To the Editor:—

The Draeger Narkomed 6000 anesthesia delivery system (Telford, Pennsylvania) incorporates self-testing features that are designed to automate the checkout process. However, while evaluating the system (S/N 10038, software version M1/7.0) for use in our hospital, I discovered three significant deficiencies, which pose a risk to patient safety.

First, in an emergency situation, with the machine turned off, the flush valve delivers high-flow oxygen, provided there is a source of oxygen (pipeline or tank) available. However, in contrast to other machines, positive-pressure ventilation cannot be administered via  the breathing circuit because the ventilator’s built-in pressure relief valve is open under these circumstances.

Second, immediately after the system is turned on, it remains impossible to deliver positive-pressure ventilation. After filling the circuit from the flush valve or flow meters, the circuit does not hold pressure, even if the pop-off valve is completely closed. Positive-pressure ventilation can only be accomplished after the self-test procedure is completed (requiring 4 min), the self-test procedure is interrupted by pressing the standby button (requires about 33 s), or the red, emergency ventilator bypass button is pressed (requires about 13 s). In the third instance, although manual ventilation is possible, the machine must be completely reset and tested before the ventilator can be used.

Third, if the machine is turned off briefly while running on battery power (which might occur accidentally, or perhaps intentionally if it is necessary to reset the computer) and then restarted, the computerized electronics may fail: The display first indicates “Please wait while system writes unsaved data to disk.” This is followed by the message “It is now safe to turn off your computer,” accompanied by a small box on the screen indicating “Restart.” Touching this box causes the computer to restart, but shortly thereafter, the machine electronics abruptly power down: The screen goes dark, the fans go silent, and even the flowmeter lights switch off. Furthermore, during the abortive start-up process, there is no indication of the AC power failure. Despite this electronic failure, the flowmeters continue to operate properly because the pneumatic switch remains in the “on” position. However, manual positive-pressure ventilation may not be possible, depending on the internal state of the ventilator pressure relief valve.

I believe that these characteristics pose a significant risk to patient safety. If an electrical or electronic failure occurs, the ventilator’s internal pop-off mechanism cannot be bypassed, and it is impossible to deliver positive-pressure ventilation manually with the rebreathing bag. Although internal battery backup power provides a measure of protection, a failure in the internal power supply circuitry could result in inability to provide positive-pressure ventilation. A mechanical switch-over device, similar to those used in previous anesthesia delivery systems, would reduce or eliminate the risk of these problems.