Post-colonoscopy tension pneumothorax resulting from colonic barotrauma in a previously unrecognised left-sided diaphragmatic hernia.

result in colonic perforation during colo− noscopy, they are always associated with pneumothorax and are always fatal [1, 2]. We report a case of colonoscopy−induced tension pneumothorax due to an undiag− nosed diaphragmatic hernia without co− lonic perforation. A 46−year−old man with no history of thoracoabdominal trauma underwent co− lonoscopy for rectal bleeding. Colono− scopic progress though initially straight− forward became difficult beyond the splenic flexure. The procedure was aban− doned due to abdominal discomfort. The patient was discharged after a satisfac− tory post−procedural examination. After 6 hours he presented with chest pain and breathlessness due to a left−sided tension pneumothorax (l" Fig. 1). Following de− compression and chest drain insertion, he was managed in the intensive therapy unit. The left lung failed to re−expand due to a left−sided diaphragmatic hernia (l" Fig. 2) that was later repaired by open surgery. The hernia contained an incar− cerated colonic loop that was resected with primary colonic anastomosis fol− lowed by closure of the hernial defect. Histological examination of the specimen (l" Fig. 3) suggested ischemic damage but not perforation. The lung re−expan− ded after 4 days and the patient was dis− charged shortly afterwards. Radiological evidence suggested the hernia was prob− ably of traumatic origin. Late diagnosis of traumatic hernias im− pedes their management, with unexpec− ted diagnoses after a latency period [3, 4] or at post−mortem [5]. The abdomen−to− thorax pressure gradient favors progres− sive herniation of viscera through undiag− nosed diaphragmatic defects. The lack of supporting evidence (radiological or his− tological) and the nonfatal outcome meant that it was unlikely colonic per− foration had occurred. The hernia con− tained a colonic loop, causing acute angu− lation and impeding the passage of the colonoscope. We are convinced that the pneumothorax resulted from the avulsion of adherent lung away from the incarcer− ated colonic segment (l" Fig. 4). Maneu− vering of the instrument and stretching of the colonic segment on air−insufflation caused avulsion of adherent colon, creat− ing a defect in the apposed lung surface along with rupture of associated alveoli. Good postprocedural advice, excellent emergency medical management and sound surgical repair ensured that the pa− tient had a good outcome.