Catastrophic Failure of Aestiva 3000 Absorber Manifold

To the Editor:—

We wish to report a potentially catastrophic failure mode of the absorber module of the Aestiva 3000 anesthesia machine (Datex-Ohmeda, Madison, WI). During anesthesia for removal of a baclofen pump in a 17-yr-old girl, the surgeon requested a Valsalva maneuver to confirm adequate dural seal. On switching from ventilator mode to bag mode, it was not possible to generate positive pressure in the breathing circuit; during efforts to troubleshoot the problem, the patient’s depth of anesthesia lightened to the point that she aroused and self-extubated. Following reintubation, the patient was successfully ventilated using the ventilator. At the conclusion of surgery, the patient was allowed to breathe spontaneously from a Mapleson D (Vital Signs, Totowa, NJ) breathing circuit and auxiliary oxygen source, as it remained impossible to generate positive pressure with the anesthesia machine in bag mode.

Examination of the machine immediately after the episode revealed a massive leak in the bag mode; activation of the oxygen flush with the bag-ventilator switch set to bag mode resulted in expansion of the ventilator bellows. The machine was pulled from service and the absorber assembly was disassembled and inspected, revealing a crack in the bracket holding the shaft of the bag-ventilator switch toggle actuator (fig. 1). A review of the design of the Aestiva absorber shows how this defect produced the failure.

In the Aestiva absorber, the bag-ventilator switch uses a rocker-arm toggle actuator (fig. 1) to close either the ventilator-exclusion valve (in bag mode) or the bag-exclusion valve (in ventilator mode). In the event that both valves are open, compression of the bag results in diversion of gas to the ventilator assembly, where it is vented through the open ventilator pop-off valve to the scavenger system, resulting in a massive leak. Activation of the ventilator with both exclusion valves open results in diversion of the inspired tidal volume to the breathing bag and/or the scavenger system (depending on the pressure limit set on the pop-off valve). The exclusion valves are plunger-type valves; to seal, a pliable O-ring must make contact with the lip of the valve orifice. The crack in the support bracket allowed enough play in the rocker-arm actuator that the ventilator-exclusion valve did not seal when the selector was set to bag mode (fig. 2).

We have also experienced recurrent episodes in which it has been temporarily impossible to deliver positive pressure with the breathing bag on the Aestiva anesthesia machine; these events have typically occurred after switching from ventilator mode to bag mode and have been consistent with a massive circuit leak. All affected machines were reported to have had minimal leak during the anesthesia machine checkout procedure. During at least one of these episodes, activation of the oxygen flush with the bag-ventilator switch in the bag mode resulted in expansion of the ventilator bellows along with the breathing bag, as in the event described above. These episodes have resolved spontaneously during troubleshooting, and despite extensive examination by anesthesiologists, hospital biomedical engineers, and company representatives, no physical defects or mechanical problems were ever identified. We believe that these sporadic, self-correcting events occurred when minor alterations in the handling or function of the bag-ventilator switch permitted the ventilator-exclusion valve to fail to seal; manipulation of the switch during troubleshooting then corrected the problem.

In an effort to confirm that failure of the ventilator-exclusion valve would result in the observed failure, we attempted to reproduce the failure mode by modifying the ventilator-exclusion valve. Removal of the ventilator-exclusion valve O-ring resulted in the ability to generate, but not sustain, positive pressure in the breathing bag. Removal of the entire ventilator-exclusion plunger assembly reproduced the failure mode as outlined above (inability to generate positive pressure with the bag, filling of the ventilator bellows with the oxygen flush when in bag mode, and proper function of the ventilator). It is interesting that it remained possible to generate positive pressure using the ventilator despite the laxity introduced into the switch actuator by the cracked bracket; examination of the valve fiber-optically (as in fig. 2) revealed that the bag-exclusion valve did indeed seal (not shown). This observation leads us to speculate that there is some difference in the operation of the bag- and ventilator-exclusion valves. This speculation is supported by the fact that all of the sporadic, self-correcting episodes mentioned above have involved the inability to generate positive pressure using the breathing bag and never the converse (i.e. , inability to generate positive pressure with the ventilator).

In conclusion, we have presented a failure mode of the Datex-Ohmeda Aestiva absorber assembly that has presented in both a catastrophic mode and in a sporadic, self-correcting mode. We speculate that both failure modes can be attributed to failure of the ventilator-exclusion valve to seal when the bag-ventilator selector switch is set to bag mode. Anesthesiologists experiencing loss of circuit integrity in an Aestiva anesthesia machine when switching from ventilator mode to bag mode should respond initially by cycling several times between bag and ventilator mode, which should correct the problem. If repeated cycling of the selector switch does not resolve the situation, then the patient should either be returned to the ventilator or be ventilated using an auxiliary oxygen source and breathing circuit. The possibility of unexpected loss of circuit integrity despite an adequate preanesthesia checkout demonstrates again the importance of having an auxiliary oxygen source and breathing circuit during administration of each anesthetic.