Prone, head down for pulmonary haemorrhage

Abstract

Editor'I read with interest the article by HÅkanson and colleagues1Håkanson E Konstantinov IE Fransson S-G Svedjeholm R Management of life-threatening haemoptysis.Br J Anaesth. 2002; 88: 291-295Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar describing the successful management of life-threatening pulmonary haemorrhage. I have been involved with an unexpected case occurring during elective general anaesthesia in a young adult with cystic fibrosis.2Savage R Prone position as a life saving measure for acute pulmonary haemorrhage in a young adult with cystic fibrosis.Anaesth Intensive Care. 2002; 30: 223-225PubMed Google Scholar I agree that death is more likely to be due to hypoxia than exsanguination. In my case, the patient became severely hypoxic (PaO2 8.1 kPa, FiO2 1.0) and hypercarbic (PaCO2 28.2 kPa) with a severe respiratory acidosis (pH 6.85) despite peak ventilatory pressures of 40 cm H2O, PEEP of 12 cm H2O and inverse ratio ventilation. All simple manoeuvres such as endotracheal adrenaline, suction via the endotracheal tube and fibreoptic bronchoscopy were ineffective. The patient was considered too unstable for lung isolation using a double lumen endotracheal tube. However, on turning the patient prone, gas exchange improved dramatically and immediately. There are many reasons for such a dramatic improvement. Blood in the airway both acts as a barrier to gas diffusion at an alveolar level as well as physically blocking larger airways, and results in significant hypoventilation with resultant shunt. Due to gravity, the haemorrhage is likely to affect mainly the posterior part of the lung, which is also the part of the lung that is preferentially perfused, worsening V˙/ Q˙ mismatch. A severe respiratory acidosis may result from decreased alveolar ventilation. The anatomy of the airways is such that drainage of blood and secretions from the lung are encouraged in the prone position and airways open up. I disagree with turning the patient onto the side of the bleeding lung because this results in one lung not contributing to gas exchange when this may not be necessary. It worsens V˙/ Q˙ mismatch as the lower lung will be preferentially perfused but not ventilated. These patients can then be extremely difficult to oxygenate and ventilate. Rigid bronchoscopy to aspirate blood clots requires an experienced operator and may be hazardous if the patient is very hypoxic. The flexible fibreoptic bronchoscope may not be able to deliver the suction required to remove large blood clots sufficiently quickly. My patient was so hypoxic that he would have been unable to tolerate such suction for more than a few seconds. The source of bleeding was identified using angiography with subsequent successful management by bronchial artery embolization. Editor'Thank you for giving us the opportunity to respond to Dr Savage's letter. We are impressed by the successful management of his patient with cystic fibrosis and life-threatening haemoptysis originating from a bronchial artery malformation. The respiratory problem in our patient with bleeding from a bronchial artery malformation was severe hypoxia (Po2 approximately 5 kPa at FiO2 1.0), without serious hypercapnia or acidosis, whereas Dr Savage's patient had moderate hypoxia with extreme hypercapnia and acidosis. The discrepancy in clinical presentation could be due to the fact that Dr Savage's patient had serious pre-existing pulmonary disease, which is likely to have influenced the impact of lung bleeding. Turning the patient prone can therefore be useful not only to improve oxygenation but also ventilation, as was demonstrated in his case report. In our paper it was suggested that it could be advisable to place the patient with the bleeding side down to prevent aspiration into the uninvolved lung. None of our patients were treated in this way. This statement was a reference to an acknowledged recommendation from the literature rather than a scientifically challenged hypothesis. Dr Savage's point that this position may in fact worsen the V˙/ Q˙ mismatch as compared with the prone position, is well taken. However, in these situations it should be appreciated that other factors than V˙/ Q˙ mismatch may also influence gas exchange. Life-threatening haemoptysis is fortunately very rare and not many doctors have experienced it in more than a few patients. We are grateful to Dr Savage for a contribution that draws attention to another therapeutic option in the treatment of these challenging cases. E. Håkanson I. Konstantinov S.-G. Fransson R. Svedjeholm Linköping, Sweden