The automated machine checkout, available on most modern anesthesia workstations, is essential to detect and address malfunctions before an anesthetic delivery and is critical for patient safety. Even when the automated machine checkout indicates readiness of the anesthesia machine, unexpected mechanical failures may occur and require providers to remain vigilant. Previous reports of machine failures despite an uneventful automated machine checkout have been documented,1–4 mostly related to the adjustable pressure limiting valve. We present a case of postanesthesia induction ventilator failure in the setting of a successful automated machine checkout caused by a defective CO2 absorbent canister.
Written Consent Statement
Written consent was obtained from the patient to publish this case report. The report complies with the standards of the Health Insurance Portability and Accountability Act of 1996.
DESCRIPTION OF THE CASE
Before a general anesthetic, a Dräger Apollo ventilator (Telford, PA) passed the automated machine checkout. Anesthesia was induced, and a laryngeal mask airway was placed successfully. The ventilator was set up to volume auto flow mode, with a tidal volume of 600 mL, fraction of inspired oxygen 30%, positive end-expiratory pressure of 5 cm H2O, and respiratory rate of 12 breaths/min. The ventilator subsequently indicated a circuit leak and a failure to achieve inspiratory pressures and tidal volumes in the aforementioned mode. The end-tidal CO2 (ETco2) increased from 39 to 58 mm Hg with an alteration in the slope of phase III, suggesting an obstructive waveform on the capnogram.
Although no obvious circuit leak was apparent, the patient was intubated endotracheally to exclude poor laryngeal mask airway fit as a leak source. Rocuronium was used for muscle relaxation. Bilateral breath sounds without wheezing were present, and ETco2 detection was confirmed. Minutes later, in the same ventilatory mode, the monitor display again indicated failure to achieve inspiratory pressures, elevated ETco2, and a circuit leak (Figure 1). We observed a mismatch between the selected and delivered tidal volumes, and we increased the fresh gas flow to compensate for a suspected leak.
Shortly thereafter, the monitor display reported a “negative pressure” and a need to “reinstall ventilator.” Further attempts to ventilate the patient manually via the machine circuit and bag revealed increased inspiratory resistance. We ruled out bronchoconstriction, inadequate muscle relaxation, and other patient-related factors. We assessed the seemingly contradictory findings between mechanical and manual ventilation. A decision was made to convert the balanced to a total intravenous anesthetic (TIVA) technique and to provide manual ventilation by bag valve, which did not demonstrate any increased inspiratory resistance. During the procedure, a biomedical engineer performed a complete automated anesthesia machine check without identifying any problems. The inspiratory and expiratory valves, as well as the link to the CO2 absorbent canister in the internal panel, also were examined without revealing a clear cause for our observations. After reconnection of the ventilator to the endotracheal tube, the issue resurfaced, and the case was completed under TIVA with bag valve ventilation.
During a subsequent detailed investigation, the ventilator failure was fully reproducible with the breathing circuit connected to the ventilation bag that served as an artificial lung. Close examination of the ventilator showed that the upper filter of the tube insert (part # M33728, Figure 2) inside the CO2 canister (part #M3319) was broken, allowing entry of absorbent granules (Figure 3), which explains the observed increased resistance to inspiratory flow. Draeger was informed of this event on May 24, 2016.
We report an anesthesia machine failure as a result of a broken CO2 absorbent canister not detected by the automated machine checkout. The upper part of the CO2 absorbent insert (a protective filter) was broken (Figure 3). During refilling of the CO2 absorber, granules must have entered the inner tube of the absorber. These granules subsequently obstructed the inspiratory branch of the ventilator, causing increased inspiratory resistance to manual ventilation. At the same time, the alarms suggested a leak within the system. Because of the latter concern, we excluded common causes of leaks, and we secured the airway with an endotracheal tube.
Manual ventilation was difficult because of increased resistance to bagging, but this was not reproducible when the patient was ventilated with bag valve ventilation independent from the anesthesia machine. Draeger was informed via email, and their input via phone call on May 26, 2016 was helpful in discerning the mechanism of the issue. The company representative referred us to the Apollo anesthesia machine operator manual, which provided us with a diagram showing the internal arrangements of the ventilator valves, pressure sensors, and the ventilation bag (Figure 4). The increased inspiratory resistance during manual ventilation using the machine circuit is best explained by the fact that the bag is upstream from the obstruction of the CO2 absorbent canister (Figure 4). Therefore, compression of the bag will encounter resistance from the downstream obstruction. At the same time, the airway pressures throughout the case were never documented as high, as the pressure sensor is located downstream from the CO2 canister. Instead, the pressure sensor detected low volumes and pressures during controlled ventilation due to the limited volumes collected from the CO2 canister. Machine breakdown was not suspected until the machine itself reported a failure. Because the “ventilator failure” alarm was recurring, we decided to complete the surgery uneventfully under TIVA with manual bag valve ventilation. The patient denied any intraoperative awareness. After a thorough investigation, the internal CO2 absorbent canister insert was replaced and no further problems occurred. To this date, Draeger has not received any reports that are similar to our observation.
The automated anesthesia machine checkout procedure is mandatory to detect malfunctions in the anesthesia machine that can be addressed before clinical use. Earlier generations of anesthesia machines allowed for manual machine checkouts. The current generation of sophisticated machines has a built-in automated checkout system that follows a computerized algorithm from the manufacturer. Many anesthesiology societies, including the American Society of Anesthesiologists (ASA), have established guidelines for the preanesthesia machine check. Because the differences between the available anesthesia work stations, the ASA website currently includes peer-reviewed checkout procedure templates from 3 different institutions specific to the Draeger Apollo anesthesia machine. For example, the ASA guidelines suggest to perform a 2-bag test of the machine and circuit to ascertain to and fro gas flow between the bag and the test-lung without requiring a specific ventilator mode other than manual for the Draeger Apollo anesthesia machine.1 In contrast, the 2-bag test proposed by the Association of Anesthetists of Great Britain and Ireland requests to turn the ventilator on to aerate the test lung.2 By following the ASA recommendations, the 2-bag test might not have discovered the issue we encountered. Use of the ventilator mode during the test as per the Association of Anesthetists of Great Britain and Ireland guidelines would have increased the chance to detect the problem.
Several reports of failure to ventilate after uneventful anesthesia machine self-checks relate to trapping of the gas sampling line under the adjustable pressure limiting valve.3–6 “Human error” has been identified as a potential contributing factor for machine malfunction in the literature, but the term is poorly defined.7 In our case, the equipment failure was caused by shearing of the insert filter in the CO2 absorbent canister. The machine check was not able to recognize the partial obstruction within this component. Distraction, training deficits regarding the anesthesia machine assembly and function, or carelessness of the personnel involved with anesthesia equipment maintenance may have contributed to this event. Public reporting of specific issues has led to modifications in the machines to reduce the risk of technical complications during subsequent use.
The use of disposable CO2 canisters could have prevented this particular event, and such an option should be considered by the departmental anesthesia safety committees. The anesthesia personnel were briefed and educated on our findings and their significance; however, our case highlights the importance of a thorough understanding of anesthesia machine equipment by both providers and anesthesia technicians, with regular updates in a constantly changing technological environment. Scheduling equipment fairs and workshops under the leadership of trained anesthesia technicians or medical engineering personnel could help to achieve this goal. Departments need to recognize the vital role of anesthesia support personnel maintaining essential equipment for patient safety and should systematically promote and support ongoing professional development opportunities.
Our experience demonstrates that even when the automated machine check indicates proper anesthesia machine function, unexpected malfunctions may lead to unforeseen mechanical failures and require providers to remain vigilant. Any personnel involved in the use and maintenance of anesthesia machines should have a thorough understanding of its parts and their assembly, and we encourage continuous education in an increasingly technology-driven environment.
Name: Ingrid Moreno-Duarte, MD.
Contribution: This author helped conduct the case, collect and analyze the data, and prepare the manuscript.
Name: Julio Montenegro, BMT.
Contribution: This author helped collect and interpret the data, and prepare the manuscript.
Name: Konstantin Balonov, MD.
Contribution: This author helped conduct the case, collect the data, and prepare the manuscript.
Name: Roman Schumann, MD.
Contribution: This author helped analyze and interpret the case, and prepare the manuscript.
This manuscript was handled by: Hans-Joachim Priebe, MD, FRCA, FCAI.
1. Recommendations for Pre-Anesthesia Checkout Procedures. Sub-committee of ASA committee on equipment and facilities. 2008. Available at: http://www.asahq.org/For-Members/Clinical-Information/2008-ASA-Recommendations-for-PreAnesthesia-Checkout.aspx
. Accessed August 30, 2016.
2. Association of Anaesthetists of Great Britain and Ireland. Checking anaesthetic equipment 2012. Anaesthesia. 2012;67:660668.
3. Hennenfent S, Suslowicz B. Circuit leak from capnograph sampling line lodged under adjustable pressure limiting valve. Anesth Analg. 2010;111:578.
4. Kibelbek MJ. Cable trapped under Dräger Fabius automatic pressure limiting valve causes inability to ventilate. Anesthesiology. 2007;106:639640.
5. Robards C, Corda D. A potential hazard involving the gas sampling line and the adjustable pressure limiting valve on the Drager Apollo Anesthesia Workstation. Anesth Analg. 2010;111:578579.
6. Vijayakumar A, Saxena DK, Sivan Pillay A, Darsow R. Massive leak during manual ventilation: adjustable pressure limiting valve malfunction not detected by pre-anesthetic checkout. Anesth Analg. 2010;111:579580.
7. Fasting S, Gisvold SE. Equipment problems during anaesthesia—are they a quality problem? Br J Anaesth. 2002;89:825831.