Using the Bruel & Kjaer Anaesthetic Gas Monitor type 1304, we have mon
itored the output of 94 anaesthetic agent vaporizers (Fluotec 3:58, En
fluratec 3:24, Isotec 3:12), in seven departments of anaesthesia, at d
ifferent dial settings and flow rates. The range of output, for one ty
pe of vaporizer and dial setting (flow: 6 litre min-1) was largest wit
h the Fluotec 3 (0.85-1.55% when dial set to 1%) and smallest with the
Isotec 3 (0.85-1.15% when dial set to 1%). In determining the number
of vaporizers with unacceptable inaccuracy, we applied acceptance limi
ts of +/- 15% relative on each vaporizer and each dial setting. Using
a flow of oxygen 6 litre min-1 17% of Fluotec 3,8% of Isotec 3 and 71%
of Enfluratec 3 vaporizers had outputs outside those limits. Even whe
n some specific conditions (vaporizers giving output beyond the limits
at any two or more dial settings; output beyond the limits in the cli
nically relevant range (0.5-2%)) were added, a substantial number of v
aporizers did not perform within the limits. We found a significantly
greater accuracy of the vaporizers after 3-monthly calibration, checks
(P < 0.05) compared with vaporizers undergoing service and calibratio
n only annually. Using a questionnaire, we found that fewer than 30% o
f the anaesthetists using the vaporizers would accept aberrance beyond
+/- 10% relative of the dial setting.