To the Editor:—

Our department recently installed new North American Draeger Narkomed 6000 anesthesia machines (Telford, PA). These are microprocessor controlled and software driven and use an internal flow-dependent, piston-driven ventilator instead of a bellows. Thorough in-service education was done before use. Within a month, we experienced an unusual but significant problem with the ventilator. Neither the cause nor the solution was obvious or intuitive, thus prompting this letter.

In a patient requiring bronchial blockade for one-lung ventilation, we planned bronchoscopy using an Olympus LF-2 fiberoptic bronchoscope (Olympus, Lake Success, NY) and a Portex swivel adapter (Concord/Portex, Keene, NH). While maintaining mechanical ventilation, secretions were suctioned from the trachea via  the swivel adapter using a 14-French Kendall-Curity suction catheter (Kendall-Curity, Mansfield, MA). After bronchoscope insertion, the Apnea-Low Pressure alarm sounded in response to the deliberate leak. However, it became apparent by observation of the patient and machine that there was no effective ventilation occurring. The reservoir bag was grossly distended and would not empty, the display panel read “resetting piston,” the control switches were unresponsive, and we were unable to convert to manual ventilation. We disconnected the circuit and finished the procedure while maintaining ventilation with an Ambu bag. We then opened the locking lever under the ventilator cover, pulled up the ventilator and piston components, and reseated them. The piston reset itself, and the ventilator function resumed according to the original settings. The procedure continued uneventfully.

North American Draeger technical support was consulted, the problem was successfully recreated, and the cause was defined: The use of suction in the airway during mechanical ventilation generates negative-pressure flows of approximately 30 l/min, causing the piston to empty and lock, shutting down the ventilator. In addition to deliberate airway suctioning, unintentional placement of a nasogastric tube in the trachea or negative pressure from chest tubes put to suction with the chest closed in the presence of a large bronchopleural communication may also produce this problem. The flow rate through the bronchoscope suction port was measured at only 4.6 l/min (at −375 mmHg wall suction)—a flow inadequate to cause the malfunction. Flow rates in larger bronchoscopes were not tested. The solution to this situation requires the maneuver described. By breaking the vacuum seal, the piston is allowed to reset. The mechanical ventilation override control present on this machine does not work in this situation.

In this case, the patient was unharmed. In review, we make the following recommendations: (1) awareness of the causes and avoidance where possible; (2) switching to manual ventilation and filling the reservoir bag before airway suctioning; (3) availability of an Ambu bag to simplify the circuit if a ventilator problem occurs; and (4) machine function and problem-solving education. Routine procedures may trigger new problems when performed with new equipment. We recognize that the benefits of new technology can be tempered by their complexities and our limited ability to achieve intuitive solutions when those problems occur.