The Oxygen Analyser: Applications and Limitations – An Analysis of 2000 Incident Reports

Abstract

The first 2000 incidents reported to the Australian Incident Monitoring Study were analysed with respect to the role of the oxygen analyser; 27 (1%) were first detected by the oxygen analyser. All of these were amongst the 1256 incidents which occurred in association with general anaesthesia, of which 48% were “human detected” and 52% “monitor detected”. The oxygen analyser was ranked 7th and detected 4% of these monitor detected incidents. This figure would have been much higher had the oxygen analyser been correctly used on more occasions. The oxygen analyser detected 10 ventilator-driving-gas leaks into the circuit, 6 hypoxic mixtures due to rotameter settings, 3 inappropriate nitrous oxide concentrations, 2 disconnections and 1 leak at the common gas outlet, and 2 partial and 1 total failure of ventilation. In a theoretical analysis of these 1256 incidents it was considered that the oxygen analyser, used on its own, would have detected 114 (9%), had they been allowed to evolve (3% before any potential for organ damage). In 4 incidents an oxygen analyser gave faulty readings, in 3 caused a leak and in one a total circuit obstruction; 5 incidents were not detected because the alarm had been disabled. Despite the advent of piped gas supplies, failure of gas delivery or delivery of a “wrong” gas mixture still occurs surprisingly frequently in current anaesthetic practice; hypoxic mixtures were supplied on 16 occasions, other “wrong” mixtures on 23 and the oxygen supply failed on 7 occasions. Failure to use and improper use of the oxygen analyser is also surprisingly common; this is mainly due to a rule-based error encouraged by the poor design of the high alarm. It is highly recommended that a suitable, correctly sited, calibrated, tested oxygen analyser be used from before pre-oxygenation until the patient is no longer breathing gas from the anaesthetic machine or circuit.