Expired Oxygen as the Unappreciated Issue in Preventing Airway Fires

Conventional wisdom and traditional habits need to be challenged continually to help understand and, thus, advance clinical practice. Many features of routine anesthesia protocols have evolved over time because they seem logical and are faithfully drilled into trainees by well-meaning faculty members who learned them from their teachers, and so on…. In this issue of the journal, Remz and colleagues1 address the conventional teaching that reducing the inspired oxygen concentration will prevent an airway fire when an ignition source is used in the airway. All trainees see images of the dramatic “blowtorch ignition” of a plastic endotracheal tube through which 100% oxygen is flowing2 (simulating a recognized danger during a “simple, routine” tracheostomy)3–5 (Fig. 1). Airway fires are usually reported by otolaryngologists or anesthesiologists and have involved all types of surgery in the airway,6 including the common tonsillectomy.7 Ever since identification of the danger of a fire in a patient’s airway, ignited by an electrocautery, a laser, or a fiberoptic light, the “wisdom” handed down has been that the inspired oxygen concentration delivered from the anesthesia machine should be made <30% (FIO2 <0.30) or reduced to the “minimum possible” to limit the dramatic increase in flammability of plastic, cloth, and tissue in an oxygen-enriched environment. Remz and colleagues1 posit that the real issue is not only the inspired oxygen concentration but also the expired concentration. Using a simulation model, they demonstrated that the relation between the 2 depends on several factors and, importantly, that expired oxygen concentration may be increased for a considerable period of time after the inspired concentration appears to be “safe,” which could create an unexpected, occult risk of airway fire. These findings necessitate a reevaluation of our wisdom and expansion of conventional teaching to recognize the role of expired oxygen in mitigating the risk of airway fires.Figure 1: Demonstration of rocket-like flames shooting from a tracheal tube caused by laser ignition of the tube with 100% oxygen flowing. Image courtesy of ECRI Institute.The American Society of Anesthesiologists Practice Advisory for Prevention and Management of Operating Room Fires8 states that the consultants and American Society of Anesthesiologists members strongly agree that for high-risk procedures where an ignition source is in proximity to an oxidizer-enriched atmosphere, the inspired oxygen fraction (FIO2) delivered to the patient should be kept “as low as clinically feasible.” Regarding use of an ignition source to enter or work in the trachea when the patient’s lungs are being mechanically ventilated via a tracheal tube, the American Society of Anesthesiologists advisory cites a survey asking practitioners how much time did they believe was needed to reduce oxygen (or nitrous oxide) concentration to a safe level before using an ignition source.8 The answers ranged from <1 to 10 minutes (mean 2.9 minutes). These values likely reflect real-world experience in clinical anesthesia practice, in which there is a wide spectrum of lung function and pathology, leading to variation in time constants and, thus, in washout time of the high concentration of oxygen. Remz and colleagues1 used a mechanically ventilated human patient simulator with routine settings. Starting with inspired oxygen concentrations of 100% and 60% in the breathing circuit, they replaced the original oxygen-rich fresh gas flows (FGF) with air (21% oxygen) and repeatedly measured inspired and expired oxygen concentrations at the Y-connector over time, until both oxygen concentrations were <30% (the alleged maximum concentration recommended when a laser is used in the airway). Using this model, they studied the replacement FGF of air at 2 L/min and 5 L/min and also compared short (fully compacted) and long (fully extended) breathing circuits. The results were illuminating. As would seem intuitive, the oxygen concentrations took longer to decrease to <30% when starting from 100% vs starting from 60% oxygen. Likewise, the oxygen concentrations decreased much more quickly at the higher FGF of air and somewhat faster in the short circuits versus the long ones (but not as much as might have been expected based on the large difference in volumes). The core results involved the comparison of the decrease over time in the inspired versus expired oxygen concentrations to <30%. With 5 L/min FGF of air into 100% oxygen (long circuit), the inspiratory concentration took <1 minute to fall, while the expiratory concentration took nearly 3 minutes. At 2 L/min FGF of air, paradoxically, the inspiratory concentration decreased in 8 minutes, and the expiratory concentration decreased in 7 minutes. The remarkably slow decreases in circuit oxygen concentration at the lower FGF of air are very important for fire risk considerations. A wait of 7 or 8 minutes will seem like an eternity to an impatient surgeon holding a cautery pencil or laser probe. Also, the observation that the expiratory concentrations track the inspiratory concentrations closely at low FGF of air (and, in fact, decreases slightly faster) may seem counterintuitive. One possibility is that this is related to the oxygen consumption rate of the simulator, but, in any case, this phenomenon deserves further study. However, it is more important to emphasize the key finding that at high FGFs (which are much more likely to be employed when the transition to air is made; otherwise, the wait would seem interminable), the decrease in expired oxygen concentration is significantly slower than for the inspired concentration. Thus, in such situations, using the inspired oxygen concentration as the basis for signaling the impatient surgeon to activate the cautery or laser would be dangerous. An oxygen-enriched environment persists (the expiratory phase is usually two thirds of the ventilatory cycle) and, thus, there is a dramatically increased risk of airway fire when cautery or a laser is used. One potentially important factor that was not studied by Remz and colleagues1 following switching from 100% or 60% oxygen to air was the effect of minute ventilation on the decrease in expired oxygen concentration. In their study, they used a minute ventilation set to 4 L. It is well established that to achieve a desired concentration of a gas in the lungs (in this case, to <30% oxygen), the 2 important variables are the FGF (in this case, of air) and minute ventilation.9 This is, after all, how we speed a patient’s emergence (anesthetic washout) following the administration of an inhaled anesthetic.10 Establishing the interaction between FGF of air and minute ventilation on the rate of decrease in expired oxygen concentration would certainly have clinical application during “deoxygenation” before the use of an ignition source. It seems intuitively obvious that increasing both FGF of air (e.g., to 10 L/min) and minute ventilation (e.g., by 50%) would substantially decrease the time to achieve an expired oxygen concentration of <30% and, thus, is recommended in such situations. Despite the conventional wisdom concerning decreasing the oxygen concentration, airway fires continue to occur. The number of these fires is unknown because there is no mandatory reporting to a centralized registry. Remz et al.1 cite the often-quoted annual estimate by ECRI of >120.11 If, for example, a “blowtorch” fire of an endotracheal tube during a tracheostomy caused a patient’s death, that would require reporting to The Joint Commission as a “sentinel event.” An airway fire should be a “never event,” something that simply should never happen and that should be reported to the appropriate regulatory and accrediting entities so that education occurs, and warnings can be generated to help prevent similar events in other patients. In that vein, airway fires are featured in the dramatic and effective Anesthesia Patient Safety Foundation’s “Fire Safety Video”a that also covers fires during monitored anesthesia care for superficial surgery on the upper body with open delivery of supplemental oxygen underneath drapes. Airway fires were also included in the analysis of the American Society of Anesthesiologists Closed Claims Database reports of operating room fires.12 Ten of the 103 cases since 1985 were airway fires in intubated patients, 4 during tonsillectomy and 6 during tracheostomy. In the tonsillectomy cases, an uncuffed endotracheal tube or a leak around an endotracheal tube cuff was reported as the oxidizer source. Thus, overall, while airway fires may be relatively infrequent, they do occur, can be severely injurious, and also should be preventable with awareness and precautions. Regarding potentiation of conventional wisdom, an erudite and extensive discussion of endotracheal tube fires in airway laser surgery13 states only: “…most clinicians recognize the need to reduce the FIO2 to <0.40 or to the minimum concentration consistent with patient oxygenation.” There are no references and no mention of time course or expired oxygen concentration, illustrating the knowledge gap addressed by the data reported by Remz and colleagues.1 Overall, in the (relatively infrequent) cases where the patient truly is dependent on supplemental oxygen for survival during the surgery, the logical approach is to decrease the inspired and expired concentrations to the absolute minimum necessary and then warn the surgeon that there is an increased risk of fire if he/she uses an ignition source. In this way, preventive precautions can be used in the surgical field, which would include not using the ignition source at all in appropriate circumstances. Furthermore, Remz and colleagues1 address the often-cited protocol response to an airway fire (previously a perennial question on the written anesthesiology board certification examination) by correctly raising the question of the implication of the lag in decrease of expired oxygen concentration after changing the FGF to air. If an airway fire should occur shortly after the switch to inspired air, initially disconnecting the breathing circuit from the tube or stopping the FGF (the traditional conventional wisdom) may actually be detrimental to the situation because that leaves in the airway only the oxygen-enriched expired gas that would be more likely to support vigorous combustion. Remz and colleagues1 recommend leaving the circuit with air flowing connected and simply immediately pulling out the tube. This is controversial and deserves research and further consideration. However, it seems logical and, until proven otherwise, should be done. The results presented by Remz and colleagues1 challenge conventional wisdom and provoke critical thinking. However, as can often be the case, how 1 dataset from a simulation experiment such as this should influence daily clinical practice is not completely clear. Another study, employing high-flow oxygen through an endotracheal tube into a gutted chicken carcass, showed no ignition by electrocautery of the tissue or tube at oxygen concentrations <45%.14 How might that observation relate to this question? It seems unlikely that there will be definitive research in the foreseeable future demonstrating that it is safe to maintain high airway oxygen concentrations during use of ignition sources. Accordingly, the reconsidered new wisdom should be that during airway cases, particularly tracheostomy, tonsillectomy, and laser surgery, the minimum possible oxygen concentration consistent with adequate patient oxygenation should be utilized. Monitoring of both inspired and expired oxygen concentrations appears to have advantages, particularly in airway cases. The current American Society of Anesthesiologists Standards for Basic Anesthetic Monitoring,15 under “Oxygenation” have the stated objective: “To ensure adequate oxygen concentration in the inspired gas,” and under “Methods: Inspired Gas”: “During every administration of general anesthesia using an anesthesia machine, the concentration of oxygen in the patient breathing system shall be measured by an oxygen analyzer with a low oxygen concentration limit alarm in use.” Thus, the concern has been to avoid delivery of a hypoxic gas mixture to the patient. Further, the most recent voluntary consensus standard applicable to the anesthesia workstation requires an oxygen monitor in the inspiratory limb or at the Y-piece.16 Most commonly, an oxygen analyzer is located in the vicinity of the inspiratory unidirectional valve in the circle breathing system. Many, if not most, anesthetizing locations now have multigas analyzers that sample respired gas from an adapter at the patient’s airway and measure both inspired and expired gas concentrations, including those of oxygen, on a breath-by-breath basis. These analyzers may be integral with the anesthesia workstation or free-standing units. In these devices, oxygen concentration is measured by a rapidly responding paramagnetic analyzer that provides inspired and end-tidal oxygen concentrations and a display of the oxygram (analogous to the familiar capnogram).17 Monitoring of end-tidal oxygen concentration has proven valuable in confirming the adequacy of preoxygenation prior to induction.18 Importantly, concerning airway fires, it serves as a monitor of deoxygenation in anticipation of activation of an ignition source. With increasing prevalence of these multigas analyzers, monitoring of expired oxygen concentration will become increasingly routine, including as a feature to help prevent airway fire. If, in anticipation of activation of an ignition source, the FGF containing supplemental oxygen is appropriately switched to air at a high-flow rate, the anesthesia practitioner should wait and watch until the oxygen concentration of both inspired and expired gas has fallen well into the safe range (arguably <30%) before authorizing the use of an ignition source. Increasing the minute ventilation in addition to FGF of air should accelerate the attainment of the safe range. All of this requires vigilance and good rapport and communication with the surgeon. Diligent application of such a protocol should, indeed, help to make airway fires a never event. DISCLOSURES Name: John H. Eichhorn, MD. Contribution: This author helped write the manuscript. Attestation: The author approved the final manuscript. Name: James B. Eisenkraft, MD. Contribution: This author helped write the manuscript. Attestation: The author approved the final manuscript. This manuscript was handled by: Sorin J. Brull, MD, FCARCSI (Hon).