Coughing induced by an endotracheal tube can complicate emergence from general anesthesia. Irritant or stretch stimuli in the trachea caused by the tube and its cuff are the presumed mechanisms. Rapidly acting receptors are found throughout the trachea and are primarily superficial (1). They are thought to be the irritant receptors involved in the cough reflex (2). These nociceptive stimuli can be blocked by topically applied anesthetics (3).
Coughing during emergence can result in hypertension, tachycardia, increased intraocular and intracranial pressure, myocardial ischemia, bronchospasm, and surgical bleeding (4–6). This can be of particular relevance in neurosurgical, ophthalmic, and vascular procedures. Maneuvers to reduce coughing include the administration of IV or topically applied local anesthetics (7–11). The administration of IV narcotics or tracheal extubation in a deep plane of anesthesia are alternatives; however, in many cases they are undesirable. The benefit of topically applied drugs before tracheal intubation is limited to a short time postapplication, as they are absorbed through the tracheobronchial mucosa (12). To apply an anesthetic effect to the mucosa over a longer time, an alternative method of topical application is required.
Gonzalez et al. (4) showed a reduced incidence of coughing using a modified endotracheal tube with a specially designed port to allow a local anesthetic to be applied to the mucosa below the endotracheal cuff. However, most stimulation may, in fact, come from the cuff itself, because it is in closest contact with the tracheal mucosa. In addition, there is a requirement for a specialized endotracheal tube to facilitate the technique. These factors may serve to limit the effectiveness of this technique. The cuff of an endotracheal tube is permeable to local anesthetics, including lidocaine. Sconzo et al. (13) demonstrated that 4% lidocaine placed in the cuff of an endotracheal tube, diffused across the cuff membrane. They concluded that the cuff could act as a potential reservoir for a local anesthetic, allowing diffusion and subsequent anesthesia of the underlying mucosa. The quantity of lidocaine that passed through the cuff was found to be dependent on the concentration of the lidocaine inserted into it and on time. Topical anesthetics applied in this manner might represent a novel technique to reduce adverse emergence phenomena, in particular, postextubation coughing. We therefore, investigated the potential benefits in a clinical setting of inflating the cuff of a standard endotracheal tube with 4% lidocaine, rather than air.
After obtaining approval of the hospital ethics committee, and written informed consent, we studied 63 ASA physical status I and II patients in a prospective, randomized, double-blinded trial. All patients were >18 yr of age undergoing elective orthopedic, urological, plastic, or general surgery. Certain specific exclusion criteria were applied, namely: increased intracranial pressure, active upper respiratory tract infection, history of laryngeal or tracheal surgery or pathology, risk of aspiration of gastric contents as deemed by the attending anesthesiologist, asthma, or any airway condition resulting in an ASA physical status greater than II.
Patients were randomized into one of three groups depending on whether air, saline, or 4% lidocaine was used to inflate the cuff of the endotracheal tube: Group C, the control group, in which air was used: Group S, in which normal saline was used: and Group L, in which lidocaine 4% was used.
Preoperative Management and Anesthesia
All patients were visited the evening before surgery, informed consent was obtained, and 10 mg oral diazepam was prescribed as a premedicant. This was administered 1 h before surgery. At the induction of anesthesia, patients breathed 100% oxygen via a face mask and then, were anesthetized according to a standard protocol. Patients received 1.5 μg/kg fentanyl, 2.5 mg/kg propofol, and 0.1 mg/kg vecuronium to facilitate tracheal intubation.
Laryngoscopy was performed and the trachea was intubated with a standard cuffed endotracheal tube (Mallinkrodt, Athlone, Ireland). Tubes of 7.5 mm diameter were used for women and 8.5 mm diameter for men. The endotracheal tube cuff was inflated, in accordance with the randomization, until such time as there was no air leak around the tube when positive pressure was administered to 20 cm H2O.
Anesthesia was maintained with 1–1.5 μg · kg−1 · hr−1 fentanyl, isoflurane 1.2% to 1.5%, and nitrous oxide/oxygen (65:35 ratio). Vecuronium was administered to maintain the ulnar nerve train-of-four at 0–3 of four twitches. The lungs were mechanically ventilated with tidal volumes of 8–10 mL/kg to maintain end-tidal CO2 concentration at 30–35 mm Hg. Esophageal temperature was maintained at 35–36°C by using warming equipment.
Anesthesia was maintained, as previously mentioned, until the last suture was placed or a cast applied. Neuromuscular blockade was then antagonized with neostigmine 0.04 mg/kg and glycopyrrolate 0.007 mg/kg and the pharynx was gently suctioned under direct vision. After reversal of neuromuscular block, isoflurane and nitrous oxide were discontinued and 100% oxygen given. Mechanical ventilation was maintained until swallowing or spontaneous respiration began, and then, converted to assisted manual ventilation. Extubation was performed when all of the following criteria were met: 1) full reversal of neuromuscular block (ulnar nerve T4/T1 ratio 1:1, with sustained tetanus at 50 Hz for 5 s and no fade); 2) spontaneous ventilation; and 3) the ability to follow verbal commands (eye opening or hand grip) or demonstrate purposeful unilateral movement (attempting self-extubation).
Assessment of Emergence
An independent, blinded observer (anesthesiologist uninvolved with the case) assessed the patient during emergence for hemodynamic indices, including heart rate and blood pressure, arterial oxygen saturation, end-tidal isoflurane concentration, and the incidence of coughing. A cough was recorded as either having occurred or not having occurred during each of the postextubation time periods.
Withdrawal of Patients from the Study
If surgery was less than one hour, patients were excluded from further study. This was done to ensure sufficient time had elapsed from the time of intubation for diffusion of lidocaine to occur.
Before beginning the study, a sample-size calculation was performed. Our sample-size calculations were based on an incidence of coughing on extubation of 96%, as seen in the control group in the study by Gonzalez et al. (4). We estimated that a decrease in the incidence of coughing on extubation with intracuff lidocaine of 35% would be of clinical importance. Based on these estimates, we calculated a sample size that would permit a type I error rate of two-tailed α = 0.05 with a type II error rate of β = 0.20, i.e., power equal to 0.80. This required recruitment of 21 patients in each group, and 63 patients were, therefore, required.
Demographic data were analyzed by using Student’s t-test. Hemodynamic data (mean ± sd) were analyzed by using analysis of variance and the incidence of coughing was compared by using the χ2 test for multiple variables. Continuous data are presented as means ± sd; categorical data are presented as frequencies. P < 0.05 was considered to be statistically significant.
A total of 63 patients were entered into the study. They were equally distributed among the groups and the three groups were demographically comparable (Table 1). There were equal numbers of cigarette smokers in each group. Three patients were excluded from both Group C and Group S, because it was later determined that their surgery lasted less than 1 h. Surgery for all patients in Group L lasted longer than 1 h and, therefore, all of these patients included.
The incidence of coughing during the time period 0 to 2 min was 38% and 44% for air and saline, respectively, whereas in the lidocaine group, the incidence of coughing was 16% (P = 0.18) (Figure 1). From 2 to 4 min, the incidence of coughing in the air group was 38%, whereas the incidence of coughing was comparable in the lidocaine and saline group, 11% and 11.1%, respectively (P = 0.055). From 4 to 8 min there were no recorded incidences of coughing in the lidocaine group. This compared with an incidence of 34% with air and 15% with saline, indicating a statistically significant difference between the groups with P < 0.05 (Figure 1). During the remaining two time periods of 15–25 min and 25–50 min, the incidence of coughing decreased and became comparable for air, saline and lidocaine (P values of 0.55 and 0.71, respectively).
There was no statistical difference in the incidence of laryngospasm and bronchospasm among the three groups (Table 1). The emergent hemodynamic and oxygenation saturation data were similar for all three groups. The tracheas of most patients were extubated at end-tidal isoflurane concentrations of 0.1%. Four patients each, in the air and saline groups were extubated at 0.2% isoflurane, whereas in the lidocaine group, eight patients were extubated at 0.1% isoflurane. This difference was not statistically significant.
A technique that would allow patients emerging from anesthesia to tolerate an endotracheal tube, while also affording airway protection with intact supraglottic reflexes, would be desirable in selected groups. Previous work has shown that topical lidocaine administered down the spray channel of a modified endotracheal tube reduced the incidence of coughing when compared with either a similar dose of IV lidocaine or control subjects (4). However, even direct topical administration was not completely effective in controlling the cough reflex. One of the difficulties cited by Gonzalez et al. (4) was the fact that the tracheal mucosa in direct contact with the tube cuff was effectively shielded from exposure to the administered lidocaine. Other problems included possible mucosal stimulation from the lidocaine administration.
The basis of our study was that lidocaine inserted into the endotracheal cuff might cause anesthesia of the trachea by diffusing across the polyvinyl chloride membrane of which the cuff is composed. Anesthesia should be confined to the mucosa in contact with the cuff, thus overcoming the difficulties experienced by Gonzalez et al.(4). In addition, the protective cough reflexes above the tube cuff and of the vocal cords should remain intact.
In addition to a lidocaine group, we included a saline group, as it was felt that some differences in variables could be attributed to the presence of a liquid rather than a gas in the endotracheal tube cuff. This has particular relevance in the presence of nitrous oxide, which was used in all patients as part of anesthesia maintenance, because nitrous oxide is known to increase the cuff volume when using air as the inflating medium (14,15). Expansion of the air-filled cuff, because of diffusion of nitrous oxide into it, could result in increased receptor stimulation in the tracheal mucosa, and thus, increase emergence and extubation phenomena. Although our protocol accounted for this possibility, there was no difference in any measured variables between the saline and air groups.
We believed the preextubation end-tidal isoflurane concentrations would be smaller in the lidocaine group because these patients would tolerate the presence of the endotracheal tube to lighter levels of anesthesia. We found no difference in end-tidal isoflurane concentrations at the time of extubation among the three groups.
Endotracheal tubes are constructed from polyvinyl chloride, which is a primarily hydrophobic plastic. The thin polyvinyl chloride membrane, which constitutes the tube cuff, allows simple diffusion of lidocaine across it. Because the coefficient of diffusion and the thickness of the material can be assumed to be standard across a specific range of endotracheal tubes, the limiting factors are lidocaine concentration and time. In vitro studies examining this diffusion process have reported an increased transfer of lidocaine after 60 minutes using the 4% formulation (13). In this study, prefilling the tube cuff with lidocaine for a fixed duration before the test period resulted in a slightly increased diffusion across the cuff. We felt that prefilling the cuffs was impractical on a busy operating schedule because of the unpredictability of the time of the induction of anesthesia for each patient.
The toxicity of local anesthetics must be considered regardless of the route of the administration. In this regard, our concerns were twofold, the risks of systemic absorption and the consequences of cuff damage with subsequent leakage of 4% lidocaine or saline into the bronchial tree. Although 40 mg/mL lidocaine (4%) was used, the mean volume used per endotracheal tube was 6.1 mL ± 0.9 mL (sd) (244 mg ± 36 mg). This is considerably less than the amount of lidocaine used in a study by Sutherland et al. (16), in which a fixed dose of 370 mg of lidocaine was used in 21 adult patients to topically anesthetize the airway for fiberoptic bronchoscopy and no incidence of toxic plasma concentrations of lidocaine was recorded. Another study by Efthimiou et al. (17) with 41 patients undergoing fiberoptic bronchoscopy, using average doses of 9.3 mg/kg of lidocaine, recorded only two patients in which plasma levels exceeded the toxic levels (5.0 μg/mL) and no complications were observed. All tube cuffs were intact postextubation.
Absorption through the cuff is time-dependent (13) and thus, we presume plasma levels would rise more slowly than they would after direct topical application, as occurs during fiberoptic bronchoscopy, thus reducing the risk of systemic toxicity. A study by Prengel et al. (12) set out to establish the pharmacokinetics of lidocaine administered directly via an endotracheal tube. They instilled lidocaine 2 mg/kg via an endotracheal tube and found plasma lidocaine levels of 1.4, 0.85, and 0.47 μg/mL at 20, 60, and 120 min, respectively. However, the plasma level required to effectively suppress coughing is larger (3 μg/mL) (8), and this is why we consider the use of the cuff as a “reservoir” to be so attractive.
There was no difference in the incidence of laryngospasm or bronchospasm between the groups. However, very large patient numbers would be required to show a difference, as these are relatively uncommon phenomena.
In conclusion, our study demonstrated a reduced incidence of coughing in the lidocaine group for the time period of 4–8 min. There was a trend toward a reduced incidence of coughing after extubation, in which 4% lidocaine, rather than air, was used to inflate the tube cuff for the other measured time periods up to 15 min postextubation. Using lidocaine instead of air is a relatively easy and safe practice. We suggest that further refinement in cuff membrane and local anesthetics may make this a simple technique to reduce postextubation coughing.
The authors would like to thank Dr. Dominic Harmon, MD, FFARCSI, for his assistance with the statistical analysis of this study.
1. Sant’ambrogio G, Remmers JE, deGroot WJ, et al. Localization of rapidly adapting receptors in the trachea and main stem bronchus of the dog. Respir Physiol 1978; 33:359–66.
2. Widdicombe JG. Respiratory reflexes. In: Handbook of physiology. Baltimore: Williams & Wilkins, 1964: 585–630.
3. Camporesi EM, Mortola JP, Sant’ambrogio F, Sant’ambrogio G. Topical anesthesia of tracheal receptors. J Appl Physiol 1979; 47:1123–6.
4. Gonzalez RM, Bjerke RJ, Drobycki T, et al. Prevention of endotracheal tube induced coughing during emergence from general anesthesia. Anesth Analg 1994; 79:792–5.
5. Bidwai AV, Bidwai VA, Rogers CR, Stanley TH. Blood-pressure and pulse-rate responses to endotracheal extubation with and without prior injection of lidocaine. Anesthesiology 1979; 51:171–3.
6. Leech P, Barker J, Fitch W. Changes in intracranial pressure and systemic arterial pressure during the termination of anaesthesia. Br J Anaesth 1974; 46:315–6.
7. Steinhaus JE, Gaskin I. A study of intravenous lidocaine as a suppressant of cough reflex. Anesthesiology 1963; 24:285–90.
8. Yukioka H, Yoshimoto N, Nishimura K, Fujimori M. Intravenous lidocaine as a suppressant of coughing during tracheal intubation. Anesth Analg 1985; 64:1189–92.
9. Stoelting RK. Blood pressure and heart rate changes during short duration laryngoscopy for tracheal intubation: influence of viscous or intravenous lidocaine. Anesth Analg 1978; 57:197–9.
10. Stoelting RK. Circulatory changes during direct laryngoscopy and tracheal intubation: influence of duration of laryngoscopy with or without prior lidocaine. Anesthesiology 1977; 47:381–3.
11. Hamill JF, Bedford RF, Weaver DC, Colohan AR. Lidocaine before endotracheal intubation: intravenous or laryngotracheal? Anesthesiology 1981; 55:578–81.
12. Prengel AW, Lindner KH, Hahnel JH, Georgieff M. Pharmacokinetics and technique of endotracheal and deep endobronchial lidocaine administration. Anesth Analg 1993; 77:985–9.
13. Sconzo JM, Moscicki JC, DiFazio CA. In vitro
diffusion of lidocaine across endotracheal tube cuffs. Reg Anesth 1990; 15:37–40.
14. Stanley TH, Kawamura R, Graves C. The effects of nitrous oxide on the volume and pressure of endotracheal tube cuffs. Anesthesiology 1974; 41:256–62.
15. Revenas B, Lindholm CE. Pressure and volume changes in tracheal tube cuffs during anesthesia. Acta Anaesthesiol Scand 1976; 20:321–6.
16. Sutherland AD, Santamaria JD, Nana A. Patient comfort and plasma lignocaine concentrations during fibreoptic bronchoscopy. Anaesth Intensive Care 1985; 13:370–4.
17. Efthimiou J, Higenbottam T, Holt D, Cochrane GM. Plasma concentrations of lignocaine during fibreoptic bronchoscopy. Thorax 1982; 37:68–71.