We have sought to define a way in which nitrous oxide can be safely an
d universally used at minimal to low flows by utilising a circle syste
m with a controlled leak provided by a standard gas analyser sampling
line and a fresh gas supply of 50% nitrous oxide in oxygen, entering f
rom a trunk interposed between the ventilator and the circle system. A
lthough preliminary calculations suggested that this arrangement was l
ikely to work, it was found that 13 of 23 patients studied prospective
ly developed an inspired oxygen fraction below 0.3. We conclude that,
although this arrangement provides a new means of introducing nitrous
oxide into the circle breathing system, it does not appear inherently
safer or more convenient than the conventional route.