Awake insertion of the laryngeal mask airway using topical lidocaine and intravenous remifentanil*


  • *

    Presented in part to the Anaesthetic Research Society, Sheffield; July 2005.

Dr M. C. Lee E-mail:


We assessed the use of intravenous remifentanil for the insertion of the laryngeal mask airway in 10 healthy awake volunteers, a technique primarily developed to facilitate functional magnetic resonance imaging studies of anaesthesia. Each volunteer received 200 μg glycopyrronium intravenously. Topical airway anaesthesia was effected by 4 ml nebulised lidocaine 4%, followed by 12 sprays of lidocaine 10%. Remifentanil was subsequently infused to achieve an initial target effect-site concentration of 2−1; increments of 1−1 were allowed with the maximum effect-site concentration limited to 6−1. Insertion of the laryngeal mask airway was successful on the first attempt in all cases. The median (IQR [range]) target effect-site remifentanil concentration at insertion was 2.5 (2–3 [2–4])−1. All volunteers were co-operative during the procedure and only one reported discomfort. Sore throat was a complication in all volunteers. We conclude that the technique allows successful insertion of the laryngeal mask airway in healthy awake volunteers under conditions that were safe and reproducible.

Functional magnetic resonance imaging is a non-invasive technique increasingly used to explore the neural correlates involved in altered consciousness induced by sedative and hypnotic agents. Anaesthetic concentrations associated with sedation or unconsciousness have been investigated [1]. However, it is difficult to study the transition from the awake to the unconsciousness state safely because airway patency without manual support or artificial means in the anaesthetised subject within the confines of the scanner cannot be guaranteed. As access to the subject during scanning is limited, the airway must be secured outside the scanner while the subject is conscious.

The technique for awake fibreoptic tracheal intubation is well known, but we chose to evaluate the laryngeal mask airway because insertion is technically simpler, less invasive and more likely to be accepted by volunteers. Previous studies suggest that the laryngeal mask airway is well tolerated with topical anaesthesia of the oropharynx alone, but its initial insertion requires further suppression of airway reflexes, usually accomplished through the use of intravenous midazolam or fentanyl [2–5]. However, the residual effects of midazolam or fentanyl are likely to alter neurophysiological and behavioural responses and confound the study of subsequently administered hypnotic agents. This consideration led us to investigate the possible use of remifentanil in this setting as termination of its intravenous infusion after the insertion of the laryngeal mask airway results in a rapid and predictable decrease of its plasma concentration and offset of its effects.

Being unaware of any published report describing topical airway anaesthesia supplemented by remifentanil, we sought to assess formally the use of a target controlled infusion of remifentanil with topical lidocaine for insertion and tolerance of the laryngeal mask airway in awake volunteers.


With Local Research Ethics Committee approval and written, informed consent, we studied 10 healthy, non-pregnant, unpaid adult volunteers of ASA physical status 1. All volunteers were anaesthetists working at Addenbrooke's Hospital. To avoid any suggestion of coercion, consultant anaesthetists were not allowed to approach training grade anaesthetists. Exclusion criteria included a history of drug or alcohol misuse, psychiatric problems and allergy to lidocaine or remifentanil.

All volunteers fasted for at least 6 h before the study, which was conducted in an anaesthetic room adjacent to the operating theatre. On arrival in the anaesthetic room, the volunteer was placed semisupine on a trolley. Intravenous access was secured and monitoring (comprising pulse oximetry, non-invasive arterial BP, ECG and Bispectral Index (Aspect Medical Systems, Newton, MA)) was commenced. The Ramsay sedation score was used to assess sedation [6].

Intravenous glycopyrronium (200 μg) was used as an antisialagogue to enhance effectiveness of the local anaesthetic [7], followed by 4 ml lidocaine 4%, delivered via a nebuliser mask. Once nebulisation was complete, an intravenous remifentanil infusion was administered using a Base Primea Target Controlled Infusion system (Fresenius Vial, Brezins, France) programmed with the Minto pharmacokinetic dataset [8]. The computerised system calculated the infusion rates required to achieve and maintain the target effect-site concentration of remifentanil, and provided a continuous read-out of the estimated blood and effect-site concentrations during and after termination of the infusion. The initial target effect-site remifentanil concentration used was 2−1. This was derived from a remifentanil infusion regimen recommended for tracheal intubation in awake patients premedicated with midazolam [9]. Using a commercially available pharmacokinetic simulation software (TIVATrainer©, F. Engbers, Leiden, Netherlands) which incorporates the Minto model [8], we found that the recommended infusion (0.75 μ−1.min−1 bolus over 30 s followed by continuous infusion of 0.075 μ−1.min−1) resulted in the equilibration of blood and effect-site concentrations of 2−1 after 8 min. We used this as our initial target effect-site concentration even though insertion of the laryngeal mask airway is less invasive than tracheal intubation. This was because our subjects were healthy volunteers who did not receive any sedative premedication.

The oropharynx was subsequently sprayed with lidocaine 10% (AstraZeneca, Macclesfield, UK). Each actuation delivers approximately 10 mg lidocaine. A total of 12 sprays were used, with the final sprays being directed caudally in the oropharynx. All volunteers were closely observed for symptoms or signs of lidocaine toxicity. The adequacy of airway preparation was assessed using the absence of gag reflex after topical lidocaine was administered and with the estimated concentrations of remifentanil in the blood and effect-site at equilibrium. If airway preparation was deemed inadequate, the target effect-site remifentanil concentration was increased by 1−1. The airway was reassessed once the estimated concentrations of remifentanil in the blood and effect-site reached equilibrium at the increased target effect-site concentration. This process was repeated until airway preparation was deemed adequate or the maximum allowed target effect-site concentration of 6−1 was reached. A ClassicTM laryngeal mask airway (Intavent Ltd, Maidenhead, UK), size 3 for females and size 4 for males, was then gently inserted with the cuff semi-inflated. An unsuccessful attempt at insertion was defined as coughing or gagging during or immediately after insertion. In the event of an unsuccessful attempt at insertion, the target effect site remifentanil concentration was increased by 1−1 and insertion re-attempted only after the estimated concentrations of remifentanil in the blood and effect site reached equilibrium at the increased target effect-site concentration. A maximum of two re-attempts were allowed. No re-attempt was allowed after the maximum effect-site concentration of 6−1 was reached.

Fibreoptic laryngoscopy was performed via the laryngeal mask airway after successful insertion to assess its position. The target effect-site concentration of remifentanil required for successful insertion was maintained throughout fibreoptic laryngoscopy and the infusion was terminated after fibreoptic laryngoscopy. The volunteers were instructed to remove the laryngeal mask airway if discomfort or dyspnoea was experienced at any point during the procedure. The laryngeal mask airway was to be removed by the investigators if there was any sign of airway obstruction or after target effect-site remifentanil concentrations reached zero. The effect-site remifentanil concentrations were recorded at removal of the laryngeal mask airway.

All volunteers were asked to complete a questionnaire adapted from a local fibreoptic endoscopy training course [10] before discharge. The experience during administration of local anaesthetic through a mask, first spray of local anaesthetic to the back of the throat, final spray of local anaesthetic to the back of the throat, insertion of the laryngeal mask airway, inflation of the cuff, fibreoptic laryngoscopy and removal of the laryngeal mask airway was rated using the scale shown in Table 1. Finally, all volunteers were contacted a day afterwards to detect late complications.

Table 1.  Descriptive scale used to rate each stage of the procedure for inserting the laryngeal mask airway in awake volunteers.
Was it painful?AgonyPainfulDiscomfortComfortableNumbed
Were you anxious?Nerve rackingAnxiousSlight concernRelaxedEnjoyable
Did you cough/gag?RetchingSevere discomfortUncomfortableComfortableNumbed


The volunteers' characteristics are summarised in Table 2. Insertion of the laryngeal mask airway was successful at the first attempt in all volunteers. Median (IQR [range]) time from the start of the remifentanil infusion to successful insertion was 9 (6–13 [6–15]) min, and from discontinuation of the remifentanil infusion to voluntary removal of the laryngeal mask airway was 12 (9–18 [4–19]) min. Effect-site remifentanil concentrations at insertion and removal of the laryngeal mask airway for each volunteer are shown in Fig. 1. The effect-site remifentanil concentration was 2.5 (2–3 [2–4])−1 at successful insertion and 0.5 (0.4–0.6 [0.3–1.4])−1 at voluntary removal. The total dose of remifentanil used was 1.7 (1.2–2.0 [1.1–3.1]) μ−1.

Table 2.  Characteristics of volunteers having the laryngeal mask airway inserted awake. Values are median (range) or number (proportion).
Age; years34 (31–41)
Gender; M : F8 : 2
Weight; kg75 (60–90)
Height; m1.77 (1.66–1.88)
Body mass index; kg.m−224.5 (21.2–26.0)
Figure 1.

Estimated effect-site remifentanil concentrations at insertion (▪) and removal (bsl00066) of the laryngeal mask airway in 10 awake volunteers.

Fibreoptic laryngoscopy was easily performed in all 10 volunteers. The endoscopist (HLS) was experienced and avoided contact of the fiberscope with the mucosae of the laryngeal inlet. The laryngeal mask airway's aperture bars, the epiglottis and the laryngeal inlet were readily identified as the endoscope was advanced. The laryngeal inlet appeared obscured by the epiglottis to a varying extent, but in all cases the vocal cords were immediately seen once the endoscope passed between the aperture bars into the laryngeal inlet. No volunteers experienced dyspnoea or showed signs of airway obstruction during the procedure.

All volunteers remained calm and co-operative throughout the procedure, which corresponded with a Ramsay Score of 2. The Bispectral Index was ≥ 94 at all times in all subjects. Cardiovascular variables are summarised in Table 3. No volunteer experienced symptoms or had signs of lidocaine toxicity. Peripheral oxygen saturation (without supplementary oxygen) was ≥ 96% throughout the study in all but one volunteer, who had a brief decrease of oxygen saturation to 80% during fibreoptic laryngoscopy. The target effect-site remifentanil concentration at that time was 4−1, and the oxygen saturation rapidly improved when he was prompted to breathe. No other therapeutic intervention was required during the study.

Table 3.  Baseline, peak and trough systolic arterial blood pressures and heart rates of volunteers during awake insertion of the laryngeal mask airway. Values are mean (SD).
Systolic pressure; mmHg130 (10.5)146 (17.1)128 (13.7)
Heart rate; min−172 (9.8)76 (10.2)63 (8.6)

Nine volunteers rated all stages of the procedure as comfortable or numbed. One volunteer experienced discomfort during insertion of the laryngeal mask airway but rated its continued presence and all other stages as comfortable. However, all volunteers complained of a moderate to severe sore throat that developed after several hours and persisted for up to 30 h after the procedure.


The laryngeal mask airway has been successfully inserted for various clinical indications in patients who were awake. Case reports or series published thus far invariably report the use topical lidocaine supplemented by intravenous fentanyl, midazolam or propofol for this purpose [2–4]. We chose to assess supplementation by remifentanil during insertion because its pharmacokinetic and pharmacodynamic properties minimise confounding drug effects on our proposed functional imaging studies of anaesthetic drug action. Previous studies have shown that the Bispectral Index is insensitive to remifentanil [11]. The Bispectral Index was essentially unchanged, as were the Ramsay sedation scores, at the concentrations of remifentanil used in our study.

The cuff of the laryngeal mask airway was semi-inflated as this was previously reported to aid insertion in awake subjects [12]. Although the current recommendation for size of laryngeal mask airway is 4 for females and 5 for males [13], a size smaller was used in this study. This was based on published experience from the largest case series of successful insertions in awake patients [3].

Fiberoptic laryngoscopy was well tolerated by all volunteers. The fiberoptic view of the laryngeal inlet was obscured by the epiglottis to a varying extent between volunteers but none experienced dyspnoea or had signs of airway obstruction. This is consistent with published data demonstrating poor correlation between these views and clinical signs of poor anatomical placement [14]. The presence of the laryngeal mask airway was well accepted even after remifentanil concentrations fell to low levels. However, we cannot exclude the contributory effect of topical lidocaine anaesthesia as the sensation in the airway after removal of the airway was not assessed. We used 280 mg lidocaine during airway preparation for every volunteer, which resulted in a topical lidocaine dose of less than 4−1 in most volunteers. Absorption of nebulised lidocaine is approximately 25%[15], and a variable amount of sprayed lidocaine (which has a high first-pass effect) is lost through swallowing. The maximum topical lidocaine dose of 8.2−1 has been recommended by the British Thoracic Society [16]. It is thus not surprising that no symptoms or signs of lidocaine toxicity were observed.

Overall, every stage of the procedure was well accepted by the volunteers. The brief episode of oxygen desaturation during fibreoptic laryngoscopy in one volunteer may have been due to breath holding or hypoventilation as the saturation improved when he was prompted to breathe. The complaint of sore throat in every volunteer a day after the procedure was unexpected as the incidence of postoperative sore throat with the laryngeal mask airway ranges from 5.8% to 34%[17]. A recently published study has implicated the additives found in the 10% lidocaine spray as a possible cause of sore throat in our study [18].