Abstract

            Objectives:

 
          This study was conducted in a level I trauma centre to review the outcome of surgical repair of Traumatic Diaphragmatic Rupture (TDR) and to identify the predictors of mortality and hospital stay

             Methods:


           Between January 1990 and January 2001, consecutive patients with TDR and ISS (Injury Severity Score) > 12 were identified from a prospective trauma registry. Hospital charts of all eligible patients were reviewed for demographic data, mechanism of injury, mode of diagnosis, type of surgical repair, need for ventilatory support, ICU & hospital stay and mortality. We conducted a stepwise regression analysis (logistic regression for mortality, and multiple regressions for hospital stay).

             Results:


            Of the 59 patients included in the study 44 (75%) were males. Their mean age was 43±18years and their average ISS was 39 ± 15 Blunt injuries (85%) and left sided ruptures (73%) were the most common. Frontal and side impacts were equally distributed. Twenty eight (66%) patients were drivers. Forty five (79%) patients were ventilated following the diaphragm repair. A significantly higher proportion of patients with blunt injuries required ventilatory support compared with penetrating injuries (93% vs. 38%, P <0.05). The mortality rate was 7%. Older age was a significant predictor of mortality (Odds ratio = 1.2, 95% CI = 1.1-1.4, P = 0.04). ISS (Odds ratio =1.1, 95% CI = 0.98 -1.2, P = 0.08) and need for ventilation (Odds ratio=1.02, 95% CI = 0.97-1.12, P=0.09) revealed trends towards mortality, but were not statistically significant. Hospital stay was predicted by the ISS (B=0.09, P=0.05) score and need for postoperative ventilatory support (B=0.26, P=0.04).


            Conclusions:


            This is the largest Canadian series from a single trauma centre. It revealed that older age is a major predictor of mortality whereas the need for ventilatory support did not predict the overall mortality.

           
Introduction

             Traumatic Diaphragmatic Rupture (TDR) is an uncommon and a challenging problem in trauma surgery. Its incidence has been reported in the literature to range from 0.8 %-15% (1,2,3,4,5). The first report of TDR was by Sennertus in 1541 who described a herniated stomach in a patient who died seven months after a self inflicted stab (6). Since then more reports of TDR have been published(7,8,9). TDR can result from both blunt and penetrating injuries and is often associated with other multiple severe injuries. It is considered as a marker of serious injuries(10,11). Prompt recognition of TDR still poses a significant clinical challenge attributed to the lack of specific clinical characteristics in the face of more obvious injuries. A delayed or missed diagnosis and subsequent treatment is associated with potential life threatening complications. The high morbidity and mortality of TDR are attributed to the severity of the associated injuries.

Objectives

           In this retrospective review, we are presenting the largest Canadian experience from a single trauma centre. The objectives of the study are to identify the mechanism of injury, the clinical characteristics, the diagnostic & therapeutic modalities and the main outcome following the surgical repair. We also aimed to define the predictors of the outcome which mainly included the length of hospital stay and the mortality rate.

Methods

           Hamilton General Hospital is the second largest adult tertiary trauma centre in Ontario, Canada. It covers most of South-Eastern Ontario; a population of 2 million. About 450 patients with ISS >12 are seen annually.

Eligible Criteria

          Only patients who underwent surgical repair for the TDR between Jan 1990 and Jan 2001 were included in the study. As we were studying the outcome after surgical repair we excluded all patients who were diagnosed at autopsy as well as those who did not un dergo a surgical repair. TDR was defined as any tear or laceration that resulted in the communication between the thoracic and abdominal cavities.

Data Collection

          Data were collected from a prospectively managed trauma data registry. A computerized database search using the terms "traumatic diaphragmatic rupture", "traumatic diaphragmatic laceration" and "traumatic diaphragmatic hernia" was utilized to identify patients with possible TDR. All hospital charts of eligible patients were retrieved and reviewed by four independent investigators. We abstracted information about demographic data, mechanism of injury, ISS, homodynamic status on presentation, diagnostic procedures, chest X-ray (CXR) findings, associated injuries, operative findings which include indication, herniation, and repair, duration of ventilation, complications, length of stay in the ICU, the total hospital stay and mortality. Data from charts and the trauma database were entered into a computerized database for statistical analysis (SPSS version 10, professional). Data were entered by one investigator and checked for accuracy by the three other investigators.

Statistical Analysis

                   Continuous variables were presented as means and standard deviations whereas categorical variables were presented as percentages. We compared the differences between continuous variables using the students' T-tests, and differences between proportions by the chi-square test. To evaluate factors associated with mortality and hospital stay, we conducted a stepwise regression analysis (logistic regression for mortality, and multiple regressions for hospital stay). We entered the following "independent" variables into the analysis: Age, ISS, gender, need for postoperative ventilation, side of rupture, mechanism of injury, pre-operative hypotension, delay of treatment, and herniation of abdominal contents into chest cavity. We considered statistical significance at the alpha=0.05 level and all statistical tests were two-tailed. 

Results

           Identification of Patients with Traumatic Diaphragmatic Rupture During the specified period from January 1990 to January 2001, we identified a total of 4670 patients admitted to the trauma center with an ISS>12. Of those, 85 patients were found by a computer search to be potentially eligible. Following retrieval of hospital charts for these 85 patients, 26 patients were ultimately excluded since they did not undergo surgical intervention, thus 59 patients were available and eligible for study inclusion. The overall incidence of TDR in our series was 1.8%, and the incidence of surgically treated TDR was 1.2%
   
Demographics

       Of the 59 patients, the majority were males (n = 44). The male to female ratio was 3:1.The average age was 43 (Ranges from 18-82). There were 50 blunt and 9 penetrating injuries. Of the 50 blunt traumas, 42 (84%) were due to motor vehicle accident (MVA), four (8%) were pedestrian, three were falls (6%) and one (2%) was the result of a motorcycle accident. There were nine (18%) stabs and two (4%) GSW. Of the 40 MVA patients, 20 (40%) had head on collision, thirteen had left sided impacts, five had right-sided impacts and four had roll over (Table I).
 

 Patient posi-tion in their vehicles and the side of impact did not reveal differences in the type of injuries sustained by the patients (Table 2).
 

 Forty-three patients (73 %) sustained an injury to the left hemi-diaphragm, whereas 15 patients (25 %) had the right hemi-diaphragm injured and there was only one patient who sustained bilateral diaphragmatic rupture. Ten patients (17%) presented in a state of hypovolemic shock on admission to the trauma suite and required emergency laparotomy.
 

 Diagnosis of TDR

The correct diagnosis was made preoperatively in 36 patients (61%), most often with chest roentgenograms (CXR) (34) (Table 3).
 

Additional findings on chest X-ray included rib fractures (51%), hemothorax (47%), pneumothorax (47%), and contra-lateral shift of the mediastinum (4%). One patient presented with clinical picture of respiratory distress, tension pneumothorax and flail chest. The patient was treated with a needle decompression that was negative. Intra-operative diagnosis was made on 23 patients (39%). Delayed diagnosis (> 24 hours) occurred in eight patients; three were diagnosed within 48 hours after trauma and two patients within three days and another three patients within (4,9) and 15 days respectively. Diagnostic peritoneal lavage was performed in only six patients (Grossly positive in five). Focused Abdominal Sonography for Trauma (FAST) was conducted on five patients. FAST was positive in one patient who had free fluid in the peritoneal cavity secondary to liver injury.

 Treatment

The majority (93%) of the patients were repaired through a trans-abdominal approach. Three patients were repaired through a Thirty-four patients had an intra-abdominal organ herniating into the chest. The stomach was the most commonly herniated organ (48%), followed by the liver and the spleen (14% each). Twenty-eight (47%) patients had a repair of an associated intra-abdominal injury in addition to the diaphragmatic repair, whereas 31 patients had a repair of the diaphragm only. Of the 44 patients who had associated abdominal injuries, splenectomy was the most common operation (38%).

Associated Injury Patterns

           Fifty-seven patients had other associated injuries (Table 4).
 

Thoracic injuries were the most commonly associated injuries (86%). The liver was the most commonly injured intra abdominal organ (57%) followed by the spleen (39%). Pelvic fracture was the most common orthopedic injury (49%) followed by spinal injuries (33%).

Outcomes

           Twelve patients were extubated after surgery and were admitted to the Step Down Unit (SDU) except for one patient who was kept in the intensive care unit (ICU) for monitoring overnight. Of Those twelve patients, seven were due to blunt injury. The mean duration of ventilation was 10.13± 14.4. A significantly higher proportion of patients with blunt trauma required ventilatory support compared to those with penetrating injuries (93% vs. 38%, P<0.05) (Table 5). There were 22 (37%) complications, primarily pulmonary complications (5 had Pneumonia, 2 ARDS and 2 Empyema) (Table 6). Two patients had pancreatitis and one patient developed an intra-abdominal abscess in the LUQ that was drained percutaneously. One patient had MI and another developed acute renal failure that required dialysis. The total LOS in the ICU was 11.3±14.8. There were only four deaths with a mortality rate of 7%. None of the deaths was due the diaphragmatic rupture itself. One death was the result of severe brain edema, one was due to acute myocardial infarction and two patients developed multiple organ failure and succumbed to their injuries weeks after the trauma.

Predictors of Mortality and Hospital Stay

           A logistic regression model was created to identify the predictors of mortality. Older age was a significant predictor of mortality (Odds ratio= 1.2, 95% CI =1.1-1.4, P=0.04). Injury severity score (Odds ratio=1.1, 95%CI, 0.98-1.2, P=0.08) and the need for ventilation (Odds ratio=1.02, 95% CI = 0.97-1.12, P=0.09) revealed trends towards mortality, but were not significant. Hospital stay was predicted by ISS score (B=0.09, P=0.05) and need for ventilation post-operation (B=0.26, P=0.04).

 Discussion:
            Traumatic rupture of the diaphragm is an uncommon entity in trauma with an incidence that ranges between 0.4% -15%.(1,2,3) . The outcome of acute traumatic rupture of the diaphragm is determined by the severity of the associated injuries. Because of the difference in pressure gradient between the thoracic and abdominal cavities which can reach up to 100 mm Hg during Valsalva maneuver 12, intraabdominal organs can herniate into the thoracic cavity. Therefore all acute TDR should be recognized and repaired to avoid the long term potential for herniation, incarceration and strangulation of abdominal viscera (12,13, 14, 15, 16). Diagnosis of TDR posses a challenge to the trauma surgeon because of the lack of specific clinical signs and symptoms and the severity of other associated injuries. Many imaging techniques have been used such as chest X-ray (CXR), CT scan, upper GI series, peumoperitoneum(17, 18, 19)but they all lack sensitivity and specificity. CXR is the most valuable investigation in the diagnosis of diaphragmatic injuries(20). Brasel 21 has classified CXR findings in TDR into diagnostic or suspicious. We have classified the findings into diagnostic, highly suspicious and suggestive (Table 3). In Our series CXR was diagnostic in 17%, highly suspicious in 30.5% and suggestive in 52.5%. Although diagnostic peritoneal lavage (DPL) is a useful and sensitive diagnostic test to role out an abdominal source in the hemodynamically unstable patient, it has a low yield in the diagnosis of TDR and has a high false negative rate that ranges between 15% to 60 % (21, 22). The egress of peritoneal fluid through a chest tube would make a definitive diagnosis. We have used DPL in seven patients and in none of them TDR was diagnosed. Since Heiberg's first report of the use of CT scan in the diagnosis of TDR by in 1980 (23), it has become the modality of choice for evaluating hemodynamically stable patients with blunt abdominal trauma (24). However, CT is an inconsistently reliable diag nostic test in the diagnosis of TDR (25, 26). It has low sensitivity and specificity for the diagnosis of TDR especially in intubated patients(27). Laparoscopy is increasingly used as a diagnostic and therapeutic tool for both acute and chronic TDR(28). Laparoscopy confirmed the diagnosis in seven patients in our series. Patients who require emergency laparotomy for homodynamic instability, peritonitis or for any other indication should have a thorough exploration and a careful assessment of the diaphragm. The entire diaphragmatic surface should be exposed and directly visualized. Preoperative diagnosis was made in 61% of our patients in keeping with what is reported in the literature which ranges between 40% to 93% (1,29, 30). The left hemi diaphragm is the most commonly injured side. In our series left sided injuries were seen in 73% of the cohorts. This is believed to be due to many factors that include the buffering effect of the liver(12) . The incidence of right sided injuries is increasingly reported in the literature and this is attributed to many factors that include the improvement in pre-hospital care, the better transportation and the increasing incidence of severe upper abdominal trauma with motor vehicle accidents (31). The position of the patient in the vehicle and the side of impact are believed to be correlated with the side of diaphragmatic rupture (11,32, 33). Kearney and colleagues (32) have noticed a higher incidence of rupture of the diaphragm with side impact compared with frontal impact. Interestingly, in our series, impacts from all sides have the same incidence. Drivers, however, are more commonly affected than other passengers.

Management:

           The traumatic injury should be repaired as soon as the diagnosis is made to avoid the potential long-term life threatening complications such as incarceration and strangulation. Because of the high incidence of associated intra-abdominal injuries that require surgical intervention the most common operative approach to most cases of TDR is through a midline abdominal incision (93%). This is in accordance with data reported by many other authors(1,4,26,31,34,35,36). The operative repair of the right diaphragmatic rupture is still under a tremendous controversy concerning the right approach (Thoracotomy vs. Laparotomy)(29). In our series, out of the fifteen right diaphragmatic rupture, one patient only had a repair via a thoracotomy incision. There are many different types of repairs reported in the literature(30). In our series all the patients were repaired in a simple fashion using either absorbable (most commonly Vicryl) or Non-absorbable material (Proline). None of our patients required the use of a prosthetic material. The choice of the type of repair and the material was dependant on the discretion of the surgeon.

Outcome:

Complications:


          The complication rate in our series was consistent with the literature where the majority are pulmonary and infectious(1,21,26,28,37,38).

Ventilatory support:

           It is believed that the diaphragm muscle will be paralyzed after trauma and the patients will require ventilation. We have found that ventilatory support was indicated by the associated injuries rather than the diaphragmatic rupture itself. Twelve patients (21%) in the total cohort did not require ventilation post operatively. This was very evident in patients with penetrating injuries (especially stab wound) who usually have less associated injuries. Seventy one percent of the stabs did not require ventilatory support. Seven patients (14%) who sustained blunt trauma did not require ventilatory support.

Length of Stay:

              The length of stay in this series was 29 +/- 22 days (34 for blunt and 17 for penetrating). This is mainly attributed to the severity of associated injuries that are seen more with blunt trauma rather than the diaphragmatic injury itself(38).

Mortality:

              Four patients (7%) died. All deaths were due to the associated injuries rather than the diaphragmatic injury itself. Diaphragmatic injuries rarely result in post traumatic death (33). Mortality ranges from 1-40.5% (21,26,33). This depends on the severity of associated injuries. We did not find the delay in treatment to increase the incidence of mortality. Of the fifteen patients who had a delay in the treatment (>24hr) only one died and the death was not related to the diaphragm itself rather to post operative myocardial infarction. This is supported by the literature (21). We performed a logistic regression analysis to determine predictors of mortality in this cohort and we found that age was the most significant predictor of mortality (P=0.04). ISS and the need for ventilation revealed a trend towards mortality, but were not statistically significant.

Conclusion:

TDR is uncommon, but requires a prompt diagnosis and repair. Because of the high incidence of associated injuries and the lack of specific clinical signs, a high index of suspicion is required to make the right diagnosis early in the course. This study revealed that older age is a major predictor of mortality whereas the need for ventilatory support did not predict the overall mortality.


 

References

1. Rodriguez-Morales G. Rodriguez A. Shatney CH. Acute rupture of the diaphragm in blunt trauma: analysis of 60 patients. Journal of Trauma-Injury Infection & Critical Care. 1986 May; 26 (5):438-444.

2. Cox EF. Blunt abdominal trauma. Blunt abdominal trauma. A 5-year analysis of 870 patients requiring celiotomy. Annals of Surgery. 1984 Apr; 199(4):467-474.

 3. Ward RE. Flynn TC. Clark WP. Diaphragmatic disruption secondary to blunt abdominal trauma. Journal of Trauma-Injury Infection & Critical Care. 1981 Jan.; 21(1):35-38.

4. Estrera AS. Platt MR. Mills LJ. Traumatic injuries of the diaphragm. Chest. 1979 Mar.; 75 (3):306-313.

5. Voeller GR. Reisser JR. Fabian TC. Kudsk K. Mangiante EC. Blunt diaphragm injuries. A five-year experience. American Surgeon. 1990 Jan.; 56(1):28-31.

 6. Reid J. Case of diaphragmatic hernia produced by a penetrating wound. Edinburgh Med. Surg. J. 1840; 53:104.

7. Bowditch H. I. Diaphragmatic hernia. Buffalo Med J. 1853; 9(1):65.

8. Hedbolm CA. Diaphragmatic hernia. J.A.M.A. 1925; 85:947-953.

 9. Hood RM. Traumatic diaphragmatic hernia. Ann. Thorac. Surg. 1971; 12:311-324.

10. Meyers BF. McCabe CJ. Traumatic diaphragmatic hernia. Occult marker of serious injury. Annals of Surgery. 1993 Dec.; 218(6):783-790.

11. Beal SL. McKennan M. Blunt diaphragm rupture. A morbid injury. Arch of Surg. 1988 Jul; 123(7):828-832.

12. Sharma OP. Traumatic diaphragmatic rupture: not an uncommon entity personal experience with collective review of the 1980's. Journal of Trauma-Injury Infection & Critical Care. 1989 May; 29 (5):678-682.

13. Wisner D.H. Abdominal injuries-Diaphragm. Surgery Scientific principle and practice; Greenfield textbook of surgery.2nd edition 1997.page 337.

14. Guth AA. Pachter HL. Kim U. Pitfalls in the diagnosis of blunt diaphragmatic injury. American Journal of Surgery. 1995 Jul; 170 (1):5-9.
 
15. Waldhausen JA, Kilman JW, Helman CH. The diagnosis and management of traumatic injuries of the diapheagm inclding the use of marlex Prosthesis. J Trauma. 1966; 6:332-343.

16. Waldschmidt ML, Laws HL. Injuries of the diaphragm. J. Trauma. 1980. 20:587-591.

 17. Brooks JW. Blunt traumatic rupture of the diaphragm. Annals of thoracic surgery. 1978 Sep; 26(3):199-203.

 18. Clay RC., Hanlon CR. Pneumoperitoneum in the differential diagnosis of diaphragmatic hernia. J Thorac Cardiovasc Surg. 1951; 21:57.

19. Gelman R. Mirvis SE. Gens D. Diaphragmatic rupture due to blunt trauma: sensitivity of plain chest radiographs. AJR. 1991 Jan.; 156(1):51-57.

 20. Ball T. McCrory R. Smith JO. Clements JL Jr. Traumatic diaphragmatic hernia: errors in diagnosis. American Journal of Roentgenology. 1982 Apr; 138(4):633-637.

21. Brasel KJ. Borgstrom DC. Meyer P. Weigelt JA. Predictors of outcome in blunt diaphragm rupture. Journal of Trauma-Injury Infection & Critical Care. 1996 Sep.; 41(3):484-487.

22. Freeman T. Fischer RP. The inadequacy of peritoneal lavage in diagnosing acute diaphragmatic rupture. Journal of Trauma-Injury Infection & Critical Care. 1976 Jul.;16(7):538-542

 23. Heiberg E. Wolverson MK. Hurd RN. Jagannadharao B. Sundaram M. CT recognition of traumatic rupture of the diaphragm. American Journal of Roentgenology. 1980 Aug; 135 (2):369-372.

 24. Federle MP, Goldberg HI, Kaiser JA, Moss AA, Jeffrey RB Jr, JC. M. Evaluation of abdominal trauma by computed tomography. Radiology. 1981 Mar; 138(3):637-644.

25. Kearney PA. Rouhana SW. Blunt rupture of the diaphragm: mechanism, diagnosis, and treatment. Annals of Emergency Medicine. 1989 Dec; 18(12):1326-1330.

26. Chen JC. Wilson SE. Diaphragmatic injuries: recognition and management in sixty-two patients. American Surgeon. 1991 Dec; 57(12):810-815.

27. Shapiro MJ, Heiberg E, Durham RM, Luchtefeld W, JE. M. The unreliability of CT scans and initial chest radiographs in evaluating blunt trauma induced diaphragmatic rupture. Clinical Radiology. 1996 Jan; 51(1):27-30.

28. Meyer G. Huttl TP, Hatz RA, Schildberg FW. Laparoscopic repair of traumatic diaphragmatic hernias. Surgical Endoscopy. 2000 Nov; 14(11):1010-1014.

 29. Morgan AS. Flancbaum L. Esposito T. Cox EF. C. Blunt injury to the diaphragm: an analysis of 44 patients. Journal of Trauma-Injury Infection & Critical Care. 1986 Jun; 26(6):565-568.

30. Pagliarello G. Carter J. Traumatic injury to the diaphragm: timely diagnosis and treatment. Journal of Trauma-Injury Infection & Critical Care. 1992 Aug; 33(2):194-197.

 31. Sukul DM. Kats E. Johannes EJ. Sixty-three cases of traumatic injury of the diaphragm. Injury. 1991 Jul.; 22(4):303-306.

32. Kearney PA, Rouhana SW, RE B. Blunt rupture of the diaphragm: mechanism, diagnosis, and treatment. Ann Emerg Med. 1989 Dec;18(12):1326-1330.

33. Boulanger BR. Milzman DP. Rosati C. Rodriguez A. A comparison of right and left blunt traumatic diaphragmatic rupture. Journal of Trauma-Injury Infection & Critical Care. 1993 Aug; 35(2):255-260.

34. Carter JW. Diaphragmatic trauma in southern Saskatchewan--an 11-year review. Journal of Trauma-Injury Infection & Critical Care. 1987 Sep;27(9):987-993.

35. Flancbaum L. Dauber M. Demas C. Boyarsky AH. Trooskin SZ. Early diagnosis and treatment of blunt diaphragmatic injury. American Surgeon. 1988 Apr; 54(4):195-199.

36. Van Vugt AB. Schoots FJ. Acute diaphragmatic rupture due to blunt trauma: a retrospective analysis. Journal of Trauma-Injury Infection & Critical Care. 1989 May.; 29(5):683-686.

 37. Reber PU, Schmied B, Seiler CA, Baer HU, Patel AG, MW B. Missed diaphragmatic injuries and their long-term sequelae. J Trauma. 1998; 44:183-188.

 38. Ilgenfritz FM. Stewart DE. Blunt trauma of the diaphragm: a 15-county, private hospital experience. American Surgeon. 1992 Jun; 58(6):334-339.

Other Topics:

Original Study # 1 -  Spectrum of Cutaneous Adverse Drug Reactions seen in the Emergency Department (ED): A Prospective Study from Kuwait
Original Study # 2 -  One Year-Study of Patients with Acute Organophosphate Insecticide Poisoning Admitted to the Intensive Care Unit of Hamad General Hospital, Doha, State of Qatar
Original Study # 4 -  Morphine Sparing Effect of Proparacetamol in Surgical and Trauma Intensive Care
Original Study # 5 -  Acute Mastitis and Breast Abscesses Among Lactating Women: Occurrence and Risk Factors