How not to do it

Abstract

We have read the paper by Dr Elsayed about a life-threatening haemothorax complicating a mediastinoscopy [1]. We think many points must be discussed. First, the author states that they have closed a patient submitted to mediastinoscopy with demonstrated active (even if very mild) mediastinal bleeding associated with iatrogenic opening of the parietal pleura. This is not cautious behaviour by a thoracic surgeon. It is well known that even mild bleeding may be the cause of massive haemothorax when the negative intrapleural pressure is transmitted to other opened neighbouring compartments, such as the mediastinum. Second, when postoperative right-sided haemothorax was demonstrated by a reduction of the haemoglobin levels and chest X-ray (Fig. 2a of the cited paper [1]) they inserted – as stated – a basal chest drain without significant improvement. The chest X-ray demonstrated the presence of right-sided thoracic clotting in the upper and medial third of the chest without involvement of the basal region. Unfortunately, the patient was well-known preoperatively to have a raised right hemi-diaphragm, explained by a history of right lower lobe pneumonia with postpneumonic scarring and traction on the diaphragm. These are all indirect signs of the presence of basal pleural adhesions, confirmed by the presence on the chest X-ray of a bloody effusion without basal involvement. The basal chest drain was very hazardous in this case, with inherent risk of further damage. After a 24 h observation period, video-assisted thoracoscopic surgery was attempted, with demonstration of many pleural adhesions (obviously), so the video-assisted procedure was converted into right thoracotomy with evacuation of 1800 ml of clotted blood. Why, on the basis of the previous chest X-ray, was a limited axillary right thoracotomy not performed at the third intercostal space with easy pleural access and eventual control of bleeding at the well-exposed upper mediastinum? Third, but not last, during the procedure to control the iatrogenic bleeding secondary to mediastinal lymph nodal biopsy, the surgeons performed a mediastinal lymphadenectomy and attempted a decortication of the right inferior lobe. These procedures are debatable, because they are not directly related to the primary indication of the reintervenction and because they may significantly increase the operative risks. Moreover, the chest X-ray at discharge revealed that decortication resulted in no improvements. Our personal opinion is that an intervention of pleural access and haemostasis via the axillary route should have been attempted first without basal pleural drain insertion, and no other collateral procedures should have been carried out because the relative risks were not justified, in particular when the diagnosis of the primary disease was still unknown (lymph nodal enlargement due to sarcoidosis). The reported paper concerns a rare, life-threatening complication of mediastinoscopy that has been managed in a dangerous manner. We think that the procedures described are the perfect example of ‘how not to do it’.