Vocalization Assessed by Electrolaryngography Is Unaffected ... : Anesthesia & Analgesia (original) (raw)
Several studies have shown changes in the electromyography of upper airway musculature and increases in airway flow resistance after the administration of topical lidocaine. These changes appear to be due to the blockade of upper airway receptors involved in reflex maintenance of airway patency.1–4 Additionally, there are case reports in which complete airway obstruction has followed simple topicalization with lidocaine.5,6 In comparison with these effects on sensory receptors, gross motor function is generally thought to be preserved. The evidence for preservation of motor function has focused mainly on power, for example, cough strength.7 There is relatively little known of the effect of local anesthesia on fine control of laryngeal motor function, with 1 published report appearing to show little pharmacological effect.8
If laryngeal motor innervation were impaired by topical local anesthesia, it would be expected that fine motor control, and consequently control of vocalization, would also be affected. Electrolaryngography is a technique used to assess vocal fold vibration in the analysis of voice disorders. The Laryngograph® (London, UK) is a system that uses a combination of acoustic and electrolaryngographic data to quantify changes in voice quality. After a pilot study using a 4% lidocaine gargle, which presumably produced pharyngeal mucosal anesthesia, there appeared to be a pharmacological effect on motor function. The present study was designed to investigate whether this effect was primarily pharmacological, or a consequence of the action of gargling itself. Additionally, if the cause was found to be pharmacological, we intended to assess whether this effect was concentration dependent.
METHODS
This prospective, crossover, randomized, placebo-controlled double-blind study was conducted after approval from the local Research Ethics Committee and the Medicines and Healthcare Regulatory Authority. The study was registered on the UK National Research Register (identification number: N0192184907).
The Laryngograph system is widely used in the analysis of voice disorders. It uses data from the variation in the electrolaryngographic signal during vocal fold vibration as well as the acoustic signal. Two Laryngograph surface electrodes placed externally over the thyroid alae (Fig. 1) measure changes in electrical conductance between their surfaces. An alternating current (AC) at 3 MHz is passed between them. This change in electrical conductance is caused by an alteration in vocal fold contact area due to vibration of the folds. An omnidirectional pressure-sensitive microphone positioned approximately 5 cm from the mouth is used to measure the acoustic variables. The Laryngograph processor analyzes the electrolaryngographic and acoustic signals. A range of variables, considered to be determinants of voice quality, can be generated from the 2 signals using Speech Studio® (London, UK) software. The average fundamental frequency (FX) corresponds to the pitch, and is calculated from the electrolaryngograph (Lx) waveform on a cycle-by-cycle basis, as opposed to Fourier analysis of an acoustic waveform (F0). “Jitter” is a measure of frequency perturbations and is also derived from the Lx waveform, while “shimmer” is a measure of amplitude perturbation derived from the acoustic waveform in the Speech Studio program.
Position of Laryngograph® detection electrodes and microphone. Written permission was given by the subject for the use of her image.
Noise observed in a sound spectrogram has been found to correlate well with the degree of perceived hoarseness and so theoretically should be a good discriminator of a pathological voice.9 Normalized noise energy (NNE) is a logarithmic ratio of noise/signal energy contained in the acoustic signal during sustained vowel phonation. NNE differs from other noise measurements in that it only measures noise in the dip areas between the harmonics in the high-frequency region (1 to 5 kHz) of the spectrogram and for short intervals. Both of these factors help reduce errors in noise estimation and improve discrimination between pathological and normal voices.10
NNE was the variable that showed least variation on repeat testing in our pilot study. The within-subject variation in our pilot study was <10%, and it also appeared to be sensitive to the gargling process. Using Altman's nomogram for within-patient studies, and a change of 10% as the difference of interest, we required 22 subjects to provide α = 0.05 and β = 0.8.11
After giving written, informed consent, 24 healthy volunteers were recruited to the study to allow for completion failure or protocol violation. Exclusion criteria were current vocal problems, pregnancy, allergy to local anesthetic, or suspected difficulty in airway management on routine airway assessment (reduced gape, neck extension, thyromental distance, and dentition). Subjects were fasted for 4 hours before and 2 hours after the procedure. They were asked to arrive at the same time of day on each occasion, and all visits were completed within a 2-month period. All recordings were made in the same sound-treated booth used for voice recording in the outpatient department. Subjects sat upright in a comfortable position.
The electrodes were calibrated, and each subject then performed 2 control vocal exercises using sustained vowel sounds at 2 pitches: a mid “ee” and a low “ee.” At the same sitting, the subject gargled with 15 mL of 1 of 3 test solutions for a period of up to 5 minutes, or until they were unable to continue, before repeating the same vocal exercises. The test solutions were prepared and labeled A, B, or C by the pharmacy department and consisted of 2% lidocaine, 4% lidocaine, or a placebo. The placebo was a pharmacy-prepared formulation consisting of water and the red dye amaranth, so that it had the same color, same consistency, and similar taste as lidocaine. All 3 solutions were aqueous in nature. A Web-based package provided by the University of Nottingham Clinical Trials Support Unit was used to produce randomization of the order in which each participant tested the 3 solutions. After the recruitment of a volunteer, the Clinical Trials Support Unit Website was accessed by 1 of the authors (M.J.M.), and a participant number was then produced. This then led to computer-based number generation, which created the order in which that volunteer tested the solutions. This order could be accessed only by a pharmacy technician through a separate Website login. Both the participant and the investigator were blinded to the test solution.
On each testing occasion, after completion of the vocal exercises, a fiberoptic nasendoscope (Olympus KeyMed ENF type P3; Southend-on-Sea, UK) was passed transnasally to assess glottic appearance. The procedure was recorded digitally onto a personal computer. The subjects were first asked to breathe normally, and then to take a deep breath, sniff, cough, and say “ee” to obtain images of the cords in various positions including maximal abduction. The volunteers were free to eat and drink 2 hours after completing the study. Volunteers returned, again fasted for 4 hours, on 2 further separate occasions at approximately the same time of day. On each occasion they gargled with one of the remaining test solutions and then repeated the same vocal exercises and fiberoptic nasendoscopy.
A blinded observer (M.J.M.) selected stable segments of sustained vowel recordings from each vocal exercise on the basis of satisfactory recording and lack of artifact. The Laryngograph software (Speech Studio) was used to analyze these segments and generate values for the vocal variables being studied (jitter, shimmer, and NNE).
The same blinded observer then reviewed the video recordings for each subject and selected images of the vocal cords at maximal abduction, which occurred during deep breathing. The angle between the vocal cords was measured at their anterior aspect using MB-Ruler® (Markus Bader Software Solutions, Iffezheim, Germany; http://www.markus-bader.de/MB-Ruler/). Images selected for measurement were taken at the same distance from the glottis as determined by image diameter using MB-Ruler (Fig. 2). The measurement of the angle between each set of vocal cords was repeated on 3 separate occasions, several days apart, to ensure accuracy, and the mean measurement was used for analysis. A separate blinded investigator (I.K.M.) then verified the accuracy of the measurements by independently measuring the angles in a random 10% sample of vocal cord images and confirming that the value lay within the range of the 3 previous measures.
Measurement of angle between vocal cords at their anterior aspect using MB-Ruler®.
One of the authors (J.A.M.), with expertise in voice disorders, and blinded to group allocation, also listened to the recorded segments to identify qualitative speech abnormality associated with laryngeal dysfunction.
The within-subject data on shimmer, jitter, and NNE from sustained vowels were analyzed using Friedman repeated-measures analysis of variance (ANOVA) for ranks. Tukey test was used for post hoc pairwise comparison of treatment and placebo groups versus control group with adjustment for multiple testing (Sigmastat V3, Systat Software Inc., San Jose, CA).
RESULTS
Four subjects were unable to complete all 3 visits at similar times of day within the time frame of the study and were withdrawn. In the remaining 20 subjects, topical laryngeal local anesthesia, performance of the vocal exercises, and fiberoptic nasendoscopy were all completed successfully. Five subjects were unable to complete 5 minutes of gargling due to coughing, at which point, adequate anesthesia was assumed to be present.
Results of vocal variables are shown in Table 1 and Figure 3. Friedman repeated-measures ANOVA performed across control, placebo, and 2 treatment allocations showed a statistically significant difference in NNE for mid “ee” vocalization (P = 0.014), as well as a tendency towards better values of shimmer and jitter in the treatment and placebo groups when compared to control. These changes were less apparent in the low “ee” group in which there was no statistically significant difference (Table 1). Inspection and post hoc pairwise testing (Tukey) of the data for the mid “ee” group shows that the difference was significant between the control group and both treatment groups, though not between control and placebo when adjusted for multiple comparisons. The difference between control and placebo was, however, in the same direction as the treatment groups. Figure 3 illustrates how the NNE became increasingly negative, indicating improved voice quality, in the placebo and treatment groups when compared to control. This pattern was repeated with the remaining variables of shimmer and jitter, both of which also tended to a show a nonsignificant but lower voice quality than in either the placebo or treatment groups (Table 1). There were no significant differences in the glottic angle (degrees [SD]) measurements between groups at maximal inspiration, which for control, placebo, 2%, and 4% groups, respectively were 57.8 (4.6), 59.7 (4.4), 58.3 (4.8), and 60.8 (5.0).
Results for Sustained Vowel Sound Vocal Exercises
Normalized noise energy between gargling groups for sustained vowel sound mid “ee.” Data are presented as median (IQR [95% CI]).
Figure 2 shows a glottic aperture image with MB-Ruler software superimposed. Blinded recordings of connected speech showed no subjective difference in voice quality, although mild dysarthric changes—for example, slurring and slowing of speech—were present.
DISCUSSION
Electrolaryngography has been used to describe and quantify changes in voice quality in a wide range of contexts, such as voice assessment after airway instrumentation postoperatively, in myasthenia gravis, and after radiotherapy.12–14 This study demonstrated changes associated with the process of gargling but showed no difference between treatment and placebo groups. Although evidence of mild dysarthria attributed to local anesthesia of tongue and oropharynx was identified, there did not appear to be an effect on voice quality as assessed by a consultant with extensive experience in voice disorders (J.A.M.).
Although we had expected the pharmacological effect to be the more important factor in any change, we included a placebo-controlled vocalization group specifically to control for any effects of the gargling process on voice. For example, state of lubrication of the vocal tract and specifically vocal folds is well recognized to be an important determinant of voice quality.15,16 The finding that changes in voice were restricted to the higher frequency after gargling is consistent with the known effects of lubrication on vocal range.15 It is possible that changes could have occurred during the period of observation, because the pre- and postgargle measurements in the pilot study were performed at a single sitting. In combination, these changes seem likely to be responsible for the difference in the results, and the finding that the values for the unlubricated control group were the only ones to differ significantly from the others. The fact that changes were detected as a consequence of the gargling process suggests that the design was powerful enough to identify a significant effect due to the local anesthetic if this had been present.
We deliberately restricted local anesthesia application to the airway at the level of the glottis and above. Authors have described using supraglottic techniques alone for tracheal intubation, and in our practice we commonly find that subglottic anesthesia is achieved by gargling.17 Supraglottic techniques might be expected to spare the recurrent laryngeal nerve, which is the primary source of motor innervation of the larynx. However, laryngeal innervation is not always consistent. Even complete recurrent laryngeal nerve palsy results in highly varied effects on voice quality,18 suggesting that motor control is variably innervated and there are published reports of airway obstruction after supraglottic topical anesthesia alone.5,6 It is likely that some techniques of topicalization of the upper airway result in more profound and more extensive local anesthesia than others, and may be more or less likely to precipitate laryngeal dysfunction and/or loss of airway patency. At least 1 study, however, has shown topicalization effects on airway physiology to be independent of minor differences in application.1 We decided not to attempt to measure whether local anesthesia had in fact been accomplished, because although endoscopic techniques (Functional Endoscopic Evaluation of Swallowing with Sensory Testing [FEEST]) are available,19 reliable results are difficult to achieve, and sensation is not tested in anesthetic practice.
Our results showed that vocal fold behavior during speech was affected by the gargling process itself, indicating that the folds were in contact with the solutions. Of note, 5 subjects failed to achieve 5 minutes of gargling because of an inability to continue gargling due to the sensation of solution in the trachea, promoting coughing. There was no evidence of respiratory harm due to aspiration in any subject.
Topical anesthesia of the airway is widely used for awake tracheal intubation, being generally regarded as having improved safety over other tracheal intubation techniques when direct laryngoscopy is expected to be difficult. This is especially the case in the compromised airway and when difficulty in bag-mask ventilation, or use of a supraglottic airway, is anticipated to be difficult or contraindicated. However, physiological studies consistently demonstrate impairment of upper airway function after application of local anesthetic, and case reports and closed claims analyses of airway complications after awake intubation attempts show that local anesthesia is not always problem free.20
Local anesthesia has repeatedly been shown to result in significant changes in upper airway physiology, including loss of “splinting” reflexes and increases in airflow resistance.1,3,4 Involuntary reflex functions preserving airway patency are dependent on intact airway receptors, and blockade of these receptors is considered the primary mechanism involved in altered upper airway physiology after local anesthesia. A predominantly afferent action of local anesthetics on upper airway mechanoreceptors would be consistent with our findings, which appear to show little effect on voluntary motor function.1 Although our findings show that vocalization and therefore fine motor control is spared in normal subjects after topical lidocaine, vocalization is largely under conscious control, and our subjects could hear their own voices, probably allowing some use of feedback and compensation, thus possibly disguising some fine motor effect.
Superior laryngeal nerve block itself has been shown to affect vocalization. This is probably via effects on the external branch of the superior laryngeal nerve, and therefore the cricothyroid muscle.21 No changes in voice were observed after topicalization in our study.
In the clinical setting of awake tracheal intubation, patients are often concomitantly treated with IV sedative and/or analgesic drugs. All these drugs have the potential for compromising airway patency, and it is likely that this tendency is exacerbated by concomitant local anesthesia; however, the interaction has not been formally studied. The case reported by Ho et al.5 nevertheless shows that obstruction can occur without sedation. Further work could investigate the effects of local anesthesia on involuntary reflex laryngeal muscle function, including in the presence of actual or simulated obstruction and increased transglottic pressure gradients. Laryngeal dysfunction could be assessed by electrolaryngography after a range of anesthetic interventions.
In conclusion, gargling can improve some measures of vocalization; however, no local anesthetic effect could be shown on these measures of fine motor laryngeal function. Our findings support the argument that known adverse effects of topical local anesthesia on the upper airway are predominantly due to blockade of afferent fibers involved in reflex maintenance of airway patency.
DISCLOSURES
Name: Melanie J. Maxwell, FRCA.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Melanie J. Maxwell has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: James D. English, FRCA.
Contribution: This author helped design the study and conduct the study.
Attestation: James D. English has seen the original study data and approved the final manuscript.
Name: Iain K. Moppett, FRCA.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Iain K. Moppett has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Julian A. McGlashan, FRCS (Otol).
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Julian A. McGlashan has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Andrew M. Norris, FRCA.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Andrew M. Norris has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
This manuscript was handled by: Sorin J. Brull, MD, FCARCSI (Hon).
REFERENCES
1. Beydon L, Lorino AM, Verra F, Labroue M, Catoire P, Lofaso F, Bonnet F. Topical upper airway anaesthesia with lidocaine increases airway resistance by impairing glottic function. Intensive Care Med 1995;21:920–6
2. Fogel RB, Malhotra A, Shea SA, Edwards JK, White DP. Reduced genioglossal activity with upper airway anesthesia in awake patients with OSA. J Appl Physiol 2000;88:1346–54
3. Kuna ST, Woodson GE, Sant'Ambrogio G. Effect of laryngeal anesthesia on pulmonary function testing in normal subjects. Am Rev Respir Dis 1988;137:656–61
4. Liistro G, Stanescu DC, Veriter C, Rodenstein DO, D'Odemont JP. Upper airway anesthesia induces airflow limitation in awake humans. Am Rev Respir Dis 1992;146:581–5
5. Ho AM, Chung DC, To EW, Karmakar MK. Total airway obstruction during local anesthesia in a non-sedated patient with a compromised airway. Can J Anaesth 2004;51:838–41
6. Shaw IC, Welchew EA, Harrison BJ, Michael S. Complete airway obstruction during awake fibreoptic intubation. Anaesthesia 1997;52:582–5
7. Mahajan RP, Murty GE, Singh P, Aitkenhead AR. Effect of topical anaesthesia on the motor performance of vocal cords as assessed by tussometry. Anaesthesia 1994;49:1028–30
8. Yang CC, Chen SH. Impact of topical anesthesia on acoustic characteristics of voice during laryngeal telescopic examination. Otolaryngol Head Neck Surg 2005;132:110–4
9. Yanagihara N, Gould WJ, Baer T. Significance of harmonic changes and noise components in hoarseness. J Speech Hear Res 1967;10:531–41
10. Kasuya H, Ogawa S, Mashima K, Ebihara S. Normalized noise energy as an acoustic measure to evaluate pathological voice. J Acoust Soc Am 1986;80:1329–34
11. Altman D. Practical Statistics for Medical Research. London: Chapman and Hall, 1991
12. Hamdan AL, Kanazi G, Rameh C, Rifai H, Sibai A. Immediate post-operative vocal changes in patients using laryngeal mask airway versus endotracheal tube. J Laryngol Otol 2008;122:829–35
13. Kazi R, Venkitaraman R, Johnson C, Prasad V, Clarke P, Newbold K, Rhys-Evans P, Nutting C, Harrington K. Prospective, longitudinal electroglottographic study of voice recovery following accelerated hypofractionated radiotherapy for T1/T2 larynx cancer. Radiother Oncol 2008;87:230–6
14. Xu W, Han D, Hou L, Hu R, Wang L. Clinical and electrophysiological characteristics of larynx in myasthenia gravis. Ann Otol Rhinol Laryngol 2009;118:656–61
15. Roh JL, Kim HS, Kim AY. The effect of acute xerostomia on vocal function. Arch Otolaryngol Head Neck Surg 2006;132:542–6
16. Sivasankar MLC. From bench to bedside: translational research on vocal fold vibration. Perspect Voice Voice Disorders 2008;18:112–8
17. Chung DC, Mainland PA, Kong AS. Anesthesia of the airway by aspiration of lidocaine. Can J Anaesth 1999;46:215–9
18. Jeannon JP, Orabi AA, Bruch GA, Abdalsalam HA, Simo R. Diagnosis of recurrent laryngeal nerve palsy after thyroidectomy: a systematic review. Int J Clin Pract 2009;63:624–9
19. Amin MR, Harris D, Cassel SG, Grimes E, Heiman-Patterson T. Sensory testing in the assessment of laryngeal sensation in patients with amyotrophic lateral sclerosis. Ann Otol Rhinol Laryngol 2006;115:528–34
20. Peterson GN, Domino KB, Caplan RA, Posner KL, Lee LA, Cheney FW. Management of the difficult airway: a closed claims analysis. Anesthesiology 2005;103:33–9
21. Roy N, Smith ME, Dromey C, Redd J, Neff S, Grennan D. Exploring the phonatory effects of external superior laryngeal nerve paralysis: an in vivo model. Laryngoscope 2009;119:816–26
© 2012 International Anesthesia Research Society