We agree with the letter from Scala and Esquinas 1 in response to the article by Schortgen and colleagues 2, who emphasised the use of non-invasive ventilation in the ICU as the best ventilatory treatment for 'do not intubate' octogenarian patients. Scala and Esquinas argued that ICU beds are scarce and that the ICU environment alters contact between the patient and family.

We do not, however, entirely accept the views of Schortgen and colleagues. Use of the non-invasive ventilation mask for palliative care patients with acute respiratory distress prevents the patient from eating and talking, and the patient's experience can be that of being smothered. When the major indication is hypoxemia, a treatment option is the administration of high-flow oxygen using up to 60 l/minute heated and humidified oxygen through a nasal cannula 3. The mouth is thus freed and the patient is able to eat and talk with his family. The cost to efficiency ratio is favourable because the Optiflow^® oxygenation system (Fischer and Paykel™, Auckland, New Zealand) costs €4,000 versus €15,600 for the V60^® ventilation system (Philips™, Amsterdam, New Netherlands). Non-invasive ventilation appears preferable in chronic obstructive pulmonary disease patients with hypercapnia.

We tested high-flow oxygen administration in 10 'do not intubate' patients receiving palliative cancer care in whom a high oxygen concentration mask failed to relieve dyspnoea (abstract accepted for the Société Française d'Anesthésie Réanimation National Congress, September 2012). The mean respiratory rate fell from 29 to 22 breaths/minute (P <0.01) and the ability to communicate (visual analogue scale) increased from 2 to 7 (P <0.01).

To conclude, high-flow oxygen is an easy and cheap way to relieve dyspnoea in 'do not intubate' patients.

Competing interests