Laryngeal mask airway and the ASA difficult airway algorithm.

Abstract

(Benumof) Professor of Anesthesia.Received from the Department of Anesthesiology, University of California San Diego Medical Center, San Diego, California. Submitted for publication May 10, 1995. Accepted for publication November 30, 1995.Address reprint requests to Dr. Benumof: Department of Anesthesiology, 8812, University of California, San Diego, 402 Dickinson Street, San Diego, California 92103.Shortly after the introduction of the American Society of Anesthesiologist's (ASA) difficult airway algorithm in 1991, [1,2]the United States Food and Drug Administration allowed the laryngeal mask airway (LMA) to be marketed in the United States, and it was immediately recognized as a major advance in airway management. [3]During these 5 yr, the literature on and the experience with the LMA has substantially increased. As a consequence, the role of the LMA both in the ASA algorithm on management of the difficult airway and in the management of the difficult airway in general has become more clearly defined. The most important understanding to emerge during this 5-yr period is that the LMA has worked well, with few complications, as a ventilatory device in patients whose tracheas cannot be intubated and/or whose lungs cannot be ventilated with conventional techniques, provided they do not have obvious periglottic pathology (massive cancer, hematoma, abscess, anaphylactic laryngeal edema). This article will sequentially, briefly review the original ASA difficult airway algorithm from a 1995 perspective, discuss the ways the LMA can be used, and conclude with a description of the multiple uses for the LMA within the ASA difficult airway algorithm. It must be noted that the ASA difficult airway algorithm can be changed only by a formal vote of approval by the House of Delegates of the ASA. Finally, it should be noted that a significant amount of the published information on the LMA is anecdotal, rather than from controlled clinical trials.(Figure 1) is a combination of two previous ways of depicting the ASA difficult airway algorithm, [1,2]i.e., the figure incorporates all previously approved ideas/concepts into one flow diagram. The algorithm begins with preoperative evaluation and recognition of the difficult airway. If a difficult airway is recognized, then it is logical to secure/guarantee the airway with awake tracheal intubation. For awake tracheal intubation to succeed, it is essential to prepare the patient properly. Once the patient is properly prepared, any one of a number of intubation techniques can achieve tracheal intubation (Figure 1), and the practitioner should make the choice based on their particular training, experience, and knowledge. Occasionally, awake tracheal intubation may fail because of lack of patient cooperation or equipment and/or operator limitations. Depending on the precise cause of failure of awake tracheal intubation, options include:(1) cancellation of surgery (patient may need further counseling regarding their degree of cooperation, different equipment/personnel obtained);(2) general anesthesia, if mask ventilation is considered likely to be easy;(3) regional anesthesia, if inadequate block can be followed by abrupt termination of surgery or awake tracheal intubation if necessary; and (4) creation of surgical airway, if surgery cannot be postponed and general anesthesia without endotracheal intubation is inappropriate. Rarely, creation of a surgical airway with the patient awake is the best first intubation choice (laryngeal/tracheal disruption, upper airway abscess, some mandibular-maxillary fractures).It is appropriate to attempt tracheal intubation after the induction of general anesthesia when a difficult airway is not recognized or a difficult airway is recognized but the patient is uncooperative and vigorously refuses awake tracheal intubation. With induction of general anesthesia in an uncooperative patient with a recognized difficult airway, one should consider the relative merits of preservation of spontaneous ventilation versus the use of muscle relaxants. In an elective case, positive-pressure ventilation via mask ordinarily will be employed before tracheal intubation, whereas in an emergency case with gastric contents present, positive-pressure ventilation via mask ordinarily will not be used before tracheal intubation. Tracheal intubation can be accomplished by any one of a number of choices. If a given intubation choice should fail, gas exchange should be controlled via mask ventilation. If the initial intubation choice was conventional laryngoscopy, the laryngoscopist should have a good idea of the degree of difficulty of laryngoscopy/intubation and should quickly (next attempt) make an optimal/best attempt at laryngoscopy. [4]Best attempt at laryngoscopy may be defined as being performed by an experienced endoscopist, with the patient in the optimal position, using optimal external laryngeal manipulation, and perhaps a single change in each of laryngoscope blade type and size. [4,5]If conventional laryngoscopy fails after a best attempt, there are acceptable alternative approaches that do not require use of a conventional laryngoscope (Figure 1). Repeated conventional laryngoscopy can cause laryngeal trauma (edema, bleeding), which may result in the loss of the ability to ventilate via mask. A best attempt at ventilation via a conventional mask, defined as a two-person effort using appropriately sized oropharyngeal and nasopharyngeal airways, should be made before abandoning mask ventilation.If the ability to ventilate via mask is lost, and it already has been demonstrated that the trachea cannot be intubated, a true emergency exists, and gas exchange must be restored immediately. The four acceptable responses to the "cannot ventilate, cannot intubate" situation are insertion of an LMA, [3,6]insertion of a Combitube, [7],* institution of transtracheal jet ventilation, [8]and creation of a surgical airway (if personnel, equipment, lighting, and positioning are appropriate). Finally, those caring for the patient must guarantee communication of this difficult airway experience to future care-takers. [9].The LMA may be used to provide a routine airway for use during general anesthesia or, less frequently, as a conduit for tracheal intubation. Before describing the multiple places that these two uses of the LMA have in the ASA difficult airway algorithm, it is appropriate to briefly review these independent functions of the LMA.Provided the depth of anesthesia is appropriate and adequate, the LMA makes a routine airway during general anesthesia in 95–100% of patients in which LMA use is deemed appropriate. [10]Compared with ventilation via mask, the LMA has been found to be a superior device during three widely different settings/conditions (routine outpatient procedures not requiring intubation, [11]neonatal resuscitation, [12]use by inexperienced personnel [13]). First and, from the perspective of user frequency, most importantly, a randomized prospective study compared the quality of the airway with the LMA with that of a standard face mask/oropharyngeal airway system. [11]Smith and White found that problems associated with airway management (difficulties in maintaining an airway and SpO2< 95%) were more common in patients in the face mask group than in the LMA group. Certainly one can expect the LMA to be superior to a face mask if the facial contours of the patient are not suited to the standard mask, [14]such as the edentulous patient with invaginated cheeks. [15]Other advantages of the LMA are related to the fact the mandible does not need to be supported, and therefore, the anesthesiologist's hands are not occupied, fatigue is less likely, and remote observation is possible. Many authors have found that the stimulation related to insertion of the LMA is approximately the same as that for an oropharyngeal airway. [6]In addition, the LMA has some advantages over a tracheal tube as a routine airway; these include rapid insertion in the hands of both anesthesiologists and nonanesthesiologists and less in situ stimulation. [6].Second, although the American Heart Association recommends that positive pressure, when indicated, be administered by bag-and-mask, [16]this can be difficult to do effectively in small neonates, [17]even for anesthesiologists. In a recent study, in all of 21 newborn infants weighing 3.4+/-0.6 kg (mean+/-SD) who required resuscitation, trained neonatologists were able to easily insert a size-1 LMA in < 12 s on the first attempt and obtain a clinically patent airway with LMA leak and peak airway pressures of 22 and 37 cm H2O, respectively. [12]All resuscitations were successful, and there were no complications related to the use of the LMA. [12]Although the LMA was not directly compared with bag-and-mask ventilation, these results are impressive. The authors commented that a randomized prospective study is underway to determine whether these good airway results with the LMA can be obtained by other medical personnel performing neonatal resuscitation. [12].Third, nursing personnel who had a 30-min instructional course with the LMA and face mask and some practice on a mannequin could achieve larger tidal volumes (435 ml, range 385–485 ml) in patients with the LMA compared to anesthesiologists (328 ml, range 277–378 ml) or nurses using a face mask (239 ml, range 170–309 ml). [13]Thus, when ventilation via mask is appropriate, inexperienced operators appear to have better ventilation capability with the LMA than with a face mask.One of the reasons the LMA works well as a routine airway is that exact positioning is not crucial for a clinically acceptable/patent airway. The usual fit of the LMA around the larynx, as assessed by using flexible fiberoptic endoscopy, [18–20]radiologic investigation, [21]and nuclear magnetic resonance imaging, [22]is variable. One definition of appropriate position of the LMA consists of the epiglottis and esophagus outside and the laryngeal opening within the rim of the LMA and is obtained only 45–60% of the time. [18–22]When the epiglottis is within the proximal rim of the LMA, the tip of the epiglottis is slightly downfolded toward the larynx 50–90% of the time, and the lateral aryepiglottic folds are infolded toward the larynx half the time. [18–22]The slight downfolding means that the proximal anterior surface and tip of the epiglottis are against the upper half of the aperture bars, which is, in essence, close to the normal anatomic position. Consequently, visualization of the epiglottis against the aperture bars (e.g., fiberoptically) causes a partial obstruction to the laryngeal view but not necessarily to gas flow. It is not surprising that, in 94–99% of adult [18,23]and pediatric patients, [19,20,24–27]there is no difficulty with ventilation, and the airway is ultimately judged to be clinically acceptable (although proper position may require two insertion attempts; see below). In fact, there are reports in which x-rays revealed the mask to be in the oral cavity or bending completely backward, neither of which had been detected clinically. [28,29]Finally, the distal rim of the LMA usually is wedged in the hypopharynx, but in 0–7% of cases, the esophagus may be seen inside the distal rim, and in some cases, the distal rim may be directly opposite the glottis. [30–32].The use of the LMA for routine airway management can be associated with a number of problems. These problems consist of clinically unacceptable nonpatent airway, the requirement for multiple insertion attempts in a small percentage of patients, aspiration of gastric contents, and suboptimal positive-pressure ventilation. The exact incidence of each of these problems, to some extent, must depend on insertion technique and skill. [33]Inadequate anesthesia may cause all of these problems, and therefore, a basic requirement for the safe use of the LMA is an adequate depth of anesthesia. [10]The most serious problem is that, in 0.4–6% of anatomically normal patients, placement of the LMA is clinically inadequate because of backfolding of the distal cuff, occlusion of the glottis by the distal cuff, complete backfolding of the epiglottis, or 90–180 degrees C rotation of the mask around the long axis. [11,18,19,23–27,34,35]The results of the report with a 6% incidence of a wholly inadequate airway with the LMA may have been related to the fact that the anesthesiologists participating in the study were second- and third-year residents [11]and thus somewhat inexperienced.The most common causes of poor LMA placement are inadequate anesthesia/relaxation (pharyngeal muscle and/or laryngeal spasm), [10]failure to negotiate the 90 degrees turn from the posterior pharynx to the hypopharynx, and choice of wrong LMA size. In 2–33% of LMA placements, more than one attempt is required, whether by residents or experienced practitioners, adult or pediatric patients. [11,22–29,34,36]Correct placement may be more difficult in patients with a small mouth, a large tongue and/or tonsils, and a posteriorly placed larynx (blocks the advancement of the tip of the LMA into the hypopharynx). Obviously, in view of these placement considerations, the use of the LMA is relatively contraindicated in patients with major local pathology in the pharynx and larynx such as tumor, abscess, edema, and/or hematoma.For two reasons, the LMA is also relatively contraindicated as a routine airway in patients who have a risk of regurgitation and/or active vomiting of gastric contents or have a large amount of blood present in the upper airway. First, the LMA does not provide an airtight seal around the larynx (usual pop-off pressures are 15–20 cm H2O), and therefore, the larynx/trachea may communicate with a fluid-filled pharynx and esophagus. Second, as described above, in 0–7% of patients, the esophagus is included within the rim of the LMA, [30–32]and in these instances, the esophagus may be directly exposed to positive-pressure ventilation. Thus, it is not surprising that studies using gastrointestinal tracers have found a large (up to 25%) but variable incidence of patients with gastrointestinal tracer within the LMA. [31,37,38]The variability in findings probably is related to a variable depth of anesthesia and/or intensity of surgical stimulus and, therefore, variation in abdominal, pharyngeal, and vocal cord muscle tone. [39,40]Only six cases of clinically obvious and serious vomiting and pulmonary aspiration have been reported: Four were thought to be associated with inadequate anesthesia [41–44]and two with movement [45]or change in position [46]of the patient. These few cases, plus cumulative data from two large prospective surveys, suggest an incidence of aspiration of 2.3 per 10,000 with the LMA in "low-risk patients." [35,43,47]This value is comparable to that found for elective anesthesia (2.6 per 10,000)[48]or outpatient anesthesia using the face mask or tracheal tube (1.7 per 10,000). [49]Two cases of massive gaseous gastric dilation have been reported. [50]The incidence of aspiration with the LMA has been subjected to metaanalysis and found to be comparable to that for outpatient anesthesia with the face mask and tracheal tube. [51]The general subject of aspiration risk with the LMA has been comprehensively reviewed. [52].The usual LMA seal/pop-off pressure is 15–20 cm H2O. In 48 anesthetized and paralyzed patients whose lungs were being ventilated via an LMA, peak pressures of 15, 20, 25, and 30 cm H2O caused mean leak fractions (fraction of inspired volume not returned) to progressively increase from 0.13 to 0.27 and frequency of gastroesophageal insufflation to progressively increase from 2% to 35%. [53]Consequently, the LMA is relatively contraindicated whenever it is anticipated that positive proximal airway pressures > 25–30, cm H2O will be required to adequately ventilate the lungs (i.e., existing or potential development of decreased lung compliance and/or increased airway resistance).Finally, because the LMA may become malpositioned or regurgitation/vomiting may occur at any time, the LMA is relatively contraindicated as the primary, electively instituted, ventilatory device whenever tracheal intubation cannot be readily accomplished (e.g., when the patient is prone, when the head of the operating table is turned away from the anesthesiologist, or when the potential for difficult intubation is recognized). [1–3,10,12]In the final analysis, judgment must be used to weigh perceived benefit from the LMA against the risk of this relative contraindication.There are numerous reports (mostly letters to the editor) of using the LMA as an airway intubator (conduit) for the blind passage of an endotracheal tube (ETT) or an intubating tracheal stylet or for the passage of a flexible bronchoscope under fiberoptic vision (an ETT may be passed over the stylet or fiberoptic bronchoscope (FOB)). [3]In view of how the LMA usually seats around the larynx, it is obvious that, when the LMA has a perfect central position (45–60% of the time, see above), any of the blind insertions has some chance of success. It is likely that the greater the degree of noncentral location of the LMA over the larynx, the less the chance of success of blind intubation. In addition, the unsuccessful (off-center) insertion of a rigid object through an already off-center, blindly inserted LMA may result in laryngopharyngeal injury. [10]When blind tracheal intubation is attempted through an LMA in a large series of patients thought to have normal anatomy, there is a 26–97% failure rate on the first attempt and a 10–70% overall failure rate with an ETT, [54–56]and there is an 18–70% overall failure rate with an intubating stylet. [57,58]With subsequent attempts, the cumulative success rate may progressively increase with change in head and neck alignment from the sniff position to varying degrees of atlantooccipital extension and flexion. [56].The use of cricoid pressure further decreases the chance of passing an ETT blindly through the LMA into the trachea for two reasons. First, when cricoid pressure is applied before the LMA is placed, the pressure prevents the tip of the LMA from fully occupying the 3.5-cm length of the hypopharynx behind both the arytenoid and cricoid cartilages (compare panels 3 and 4 of Figure 2and the top and bottom panels in Figure 3). [59,60]Thus, with cricoid pressure, the LMA may be wedged in the hypopharynx, but it can occupy only the 1.5 cm of the hypopharynx behind the arytenoid cartilages and is therefore 2 cm more proximal than usual. [59]Variable obstruction to the passage of the LMA by cricoid pressure (and perhaps caused by intermittent or momentary relaxation of cricoid pressure; see section IVB below) may explain why cricoid pressure has resulted in widely variable success in simply inserting the LMA from a low of 15%[61,62]to highs of 85%[63]and 90%. [64]If cricoid pressure impedes passage of the LMA, the resultant more proximal than usual, off-center location of the LMA should further decrease the success rate of blind intubation through the LMA.Second, the wedging into and subsequent full inflation of the cuff of an LMA in a hypopharynx constricted by the application of cricoid pressure causes the plane of the laryngeal aperture to tilt approximately 40 degrees anteriorly around the fulcrum of the inflated LMA distal cuff (Figure 3). [60]It seems likely that it would be more difficult for a blindly inserted intubating device to enter an aperture that is in a plane 40 degrees off the perpendicular to the insertion pathway than an aperture that is in the plane perpendicular to the insertion pathway. Failure to enter the trachea with an intubating stylet or ETT (both of which are semirigid) may result in periglottic trauma. In summary, blindly trying to intubate through a blindly inserted LMA may be unsuccessful and dangerous, especially in the presence of cricoid pressure.Conversely, passage of an FOB through the LMA is nearly 100% successful in most series. [3]A 6.0 mm-ID cuffed ETT may be passed over the FOB and through the shaft of the size-3 and -4 LMA and a 7.0 mm-ID cuffed ETT may be passed over the FOB and through the shaft of the size-5 LMA. If a larger ETT is desired, the LMA and the 6.0 or 7.0 mm-ID cuffed ETT may be exchanged for a larger ETT over a jet stylet. [1]Table 1shows the various ETT sizes that fit through all LMA sizes and the FOBs that fit through these ETTs. If a 4.0 mm-OD or smaller fiberscope is used with the 6.0 or 7.0 mm-ID ETT, the lungs can be continuously ventilated around the fiberscope but within the ETT by passing the fiberscope through the self-sealing diaphragm of a bronchoscopy elbow adapter; the distal and proximal ends of the bronchoscopy elbow adapter are connected to the ETT and ventilatory apparatus, respectively (Figure 4). With a 4.0 mm-OD fiberscope/6.0 mm-ID ETT combination, the space available for ventilation around the fiberscope corresponds to a 4.5 mm-ID ETT.There are three interrelated problems with passing an ETT over an FOB through the shaft of the LMA. First, in patients with a long incisor-to-glottis distance, the cuff of the 6.0 mm-ID ETT might be near or between the vocal cords. Table 2shows the length of the LMAs (proximal edge of LMA adapter to grille), the maximum length of the various ETT sizes with the adapter inserted, and the difference between these two lengths (i.e., the distance the tip of the ETT projects beyond the grille of the LMA). With the size-3 and -4 LMAs and the 6.0 mm-ID cuffed ETT combination, the tip of the ETT projects only 8–9 cm, and in a normal sized adult, the cuff cannot be in the trachea more than 3–4 cm (proximal border of ETT cuff to tip is approximately 5 cm;Figure 4, bottom). These calculations correlate with the finding that the mean +/-SD distance between the grille of the LMA and the vocal cords is 3.6+/-0.5 cm in males and 3.1+/-0.5 cm in females. [65]Second, the internal diameter of the shaft of the LMA limits the size of the ETT that can be passed through the LMA. In some adults who require large tidal volumes, a 6.0 mm-ID cuffed ETT may not allow generation of flow rates adequate for effective ventilation (e.g., those with V/Q mismatching, high metabolic rates). A change to a larger-size ETT requires use of a tube changer, which involves new dangers and risks. Third, a maximally inserted but still marginally long ETT does not allow for removal of the LMA over the ETT without risking extubation.Solutions for these three problems vary according to which of the three problems they solve and according to the size of the patient (and therefore the size of the LMA and the ETT). With regard to a standard cuffed 6.0 (or 7.0) mm-ID ETT inside a size-4 (or -5) LMA, one solution that solves all three problems is to change the 6.0 (or 7.0) mm-ID ETT and LMA for a larger ETT by using a tube exchanger. [1]The total intubating sequence is (1) insertion of the LMA into the patient with the 6.0 (or 7.0) mm-ID ETT inserted into the size-4 (or -5) LMA so that the tip of the ETT is between the aperture bars (Figure 4, top);(2) passage of an FOB successively through a bronchoscopy elbow adapter, the 6.0 (or 7.0) mm-ID ETT, and the bowl of the 4 (or 5) LMA into the trachea (one may continuously ventilate around the FOB (4.0 mm-OD or smaller) but within the lumen of the ETT and LMA);(3) passage of the 6.0 (or 7.0) mm-ID ETT over the FOB into the trachea until the adapter of the 6.0 (or 7.0) mm-ID ETT is flush against the adapter of the LMA (Figure 4, bottom);(4) withdraw the FOB;(5) insert the tube exchanger into the 6.0 (or 7.0) mm-ID ETT to the appropriate depth;(6) remove both the 6.0 (or 7.0) mm-ID ETT and the LMA over the tube exchanger;(7) insert the larger ETT over the tube-exchanger to the desired depth; and (8) remove the tube exchanger from the larger ETT. The limitation to this solution is that a tube exchanger has its own inherent risks and limitations (failure of larger ETT to follow the tube exchanger into the trachea, barotrauma from subcarinal jet ventilation). [1].A second solution for the three problems associated with having a standard 6.0 (or 7.0) mm-ID cuffed ETT inside a size-4 (or 5) LMA is to use a large-bore LMA [66]or split-tube LMA, [67,68]both of which allow passage of a larger diameter, longer ETT and, with the split-tube LMA, removal of the LMA. Neither of these two versions of the LMA are commercially available.A third solution to the problem of having a marginally long standard 6.0 mm-ID ETT in situ is to use the extra-long (by 2–3 cm) Mallinckrodt (St. Louis, MO) microlaryngeal 6.0 mm-ID cuffed ETT. This extra-long ETT allows deeper insertion of the cuff of the 6.0 mm-ID tube but does not solve the problems of having an in situ narrow tube or removal of the LMA over this ETT, and the extra long ETT is not commonly available.A fourth solution to the problem of having a marginally long standard 6.0 (or 7.0) mm-ID ETT in situ is to deflate the cuff of the size-4 (or -5) LMA and advance the LMA and ETT together. The LMA cuff deflation maneuver will gain approximately 1.0–1.5 cm of depth but does not solve either of the other two problems.A fifth solution to the standard cuffed 6.0 (or 7.0) mm-ID ETT/size-4 (or -5) LMA combination is to jacket the proximal end of the FOB with a long hollow bougie (such as a 20-Fr jejunostomy tube). [69]The hollow bougie can be used as a tube exchanger after passage into the trachea. This technique suffers from the same limitations of a standard tube changer, and these hollow catheters are not readily available.A sixth solution for the problem of removing the LMA from a noncuffed 3.5–5.0 mm-ID ETT is to lengthen the ETT by taping, without ridges, the primary ETT to another length of the uncuffed ETT or by inserting the tip of an ETT that has an internal diameter 1 mm less than the internal diameter of the primary ETT into the proximal end of the primary ETT. [70,71].A final solution is to pass a guidewire through the suction port of the FOB into the trachea, remove both the FOB and the LMA, and pass the desired ETT over the guidewire. [72]However, the small outside diameter and flexibility of the guidewire can render it unreliable as an ETT guide.If the ETT that has been passed through the LMA is of adequate size, there is no primary or inherent reason to remove the LMA and there is reason to leave the LMA in place. At the end of the case, the patient may be extubated of the ETT while the LMA remains in situ. This strategy accomplishes two important goals: First, on extubation, a well functioning upper airway device is already in place. Second, a proven mechanism for reintubation is immediately available if necessary. If this airway management strategy is used (which may prove common), the cuff of the LMA should be deflated while the ETT is in the trachea.The laryngeal mask airway, as a ventilatory device and/or intubating conduit, can be placed into the ASA difficult airway algorithm in five places (Figure 5; compare with Figure 1):(1) on the awake intubation limb of the algorithm as a conduit for fiberoptic tracheal intubation, (2) on the nonemergency pathway of the anesthetized limb as a ventilatory mechanism by which to do the case or (3) as a conduit for fiberoptic tracheal intubation, and (4) on the emergency anesthetized pathway as both a life-saving ventilatory device and (5) as a conduit for fiberoptic tracheal intubation.There are numerous descriptions of using the LMA in properly prepared awake patients as a conduit for fiberoptic intubation (Figure 4and section IIIB). [73–84]This technique is suitable when the preoperative evaluation indicates the patient should be tracheally intubated awake for several reasons. First, even though awake tracheal intubation can be performed by many techniques in a properly prepared awake patient, including the very stimulating methods of conventional rigid laryngoscopy and bronchoscopy, insertion of an LMA is a relatively moderate stimulus that results in little hemodynamic change. [6]The relative lack of stimulation in passing an LMA reduces the amount of preparation (topicalization, sedation) that an awake patient requires. The act of swallowing facilitates the insertion of the LMA. [85]Second, once the LMA is in good position, the shaft of the LMA is usually well directed toward the larynx, and visualization of the laryngeal aperture with an FOB is easy (see above). Third, and perhaps most importantly, with the patient awake, no options are eliminated, and risk remains low even if there is difficulty in inserting the LMA or the FOB. Proper placement of the LMA into an awake patient also allows the LMA to be used as a primary means of airway control in patients whose airways are potentially difficult but who otherwise do not require endotracheal intubation. The importance of the latter consideration is increased by several descriptions of failure to correctly place and ventilate through the LMA after the induction of general anesthesia in patients with difficult airways. [86–88].In a patient undergoing general anesthesia who cannot be tracheally intubated but whose lungs can be ventilated via mask, one of the options in the ASA difficult airway algorithm is to provide anesthesia via mask. Under these circumstances, the LMA is an alternative to mask ventilation. First, the LMA may provide a better airway, with respect to ventilation and oxygenation, than a conventional mask and oropharyngeal airway system. [11]Second, use of the LMA frees the anesthesiologist's hands for other vital tasks. Third, the LMA will provide a conduit for fiberoptic intubation (see section IVC). Notable examples of use of the LMA as a ventilatory device during complex cases in patients whose trachea could not be conventionally intubated are a 15-month-old patient undergoing patent ductus arteriosus ligation [89]and an adult patient undergoing coronary artery bypass grafting. [90].One exception to the above is a patient in whom cricoid pressure must be continuously applied; cricoid pressure may interfere with correct placement of the LMA. [62]In the obstetric patient who can be ventilated via conventional mask while cricoid pressure is continuously applied, placement of the LMA would have little benefit and might induce vomiting and aspiration. [91–95]A reasonable alternative between competing concerns of continuously maintaining cricoid pressure in a patient at risk for aspiration and failure to properly insert the LMA during cricoid pressure is to momentarily release cricoid pressure as the distal tip of the LMA reaches the hypopharynx [95]; this maximizes the chance of correct LMA placement while minimizing risk of aspiration. Once the LMA is in situ, it probably does not interfere with the efficacy of cricoid pressure. [96].The LMA works just as well as a conduit for tracheal intubation (Figure 4and section IIIB) in the anesthetized patient as it does in the awake patient because the LMA prevents apposition of the tongue to the pharynx, which is the reason why other airway intubation conduits (such as the Williams, Ovassapian, and Berman airways) do not always work well in anesthetized patients. As in the awake patient, use of an FOB as an intubating stylet will guarantee a high degree of success. Because the shaft and the bowl of the LMA are usually well directed toward the laryngeal aperture, the intubation process usually takes less than 20–30 s even in anesthetized and paralyzed patients. The same reservations and concerns about passing and positioning an LMA in the presence of cricoid pressure (see section IVB) apply here. [97].Because the LMA works well as a routine airway/ventilatory device in most patients, including those with a class 3 or 4 oropharynx and/or a grade III or IV laryngoscopic view, [98,99]it is not surprising that the LMA has been found to be a life-saving emergency airway/ventilatory device in many patients whose lungs could not be ventilated using a bag and conventional mask and whose trachea could not be conventionally intubated. [100–114]These patients include obstetric patients undergoing emergency cesarean section, [100–105]neonates [106,107]and infants, [89]trauma patients with a maxillary-mandibular fracture, [108]and other adult patients. [109–113]In all of these patients, there was no known or subsequently discovered periglottic pathology (cancer, hematoma, abscess), obstruction to mask ventilation was presumably supraglottic (e.g., tongue, palate, pharyngeal apposition), and difficulty with intubation was presumably due to unfavorable anatomy (e.g., failure to align the oral, pharyngeal, and laryngeal axes). There have been no reports of failure to correctly place or ventilate through the LMA in patients in a "cannot ventilate, cannot intubate" situation; however, it is possible that this perfect record reflects a reporting bias. Consequently, in such a situation, insertion of the LMA is a reasonable maneuver to quickly try first, except when local pathology in the pharynx or larynx (see above) precludes a reasonable chance of proper placement and/or even limited gas exchange. The reasons the LMA is a first choice over transtracheal jet ventilation in the supraglottic obstruction, unfavorable natural anatomy, nonpathologic situation are that the vast majority of anesthesiologists are familiar with the LMA and the LMA is effective and, compared to transtracheal jet ventilation, has far less trauma potential. As such, the LMA should be readily available wherever a "cannot ventilate, cannot intubate" situation may arise. In the United Kingdom, 72% of obstetric anesthesiologists favor using the LMA as the first treatment option for the "cannot ventilate, cannot intubate" situation. [114]If insertion of the LMA does not affect gas exchange quickly, transtracheal jet ventilation should be immediately instituted or a surgical airway immediately created. [1,115]If the LMA does permit some gas exchange, a form of stable permanent airway should be obtained as quickly as possible (awake spontaneous ventilation, tracheal intubation (see below), surgical airway), because of the risk of aspiration as well as the possible need for high positive-pressure ventilation.If gas exchange is obtained by insertion of the LMA in a "cannot ventilate, cannot intubate" situation, precious time has been obtained to subsequently use the LMA as a conduit for tracheal intubation. The usefulness of the LMA as a conduit for tracheal intubation after a "cannot ventilate, cannot intubate" situation is the same as that after a "cannot intubate, can ventilate" situation (see section IVB). If adequate ventilation is possible through the LMA, it is probable that the bowl of the LMA surrounds the larynx and fiberoptic-guided tracheal intubation will be successful. Many of the patients cited in references 102–116 were tracheally intubated through the LMA. Thus, if the conventional intubation difficulty was simply due to unfavorable anatomy and not periglottic pathology and one can ventilate through the LMA, the LMA makes an excellent conduit to the trachea. If ventilation is poor after LMA insertion, because the LMA is not well aligned with the laryngeal aperture, or there is major periglottic pathology, fiberoptic-guided tracheal intubation may be difficult. If ventilation is poor after LMA insertion because the patient has high airway resistance and low lung compliance, fiberoptic-guided tracheal intubation likely will be successful.The ASA difficult airway algorithm has been presented to and used by the anesthesia community for approximately 5 yr and seems to be well accepted, as documented by minimal published criticism. The algorithm stresses awake intubation if difficulty with the airway is recognized, waking the patient whenever serious difficulty is encountered after the induction of general anesthesia, and always having an alternate plan(s) in place and allows many techniques to be employed based on the training, knowledge, and experience of the practitioner. The laryngeal mask airway has two major uses, as a routine airway during general anesthesia and as a conduit for tracheal intubation. As such, the laryngeal mask airway fits into the ASA difficult airway algorithm in five places: as a conduit for fiberoptic tracheal intubation in the awake and the anesthetized patient who cannot be conventionally intubated (mask ventilation may or may not be possible) and as both a nonemergency and an emergency airway in the anesthetized patient. With multiple uses and multiple places of use, the laryngeal mask airway is an important option within the ASA difficult airway algorithm. More importantly, the clinical record of LMA use in "cannot ventilate, cannot intubate" situations has been excellent, and in patients whose lungs cannot be ventilated because of supraglottic obstruction and whose trachea cannot be intubated due to unfavorable anatomy (but not periglottic pathology), the LMA should be immediately available and considered as the first treatment choice.*The Combitube (Sheridan Catheter, Argyle, NY) is a blindly inserted plastic twin-lumen, double-ballooned (pharyngeal and either esophageal or tracheal) tube. If the tube locates in the esophagus (which it does the vast majority of the time), the lungs can be ventilated via the hypopharyngeal perforations in the "esophageal lumen" between the two balloons. If the tube locates in the trachea, the lungs can be ventilated via the "tracheal lumen."