almost half of the cases; however, the trauma team must improve their roentgenogram interpretation skills. Despite this, penetrating TDI accounted for a higher proportion of injury than was found in prior studies, possibly due to a heightened awareness of this condition and a liberal use of diagnostic studies. The high association of intraabdominal injuries, irrespective of the location of penetrating wounds, mandates that TDI be approached from the abdomen in patients who require exploration. Stable patients with a suspicion of diaphragmatic injury may be evaluated by thoracoscopy. Blunt trauma by itself is no longer a predictor of death from TDI. Rather, severe injuries, as evidenced by ISS exceeding 15, are associated with a higher mortality rate.


   References
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Comment
 References
 

  1. Epstein LI, Lempke RE. Rupture of the right hemidiaphragm due to blunt trauma J Trauma 1968;8:19-28.[Medline]
  2. Shah R, Sabaratnam S, Mearns A. Traumatic rupture of diaphragm Ann Thorac Surg 1995;60:1444-1449.[Abstract/Free Full Text]
  3. Ward RE, Flynn TC, Clark WP. Diaphragmatic disruption due to blunt abdominal trauma J Trauma 1981;21:35-38.[Medline]
  4. Rubikas R. Diaphragmatic injuries Eur J Cardiothorac Surg 2001;20:53-57.[Abstract/Free Full Text]
  5. Rodriguez-Morales G, Rodriguez A, Shatney CH. Acute rupture of the diaphragm in blunt trauma: analysis of 60 patients J Trauma 1986;26:438-444.[Medline]
  6. Murray JA, Demetriades D, Cornwell 3rd EE, et al. Penetrating left thoracoabdominal trauma: the incidence and clinical presentation of diaphragm injuries J Traum 1997;43:624-626.
  7. Beauchamp G, Khalfallah A, Girard R, Dube S, Laurendeau F, Legros G. Blunt diaphragmatic rupture Am J Surg 1984;148:292-295.[Medline]
  8. Murray JA, Demetriades D, Asensio JA, et al. Occult injuries to the diaphragm: prospective evaluation of laparoscopy in penetrating injuries to the left lower chest J Am Coll Surg 1998;187:626-630.[Medline]
  9. Ochsner MG, Rozycki GS, Lucente F, Wherry DC, Champion HR. Prospective evaluation of thoracoscopy for diagnosing diaphragmatic injury in thoraco-abdominal trauma: a preliminary report J Trauma 1993;34:704-710.[Medline]
  10. Payne Jr JH, Yellin AE. Traumatic diaphragmatic hernia Arch Surg 1982;117:18-24.[Abstract/Free Full Text]
  11. Waldschmidt ML, Laws HL. Injuries of the diaphragm J Trauma 1980;20:587-591.[Medline]
  12. Grimes OF. Traumatic injuries of the diaphragm. Diaphragmatic hernia. Am J Surg 1974;128:175-181.[Medline]
  13. Broos PLO, Rommens PM, Carlier H, Van Leeuwen JE, Somville FJ, Gruwez JA. Traumatic Rupture of the diaphragm. Review of 62 successive cases. Int Surg 1989;74:88-92.[Medline]
  14. Demetriades D, Kakoyiannis S, Parekh D, Hatzitheofilou C. Penetrating injuries of the diaphragm Br J Surg 1988;75:824-826.[Medline]



This article has been cited by other articles:


Home page
 ICVTSHome page
R. E. Al-Refaie, E. Awad, and E. M. Mokbel
Blunt traumatic diaphragmatic rupture: a retrospective observational study of 46 patients
Interactive CardioVascular and Thoracic Surgery, July 1, 2009; 9(1): 45 - 49.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Author home page(s):
Lorenzo E. Ferri
David S. Mulder
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hanna, W. C.
Right arrow Articles by Mulder, D. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hanna, W. C.
Right arrow Articles by Mulder, D. S.
Related Collections
Right arrow Diaphragm

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
ANN THORAC SURG ASIAN CARDIOVASC THORAC ANN EUR J CARDIOTHORAC SURG
J THORAC CARDIOVASC SURG ICVTS ALL CTSNet JOURNALS