Anesthesia equipment malfunction can result in severe morbidity or mortality if not treated and corrected in a timely manner. According to a report from the American Society of Anesthesiologists Closed Claims Project published in 1997, the most frequent adverse outcomes due to gas delivery equipment malfunction were death (47%) and brain damage (29%). Within this category, problems with the patient breathing circuit were the most common source of injury (39%).1 To address these problems, our scope of monitoring has greatly expanded, and anesthesia machine manufacturers have automated many components of the preuse checkout in their newest generation of products. Here we report a case where the automated preanesthesia checkout procedure failed to recognize a near-complete circuit obstruction before the induction of anesthesia. We discuss how automated machine check modules are not designed to comply with the 2008 American Society of Anesthesiologists Recommendations for Preanesthesia Checkout Procedures and how thoughtfully incorporated manual tests still need to be essential components of our practice.a
Written parent/guardian permission to publish this report was obtained before its submission. IRB approval was not obtained.
An 11-year-old boy (38 kg, 150 cm) presented for an elective tonsillectomy and adenoidectomy. His surgery was scheduled as the last case of the day. According to our institution’s preanesthesia checkout protocol for the GE Datex-Ohmeda Aisys Carestation, an automated System Machine Check was performed at the start of the day to evaluate the anesthesia machine, ventilator, and vaporizer. In addition, at the beginning of the day and after each new disposable pediatric anesthesia circuit was installed, an automated Circuit Machine Check was also performed to evaluate the integrity of each circuit. All routine checks yielded no errors.
After the placement of standard monitors, anesthesia was induced via mask using sevoflurane with nitrous oxide and oxygen. Induction of anesthesia was normal, and after insertion of an IV cannula, the trachea was intubated without difficulty. Following intubation however, a sustained but dampened and irregular end-tidal CO2 waveform was noted during mechanical ventilation. There was limited chest rise, and auscultation revealed faint breath sounds without wheezing. High inspiratory pressures of up to 40 cm H2O were required to yield 10 mL/kg tidal volume using manual ventilation. Bronchospasm or tracheal tube obstruction were initially suspected. However, subsequent treatment with inhaled albuterol, followed by tracheal reintubation both failed to improve ventilation mechanics.
The patient was hemodynamically stable, and oxyhemoglobin saturation remained normal during our initial treatments. While waiting for a flexible fiberoptic bronchoscope to be delivered to the room, a presumptive diagnosis of an equipment failure was considered. The tracheal tube was disconnected from the anesthesia circuit, and the lungs were ventilated via a Jackson-Rees circuit (Armstrong Medical, Coleraine, Ireland). This immediately resulted in normal chest rise and breath sounds. Inspection of the anesthesia circuit identified a manufacturing defect distal to the Y-piece that left a clear, hard piece of plastic forming a near-total obstruction of the lumen (Fig. 1). After this was recognized, the circuit was replaced, and the remainder of the case proceeded uneventfully.
Between 1956 and 2002, 41 cases of circuit obstructions by foreign bodies were reported to the Department of Health in Great Britain. Two of these incidents resulted in the death of a patient.b Similar instances have been reported elsewhere, where circuit obstructions due to plastic packaging,2,3 IV set caps,4,5 or manufacturing defects,6,7 have been extensively described.
After this event, our first goal was to analyze how this potentially lethal equipment malfunction was missed by our departmental preanesthesia checkout protocol. First, we reexamined the components of our 2 automated machine check tests.
The System Machine Check is intended to evaluate the anesthesia machine, ventilator, and vaporizer. It is performed while the machine is in the ventilator mode. It is a 2-step check where, after connection to the inspiratory and expiratory ports of the workstation, the circuit is initially left open to air at the Y-piece. During this first, open-circuit phase, the pipeline oxygen supply pressure is checked, and the ventilator function is tested by generating a few ventilator breaths. After the bellows are confirmed to be empty by the machine, the circuit is then purposefully occluded just distal to the Y-piece on a machine-mounted leak test plug for a second series of checks. The System Machine Check then continues to evaluate the following variables: (1) circuit leak (mL/min), (2) circuit compliance (mL/cm H2O), (3) proper mechanical ventilator function, (4) accurate O2/Air/N2O flow control operation, (5) proper vaporizer function, and (6) presence of battery backup and electrical power.
The Circuit Machine Check is intended to evaluate the integrity of the circuit. After installation onto the inspiratory and expiratory ports of the workstation, all components of this check are performed with the distal circuit occluded on a machine-mounted leak test plug. During this check, the anesthesia ventilator is turned off, and the adjustable pressure-limiting valve is set to between 30 and 70 cm H2O. While in the manual ventilation mode, the following variables are checked: (1) adequate pipeline oxygen supply pressures, (2) proper airway pressure transducer function, (3) circuit leak (mL/min) that is within an acceptable range, and (4) proper functioning of an alternate oxygen flow source.
After the completion of our case, the system and circuit machine checks were reperformed on our partially obstructed, defective circuit. Both checks were completed again without reporting errors. Next, when a complete obstruction distal to the Y-piece was experimentally created (by occluding the circuit with modeling clay), an error was generated during the System Machine Check due to inadequate emptying of the ventilator bellows between the first and second phase of the test. Performance of a circuit machine check, however, on a completely obstructed circuit by design, yielded no errors.
We then evaluated circuit obstructions proximal to the Y-piece. When complete occlusions were created with modeling clay in either the inspiratory limb or the expiratory limb, both the System Machine Check and Circuit Machine Check yielded errors. Experimentally created incomplete circuit obstructions in the inspiratory or expiratory limb, however, yielded Circuit Machine Check errors but no System Machine Check errors. Table 1 summarizes the specifics of our simulated circuit obstruction findings and the type of errors produced.
Incomplete circuit obstructions distal to the Y-piece, like the one found in our defective circuit, failed to produce errors during both machine checks. The manufacturer was contacted to discuss ways to better identify these difficult-to-detect obstructions. During our discussions, engineers suggested the following additional tests: (1) performing a manual ventilation oxygen flush test; (2) testing for appropriate ventilator function with a test lung connected to the patient Y-piece; or (3) visually inspecting the circuit distal to the Y-piece.
A manual ventilation oxygen flush test identifies an obstructed circuit by the impeded outflow of gas when the oxygen flush valve is triggered through an open circuit. While in the manual ventilation mode, with the adjustable pressure-limiting valve completely closed and the circuit opened to the air, a 10-second circuit flush via the oxygen flush button results in increasing peak airway pressures with increasing levels of circuit obstruction. We simulated the results of this test using a GE Vital Signs Pediatric Breathing Circuit (with a 2-L reservoir bag; General Electric, Fairfield, CT) and obstructed the circuit with occlusive tape distal to the Y-piece. Occlusion area was estimated by incrementally dividing the nonobstructed surface area by half. Progressive reservoir bag distension due to increasing airway pressure was easily evident only at high levels of circuit obstruction (approximately 97% occlusion). The defective pediatric circuit described in this case also produced notable reservoir bag distension during this test. Typical findings during this test are summarized in Table 2.
The manufacturer also suggested a check of ventilator function with a test lung at the start of the day and with each circuit change. At this point, our discussions with the manufacturer and our technical department made it clear that several versions of the Aisys User’s Reference Manual had been released since the time of our original purchase. As detailed in more recent versions (Software Revision 3.X) of the User’s Reference Manual, the test lung check is performed with the ventilator in volume-control mode and with standard settings (tidal volume 400, respiratory rate 12, I: E 1:2. Tpause off, positive end expiratory pressure off, Pmax at 40). For the test lung, we used a 2-L reservoir bag included in the GE Vital Signs Pediatric Breathing Circuit. With increasing levels of obstruction, the ventilator was unable to deliver the set tidal volume without triggering a high airway pressure alarm. Interestingly, when tested on simulated obstructions, only severe obstructions of approximately 97% of the circuit lumen yielded alarms during this test. We also confirmed that the defective circuit described in our case produced an alarm when evaluated with this test. These results are also summarized in Table 2.
Unfortunately, our departmental preuse checkout protocols had not been updated to reflect these changes in the User’s Reference Manual. In this particular case, addition of the test lung procedure would have detected the circuit obstruction before the induction of anesthesia. Our departmental preuse checkout protocols have since been updated to include a test lung evaluation. However, this scenario highlights that, in this digital era when software revisions and updates to a workstation user’s reference manual are common, close and continuous communication with the manufacturer is more important than ever.
In 2008, the American Society of Anesthesiologists published recommendations for preanesthesia checkout procedures. At that time, they stated that “Anesthesia delivery systems have evolved to the point that one checkout procedure is not applicable to all anesthesia delivery systems.” Thus, their goal was simply “to provide guidelines applicable to all anesthesia delivery systems so that individual [anesthesia] departments can develop a preanesthesia checkout that can be performed consistently and expeditiously.” Fifteen general items essential to the preuse checkout were highlighted in these guidelines. However, the specific steps needed to successfully perform each item were left to be analyzed and formulated by individual anesthesia departments based on their particular anesthesia workstation and clinical practice setting. It is our view that anesthesia departments must similarly be responsible for evaluating how changes in manufacturer recommendations affect compliance with these guidelines. Though one may argue that the anesthesia care provider is ultimately responsible for ensuring that workstations are safe and ready for patient use, the complexity of our current workstations make system updates or changes to the user’s reference manual difficult to identify by the end-user during daily clinical care. Furthermore, multiple providers and technicians are often involved in performing different components of the preuse checkout protocol. These factors create significant complexities that must be thoughtfully analyzed, a task that anesthesia departments are uniquely able to perform.
The GE Datex-Ohmeda Aisys Carestation is one of the most sophisticated anesthesia workstations currently marketed. It features digitally controlled ventilator, vaporizer, and gas delivery systems. However, despite this technological sophistication, the 2008 American Society of Anesthesiologists Recommendations for Preanesthesia Checkout Procedures still cautioned against relying solely on automated checks alone, stating that “[s]imply because an automated checkout procedure exists does not mean it can completely replace a manual checkout procedure or that it can be performed safely … without a thorough understanding of what the automated checkout accomplishes.” In this particular case, the Circuit Machine Check is designed to be performed on an obstructed circuit and thus cannot detect obstructions distal to the Y-piece. This highlights that Item #13 of the 2008 American Society of Anesthesiologists Preanesthesia Checkout Procedures: Verifying that gas flows properly through the breathing circuit during both inspiration and exhalation requires the performance of supplementary manual checkout procedures and that the automated checkout procedures of the Aisys Carestation cannot satisfy this item when performed alone.c
Furthermore, our simulated circuit obstructions also demonstrate that the most current preuse testing protocols can still only detect severe distal obstructions of ≥97%. Therefore, routine performance of these tests will still fail to detect severe and clinically significant circuit obstructions. Even in this era of sophisticated automation, a visual inspection of the circuit remains the most reliable method of identifying these obstructions, and a manual check for the presence of a functioning self-inflating manual ventilation device remains an essential first step in all preuse checkouts.
In this era of increasingly sophisticated machine check technology, this case highlights that automated tests are designed to function only as a part of a preuse checkout. To assure the highest safety of clinical care, institutions must continually and systematically assess evolving manufacturer recommendations and software changes in the context of the 2008 American Society of Anesthesiologists Recommendations for Preanesthesia Checkout Procedures. In the end, anesthesia providers must not be lulled into a false sense of security by improved technology but must acknowledge that these technologies are imperfect when used alone and that manual checks are still essential components of safe care.
a 2008 ASA Recommendations for Pre-Anesthesia Checkout: Available at: http://www.asahq.org/For-Members/Clinical-Information/2008-ASA-Recommendations-for-PreAnesthesia-Checkout.aspx. Accessed April 15, 2013
b Protecting the breathing circuit in anesthesia. Available at: http://www.frca.co.uk/documents/Protecting%20the%20PBC.pdf. Accessed April 15, 2013
c Feldman JM, Olympio MA, Martin D, Striker A. New Guidelines Available for Pre-Anesthesia Checkout. Anesthesia Patient Safety Foundation (APSF) Newsletter 2008; 23:6–7: Available at: http://www.apsf.org/newsletters/pdf/spring2008.pdf. Accessed August 29, 2013
1. Caplan RA, Vistica MF, Posner KL, Cheney FW. Adverse anesthetic outcomes arising from gas delivery equipment: a closed claims analysis. Anesthesiology. 1997;87:741–8
2. Chacon AC, Kuczkowski KM, Sanchez RA. Unusual case of breathing circuit obstruction: plastic packaging revisited. Anesthesiology. 2004;100:753
3. Hindman BJ, Sperring SJ. Partial expiratory limb obstruction by a foreign body abutting upon an Ohio 5400 Volume Monitor Sensor. Anesthesiology. 1986;65:349–50
4. Carr ZJ, Deal ER, Goldberg ME. Ball valve effect precipitating an anesthesia circuit obstruction by an intravenous set cap within the elbow connector. J Clin Anesth. 2010;22:486–7
5. Krensavage TJ, Richards E. Sudden development of anesthesia circuit obstruction by an end-tidal CO2 cap in the gas sampling elbow. Anesth Analg. 1995;81:207
6. Yang CH, Chen KH, Lee YE, Lin CR. Anesthetic breathing circuit obstruction mimicking severe bronchospasm: an unusual manufacturing defect. Acta Anaesthesiol Taiwan. 2012;50:35–7
7. Eckhout GV, Bhatia S. Another cause of difficulty in ventilating a patient. J Clin Anesth. 2003;15:137–9