I would like to thank Alistair Steel and James Varley for their interest in my paper and for their extremely useful comments.1 2 I would agree with them that infusions should be purged before connection—something which I suspect is commonly done—and that higher doses of glyceryl trinitrate should be used, a point which I make in my paper. I also agree with them that a bolus should be given in critically unwell patients while the infusion is being set up; however, I suspect this is not done routinely. For this reason, I still believe that it is very useful to make practitioners aware of the extreme differences in dead space depending on the cannula size and initial infusion rate used. I hope that, after the considerable interest the article has generated, an initial bolus, a higher initial infusion rate and a move away from large diameter cannulae will become common practice.