Diaphragmatic injuries were first described in 1541 by Sennertus. This condition is not uncommon with rates as high as 5% for patients hospitalized after motor vehicle accidents, and 15% for patients after penetrating injuries to the lower chest and upper abdomen. 8–10 Almost half of these injuries develop into hernias 1.
A number of series showed more patients with traumatic diaphragmatic injury(TDI) suffered from penetrating trauma as compared with blunt trauma1,11 while other series have suggested that the majority of TDI, up to 80%, resulted from blunt trauma4,12−15. Lewis1 thought that these discrepancies in frequency are likely the result of variations in geographic and socioeconomic factors of the study populations. In our TDH cohort, blunt trauma was predominantly higher for we had selectively excluded the cases of TDI without herniation. So we speculate that since a lot of TDI were not able to be recognized promptly, especially for the penetrating injuries which may not present with herniation, these cases might have been missed in some previous studies. And it may also have added to the discrepancies in frequency of the mechanisms. The most common causes of TDH in our series (MVC, GSW, and SW) are consistent with previously reported data.4,11,14,16,17 MVC comprised 61.4% of all cases, which was by far the most common cause. All1, or at least a large proportion14, of penetrating injuries were due to stab and GSWs, which was consistent with our estudy. It is desirable that diaphragmatic injuries be diagnosed before the complications like diaphragmatic hernia and strangulation occur since mortality and morbidity increase after the herniation and strangulation of the abdominal viscera in the thoracic cavity. However, according to our experience, in a significant proportion of the cases diaphragmatic injuries were insidious with no visible signs on Chest X-ray, which was reported to be very useful and with sensitivities up to 94% in the presence of a hernia18. In the early period, we utilized conventional chest X-ray, combined with abdominal ultrasound and barium meal trying to achieve early diagnoses. Though all cases enrolled in the study presented with herniation, as reported previously19, some of them were not diagnosed in a timely fashion. The most common cause of delayed diagnosis was the right sided diaphragmatic rupture20. According to reports, only 17% of the cases of right sided rupture are diagnosed with chest X-ray.21 In our study, 6 (60%) cases of delayed diagnosis were of the right sided hernia. Another important reason probably is the films were interpreted by surgeons. Although they were able to identify all cases of diaphragmatic hernia, many cases of injury without an obvious hernia were misseded1,22,23. Furthermore, sometimes, the hernia might have mimiced an elevation of diaphragm especially in the right hemithorax with the liver herniating into the chest or it might have been misdiagnosed with hemothorax or encapsulated effusion. As mentioned above, there were 4 delayed diagnoses which were confirmed 4 months to 12 years later after discharge from hospital. Delayed or missed diagnoses may place patients at risk for morbidity and mortality. In some cases, it could be insidious for delayed hernia formation or small openings. In the recent period, we have been using CT scan (initially, 16 multi-slice, then 64 multi-slice) characterized by high space definition and multi-planer reconstruction ability to help diagnose and have achieved a minimal number of missed diaphragmatic injuries. CT has greatly enhanced the ability of early diagnosis and is reported to have a sensitivity of 71% (78% for the left and 50%for the right) and a specificity of 100% and an accuracy of 88% for the left and 70% for the right sided injuries.24 With CT scan, the early diagnosis rates were significantly higher than it had been before (p = .042) (Table II).
The incidence of associated injuries was 52.4%, and the most common associated injury was rib fracture followed by intra-abdominal injuries, which are similar to previous series 15,25. Some studies have suggested nearly 100% incidence of associated injuries, implicating a discrepancy from ours. This may be explained by the different enrollment criteria. In our study, we excluded all patient died before admission while some previous studies included such patients. Since associated injuries add to ISS scores leading to higher mortality, those who died before hospitalization probably had higher ratio of associated injuries.
The operative approach of choice is affected by whether there are associated injuries. Though it was reported that the incidence of associated additional intra-abdominal injuries was up to 100% of patients14,26−28, we chose a thoracotomy as our main approach(71.4%)(Table III) for an incidence of associated abdominal injuries of 25.4%(16/63) in our series. A posterolateral thoracotomy warrants a good exposure of diaphragm as well as part of the abdominal organs. We have found it easier to reduce the herniated contents and to repair the diaphragm through a thoracotomy when there are no intra-abdominal injuries29. Importantly, in cases of delayed presentation, thoracotomy is an accepted approach as it is difficult to release the intra thoracic adhesions through a laparotomy.30 Many authors choose laparotomy or laparoscopy as the route of choice4–6, 31. A Laparotomy may be used in patients presenting with abdominal symptom or physical signs and will be more frequently used in our future practice as more evidence coming up in favor of a laparotomy. In patients with severe injury of the junction between thorax and abdomen, a thoracoabdominal incision may be the approach of choice. Another important factor impacting surgeon’s choice of approach is the familiarity and comfort level for a surgeon 32, which is also the case in this study. Physicians should keep in mind that injuries might happen in both sides. A case in our cohort presented with shortness of breath and low saturation of blood oxygen after surgical repair of left diaphragm injury. Further investigations suggested right diaphragmatic rupture and hepatic herniation into the chest. The patient had to undergo another right thoracotomy and recovered without major morbidity.
The mortality in our study (11.1%) is much lower than other published series11,17,33. Several reasons may have contributed to it. First, our study only enrolled the patients who had survived to hospital admission, yet many had died before admission. Second, the referrals from other hospitals were sometimes limited to relatively less severely injured patients, so a lot of severely or critically injured patient were not included. Third, our study focused specifically on TDH, which is only a subset of TDI. However, we found risk factors for death similar to published studies 11,28,34. Older age (p = .017), higher ISS (p = .026) and AAST grade (p = .014) by univariate analysis were associated with nonsurvival. Multiple logistic regression analysis suggested that advanced age (p = .013) and higher ISS (p = .028) were predictors of death.
Table I shows older mean age, higher ISS score and AAST grade in blunt group than in penetrating group. The age difference implicates a socioeconomic impact on the mechanisms of injury. According to the mechanisms, more associated injuries, which apparently add to ISS score, might be found in blunt injuries. But same as previous report1, no difference in mortality was found between the two mechanisms. Since the volume of cohort is small, large-scale study is warranted.