Article Text

Download PDFPDF

Stabbing chest pain: a case of intermittent diaphragmatic herniation
  1. P Ransom,
  2. P Cornelius
  1. Emergency Department, Hastings Conquest Hospital, East Sussex, UK
  1. Correspondence to:
 Mr P Ransom
 Specialist Register, Accident and Emergency, 24 Friars Walk, Lewes BN7 2LF, UK;


A case of traumatic diaphragmatic herniation is described in which gross mediastinal shift was caused by a combination of the herniated abdominal organs, purulent exudate, and air. This complex presentation might best be described as a case of gastrocolopyopneumothorax, diagnosis of which was further complicated by the intermittent nature of the herniation.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

A case of traumatic diaphragmatic herniation is described in which gross mediastinal shift was caused by a combination of the herniated abdominal organs, purulent exudate, and air.


A 21 year old man presented to our emergency department complaining of a 2 day worsening of anterior and posterior chest pain with shortness of breath, and a 2 month history of intermittent vomiting, dysphagia, and weight loss. Three months previously, he had sustained a stab wound to the left chest and presented to hospital. A chest radiograph taken on that admission was reported as normal, and an abdominal ultrasound had shown only a small left pleural effusion. Shortly after, he had self discharged against medical advice. One month later he began the first of six visits to two separate emergency departments complaining of episodic abdominal pain and vomiting, leading to admission on two occasions. Blood results for these visits showed white cell counts of between 16 and 23×109/l, and varying degrees of dehydration, on one occasion leading to acute pre-renal failure with urea of 33 mmol/l. During the later visits, confusion had been noted. A chest radiograph taken 3 weeks prior to his presentation to us had been reported as normal by the radiologist and treating doctors, and oesophagogastroduodenoscopy (OGD) performed during that admission found a normal stomach, pylorus, and duodenum, with a degree of oesophageal candidiasis. This, together with his emaciated appearance, had led to a suspicion of immunocompromise. HIV testing was subsequently performed, which had proved to be negative. In the light of the normal chest radiograph and unremarkable gastroduodenal examination, a diagnosis of self induced vomiting had been given on his last discharge.

On examination at our hospital, he was emaciated and distressed, tachypnoeic at 30 breaths/min, with pulse oximetry on air of 97%. He was normotensive at 110/70, with a pulse of 110 beats/min. He was vomiting an almost clear fluid. Examination of the chest showed a trachea deviated to the right, absent breath sounds on the left side, with percussion note being resonant over the left upper chest and dull from the midline down. A small, healed scar was seen below the anterior axillary line of the 10th rib.

The initial portable chest radiograph showed massive diaphragmatic herniation of stomach and bowel, with a left sided pneumothorax and what appeared to be a collection of fluid reaching to the mid-thorax, all of which had caused marked rightward mediastinal shift (fig 1A).

Figure 1

 Chest radiographs taken before (A) and after (B) insertion of drain.

As the patient’s oxygen saturation was normal and the radiographic appearances were not those of a classical tension pneumothorax, we elected to place a small (size 12) chest drain anteriorly through the second intercostal space rather than perform a needle thoracentesis. A hiss of air was noted upon drain insertion, and during the patient’s initial breaths, continuous bubbling of the drain bottle was seen throughout the respiratory cycle. Within minutes, his chest pain had decreased. Over the next 15 minutes, 1.4 litres of thick, purulent fluid was discharged. A second radiograph taken 40 minutes after the first showed the stomach partially decompressed, some remaining fluid, and the mediastinum shifted back towards the midline (fig 1B).

Thoracotomy revealed that the greater part of the stomach and a loop of colon had herniated through a 6 cm defect of the left diaphragm. There was still an appreciable amount of purulent exudate, but no perforation of the abdominal organs and no damage to the lung structures. A sample of the fluid showed white cell debris, and culture revealed scanty growth of upper respiratory tract flora. He made an uneventful recovery, and once again self discharged against medical advice.


Herniation though a diaphragmatic stab wound is a well recognised, although infrequent cause of cardiothoracic embarrassment,1 while tension gastrothorax or colothorax appears in the literature occasionally as a case report.2 Intrapleural effusion caused by herniation of the abdominal contents has likewise been noted on occasion.3,4 Radiological diagnosis of diaphragmatic rupture is difficult, and many case series have demonstrated the unreliability of radiographs or computed tomography scans to pick up this injury.5

This case is unusual, however, in that the abdominal herniation, exudate, and the air under tension had all combined to cause respiratory compromise in what would best be called a tension gastropyopneumothorax. To our knowledge, no previous cases have been described in which these three features have co-existed simultaneously.

The contribution of the air to the tension is of interest. Classically, a tension pneumothorax is caused by a one way valve in a breached lung, allowing air into, but not out of, the pleural cavity. However, the lung in this case had not been breached, but had collapsed. Here, presumably, the combined volume of the bowel, the inflated stomach, and the exudate had together exerted an increasing pressure in the left chest that was only partially compensated by the rightward shift of the mediastinum. The hissing of air as the drain was inserted, and the initial bubbling of the chest drain independently of the respiratory cycle provide evidence for the pleural cavity being under a degree of air tension.

Perhaps the most diagnostically challenging feature of this case concerns the unremarkable chest radiograph and OGD performed 3 weeks before presentation to our hospital. The normal results of these investigations argue strongly that the herniation was intermittent in nature, as it is extremely unlikely that another pathology existed for his symptoms. Presumably the defect in the diaphragm had on previous occasions allowed the herniated abdominal contents to reduce spontaneously, until he presented to our hospital and reduction was this time no longer possible.

Stabbings are becoming a more frequent presentation to the emergency department.6 Patients with thoracoabdominal stab wounds or blunt abdominal trauma should perhaps be warned before discharge about the possibility of diaphragmatic herniation in the future, as the symptoms may be delayed by many years and the originating trauma forgotten or disregarded as the cause of subsequent, potentially life threatening, symptoms.



  • Competing interests: none declared