Novel Breathing Circuit Architecture: New Consequences of Old Problems

To the Editor:—

New anesthesia machines incorporate novel designs of familiar components to provide additional functionality. The current literature and the U.S. Food and Drug Administration’s MAUDE database *contain periodic reports of difficulties in operating this modern equipment. 1,2The onus is on anesthesiologists to remain vigilant to possible new problems with this equipment and to new manifestations of old problems. 3We recently experienced a condition in which the design of the Drager Fabius GS (Draeger Medical Inc., Telford, PA) anesthesia machine’s breathing circuit exposed a loss of redundancy that was heretofore present in anesthesia machines.

After induction of anesthesia, mechanical ventilation and 2% desflurane in nitrous oxide/oxygen were provided with a Drager Fabius GS anesthesia machine. Set gas flows were 2 l/min of nitrous oxide and 1 l/min of oxygen. The anesthesia machine had been completely checked according to the manufacturer’s directions immediately before the case. This included leak tests of the circuit, breathing bag, and ventilator, all of which showed satisfactory results. Immediately after induction, we realized that the patient was latex-allergic and replaced the breathing bag with a nonlatex bag. A few minutes later, we noted that the breathing bag was empty. This is never a normal condition with the Fabius GS, as this machine has a piston ventilator and uses the breathing bag as the gas reservoir during mechanical ventilation. 3A scan of the monitors revealed that the circuit gas sampled by the gas analyzer at the Y piece contained 23% oxygen, 52% nitrous oxide, and 1.3% desflurane, all lower than the machine was set to deliver (fig. 1). The oxygen sensor on the machine also showed that less oxygen was being delivered than the machine was set for. We inferred that the machine was entraining room air, as it is designed to do under conditions of insufficient fresh gas flow, but this appeared to be inconsistent with the 3-l flows set on the machine, and we immediately suspected a breathing circuit leak. We disconnected the patient from the anesthesia machine, switched to manual ventilation mode, closed the adjustable pressure-limiting valve, and attempted to pressurize the circuit. This revealed a large hole hidden in the folds of the nonlatex breathing bag. The breathing bag was quickly replaced, the circuit was successfully pressure-tested, and mechanical ventilation and delivery of anesthetics were resumed. Thereafter, gas monitor values were consistent with machine settings. Afterwards, the patient denied recall of intraoperative events.

This is not the first report of a sudden inability to pressurize the breathing circuit of the Fabius GS. In 2002, several reports were made to the Food and Drug Administration of an inability to pressurize the breathing circuit, sometimes after successfully completing a leak test. This problem was traced by Drager’s engineers to a small change in a proven design, and the problem has been definitively addressed.

The Drager Fabius GS uses a piston-driven ventilator, and the piston is not the breathing circuit gas reservoir. Instead, the breathing bag serves as the reservoir during both mechanical and manual ventilation. Breathing bag failure compromised the function of our anesthesia machine in both ventilation modes: the inability to pressurize the circuit in manual mode and decreased oxygen and anesthetic concentrations in mechanical mode.

Our failure was not so catastrophic that the patient could not be ventilated (at least mechanically), and it had nothing to do with the anesthesia machine itself. Adequate mechanical ventilation continued throughout the event. However, the anesthesia machine entrained room air presumably freely through the hole in the breathing bag. This path of air entrainment explains another important feature of this event: no alarms sounded to alert clinicians of the failure.

The Fabius GS has a “low fresh gas flow” alarm to alert users that insufficient gas is entering the circuit to make up for removal due to uptake and leaks. With inadequate fresh gas flow, the piston empties the breathing bag and then room air is entrained via  a one-way valve into the piston chamber so that the ventilator delivers the set tidal volume. A small negative pressure is generated when the ventilator attempts to draw from an empty breathing bag, triggering the alarm. In our case, the circuit was always in continuity with the atmosphere because of the large hole in the breathing bag, no negative pressure was developed, and no alarms sounded, but room air was entrained. Instead, we detected and corrected a massive circuit leak (induced after machine checkout) by observation of a constellation of inconsistencies. Despite improved alarm systems, previously secondary monitors such as agent analyzers are sometimes the only alerts to impending failures (in this case, awareness). 4 

Relative to older designs, subtle changes in the external operation of the Fabius GS conceal radical internal changes that both add functionality and completely alter its response to failure conditions. The use of a piston ventilator with room air entrainment as a backup gas source is safe because it ensures adequate ventilation. However, this design could lead to awareness 5without the use of agent analysis or some measure of anesthetic depth, and is also of concern when considering the use of the machine to care for patients dependent on high inspired oxygen concentrations.