A hidden gas sampling line fault

An 8‐year‐old boy with a past medical history of scoliosis presented for elective percutaneous growth rod distraction. General anaesthesia was induced uneventfully and positive pressure ventilation commenced with an anaesthetic machine ventilator (Dräger Medical GmbH, Lübeck, Germany). The capnography waveform displayed an abnormal phase three, but immediately returned to baseline at the beginning of the phase. There was also an overall low end‐tidal carbon dioxide concentration (ETCO₂) of around 2.5 kPa. A large difference in measured inspiratory and expiratory oxygen and nitrous oxide concentrations of 35 % and 50 %, respectively, was also noted. There was an audible air leak near the Y‐piece of the breathing circuit, thus the circuit was changed. However, there was no improvement in the capnography waveform and there was a persistent leak alarm on the anaesthetic machine. We checked and eliminated a blocked water trap or a loose connection as the cause. Finally, we decided to replace the gas sampling line (Dragerwerk AG & Co. KGaA, Lubek, Germany) with a new one. The capnography waveform and ETCO₂ concentration gradually returned to normal. On careful inspection of the replaced gas sampling line, we found the male‐to‐male luer lock connector at both ends to be defective. The inner small tubular prolongation was missing at both ends (Fig. 1).

Figure 1.

Defective male luer lock connectors with absence of tubular prolongation.

Numerous causes have been previously reported for erroneous ETCO₂ concentrations, such as a blocked or damaged gas sampling line or low sampling rates with a long sampling line. This can be easily corrected by increasing the sampling rate or purging of the line and recalibration [1, 2]. An atypical tails‐up capnography trace is commonly observed when there is a leak between the gas sampling line and the gas analyser [1]. We did not observe such a trace. This was likely due to there being a significant breach in the gas sampling line at the patient end rather than the machine end. We believe entrainment of atmospheric air into the line through the faulty luer lock connector at the patient end during expiration led to dilution of gases and thus the abnormalities we observed.

This case highlights the importance of examining equipment thoroughly when monitoring abnormalities and alarms are observed. Although the luer lock connection at both ends of our gas sample line appeared to be secure, a structural defect led to erroneous expired gas measurements and an audible air leak. The manufacturer has been contacted regarding the fault.