International Journal of Pediatric Otorhinolaryngology (2007) 71, 949—958

www.elsevier.com/locate/ijporl

The role of laryngeal electromyography in thediagnosis of vocal fold immobility in children

Antonio Ysunza *, Luis Landeros, Ma. Carmen Pamplona,Hector Prado, Jose Arrieta, German Fajardo

Department of Otolaryngology, Hospital Gea Gonzalez, Mexico City, Mexico

Received 27 December 2006; received in revised form 7 March 2007; accepted 9 March 2007

KEYWORDSElectromyography;Larynx;Voice;Vocal fold

Summary

Introduction: Pathology may affect the muscles that control vocal function directlyby affecting peripheral function or indirectly by affecting the central nervous system.Clinically, muscle function can be assessed by observing the movements of thestructures themselves or by recording the electrical activity of the muscles (electro-myography, EMG). Since EMG is an invasive technique, it has enjoyed limited use in thediagnosis and management of voice disorders, especially in children. Laryngeal EMGmay be helpful in those patients with voice problems of suspected neurological orneuromuscular etiology.Objective: The purpose of this paper is to study the role of laryngeal EMG in theclinical evaluation of unilateral vocal fold immobility in children.Materials andmethods: Twenty-five children with unilateral vocal fold paralysis werestudied. Twenty-five patients with vocal pathologies secondary to vocal abuse andmisuse were studied as controls. The sensitivity and specificity of the EMG as adiagnostic marker for vocal fold paralysis were obtained. Additionally, nine patientswith traumatic arytenoid dislocation were also studied. All patients were subjected tolaryngeal EMG. EMG showed a sensitivity of 100%, and specificity of 92%. Only twopatients, present with a functional voice disorder, showed abnormalities in the EMGrecordings. In the nine patients with arytenoid dislocation, EMG showed normalparameters.Conclusion: EMG seems a safe and reliable test for evaluating patients with vocal foldimmobility. Moreover, the EMG recordings were helpful in differentiating vocal foldparalysis from arytenoid dislocation. EMG can provide useful data concerning muscledenervation, and more importantly, reinnervation. Besides its utility for the diag-nosis, serial EMG can be useful for monitoring recovery and for establishing a reliableprognosis, and hence, an adequate treatment plan.# 2007 Elsevier Ireland Ltd. All rights reserved.

* Corresponding author.E-mail address: amysunza@terra.com.mx (A. Ysunza).

0165-5876/$ — see front matter # 2007 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.ijporl.2007.03.007

950 A. Ysunza et al.

1. Introduction

Lesions that affect the vagus nerve somewherealong its course from the base of the skull to thelarynx will result in a paresis or paralysis of musclesin the larynx. These peripheral lesions of the vagusare the most common cause of vocal fold paralysis.In the neck, the vagus nerve proper descends withina sheath common to the internal carotid artery andthe internal jugular vein. At the root of the neck,the recurrent laryngeal nerves are given off andpursue different courses on the two sides. The rightrecurrent laryngeal nerve bends upward behind thesubclavian artery to ascend in the tracheo-esopha-geal sulcus, whereas the left recurrent laryngealnerve passes beneath the aortic arch to attain thissulcus. The recurrent laryngeal nerves then divideinto anterior and posterior rami, which supply allthe muscles of the larynx except the cricothyroidmuscle, which is supplied by the external ramus ofthe superior laryngeal nerve. This nerve arises fromthe vagus near the nodose ganglion and divides intoa predominantly motor external ramus to the cri-cothyroid muscle [1,2].

The recurrent laryngeal nerve is susceptible toinjury throughout its intrathoracic course by severalpathologies. The left recurrent laryngeal nerve islonger than the right and is thus damaged moreoften. In up to one third of the cases of isolatedrecurrent laryngeal palsy, there is no discoverablecause. However, with the advent of modern neuroi-maging, the group of patients with idiopathic vocalfold paresis should become smaller [3].

Among the most common etiologies of vocal foldparesis in children, there are: acute externaltrauma to the neck, surgery and idiopathic etiolo-gies [4].

The gold standard for the diagnosis of vocal foldparesis is laryngoscopy. The typical laryngoscopicview in unilateral paralysis is that of one relativelyimmobile vocal fold that does not adduct duringphonation, while the unaffected fold moves to themidline. There may be some apparent movement ofthe affected vocal fold caused by movements ofother structures of by contraction of the cricothyr-oid (assuming an intact superior laryngeal nerve)[5].

Electromyography (EMG) is a procedure in whichneedle electrodes are inserted into the bellies ofselected muscles for measuring their electricalactivity. This diagnostic tool has been used routinelyfor the diagnosis of various neuromuscular and neu-rological disorders. EMG is an invasive technique andrequires expertise. Hence, it has enjoyed limiteduse in the diagnosis and management of voice dis-orders, especially in children. However, laryngeal

EMG can be helpful in those patients with voicedisorders of suspected neurological or neuromuscu-lar etiology [6].

Laryngeal EMG has proved to be a safe and effec-tive procedure in the diagnosis of laryngeal neuro-pathy [7].

Laryngeal joint injury or arytenoid dislocation isnot an uncommon complication resulting from oro-tracheal intubation trauma or acute externaltrauma to the neck. This disorder can be bestevaluated by laryngeal EMG combined with laryngo-scopic examinations and CTscans. Early diagnosis ofarytenoids dislocation is important for appropriateconservative or surgical management and betterprognosis [8].

The purpose of this paper was to study the role oflaryngeal EMG in the clinical evaluation of vocal foldimmobility in children.

2. Materials and methods

This study was carried out with patients evaluatedat the Otolaryngology Department of the HospitalGea Gonzalez in Mexico City. All patients with uni-lateral vocal fold paralysis from January 2002 toDecember 2006 were selected for the study. Duringthis period of time, 45 patients with unilateral vocalfold paralysis were evaluated. The protocol wasapproved by the Bioethics and Clinical ResearchCommittees of the Hospital Gea Gonzalez. Beforebeing recruited for the study, the protocol and theprocedures involved in the study were carefullyexplained to all the patients and their parents orlegal guardians.

The diagnostic indices of sensitivity, specificity,positive predictive value and negative predictivevalue, of EMG as a diagnostic marker for vocal foldparalysis, were obtained. An active group ofpatients affected by the disorder (vocal fold paraly-sis) and a control group of patients with symmetricalmobility of both vocal folds were studied. A goldstandard (videonasolaryngoscopy) was used fordetermining the presence or absence of the disorderin both groups of patients. The diagnostic test understudy (laryngeal EMG) was applied to both groups ofpatients [9,10].

It has been described that these values (specifi-city, sensitivity, negative and positive predictivevalues) correspond to the phase II of the evaluationfor diagnostic tests. In this phase, it is necessary tofind out the diagnostic efficacy in a sample ofpatients with the disorder, i.e. vocal fold paralysis,demonstrated by a gold standard. For this paper, aflexible videonasolaryngoscopy was considered asthe gold standard for the diagnosis of vocal fold

Role of laryngeal EMG in diagnosing vocal fold immobility in children 951

paralysis or functional voice disorder. Moreover, aCT scan of the larynx and complete clinical historywere also obtained in all patients in order to havecomplete data of all cases. The first two phases ofthe evaluation of a diagnostic test, must be com-pleted before evaluating the diagnostic marker (lar-yngeal EMG) in a more wide spectrum of patients(phases III—V) [9]. The evaluation of laryngeal EMGin these latter phases will be reported in furtherpublications.

In order to qualify for the active group of thisstudy, the patients had to fulfill the following cri-teria:

(1) A

unilateral vocal fold paralysis had to bedemonstrated using a clinical history, flexiblevideonasolaryngoscopy and a CT scan of thelarynx.

(2) T

he chronological age of the patients had to bebetween 11 and 16 years of age. This age rangewas selected in order to include only patientswithin the pediatric age groups, and also, toassure appropriate compliance during the pro-cedures.

(3) A

ll patients with associated systemic neurologi-cal disorders were excluded.

(4) A

ll patients with associated systemic neuromus-cular disorders were excluded.

A total of 25 patients were selected for the activegroup, all the parents or legal guardians agreed toparticipate in the study and signed an informedconsent form.

A total of 25 patients matched by sex, and withinthe age range of the active group were selectedfrom the patients being attended in the same periodof time at the Otolaryngology Department. Thesepatients showed various vocal pathologies second-ary to vocal abuse and misuse. In all these patients,vocal fold symmetric mobility was demonstratedwith the same procedure used in the patients fromthe active group (videonasolaryngoscopy).

Before being recruited, the protocol and theprocedures involved in the study were carefullyexplained to all these patients and their parentsor legal guardians. All 25 parents or legal guardiansagreed to participate in the study and signed aninformed consent form.

All patients from both groups were subjected tolaryngeal EMG. Before the procedure, all the man-euvers and materials were explained in detail toeach patient. It was stressed that their compliancewas necessary for accomplishing the EMG record-ings. It was explained that the procedure was goingto produce minor discomfort and they could requesta break at any point in time. After local topical

anesthesia of the skin over the larynx, bipolar nee-dle electrodes were inserted into specific muscles tomeasure their electrical activity. The cricothyroidand thyroarytenoid muscles on each side were usedfor assessing electrical activity of the larynx at rest,and during phonation. The cricothyroid was easilyaccessible since its superficial location. The cri-cothyroid was approached through the spacebetween the thyroid and cricoid cartilages, follow-ing the anatomical disposition of the thyroaryte-noid.

According to the previously reported protocolused in our center [11—15], the electrodes wereled to a Nicolet Viking System. A gain between 1000and 50,000 (60—94 dB), with filters of 20 Hz—10 kHzwere used for all recordings. A rejection mode wasused for preventing the amplification of 60-cycleelectric fields. The EMG activity was evaluated usingan audio amplifier connected to a speaker for carry-ing out the acoustic analysis of the EMG signal. Avisual monitor was also used for obtaining samples ofthe online recordings on a screen. The EMG data ofthe laryngeal muscles were acquired simultaneouslywith four EMG channels with a sampling rate of20,000 samples per second, which is sufficient forallowing the resolution of the fine details of themotor unit action potential (MUAP) waveformrecordings. A representative sample of MUAP wasrecorded with one skin penetration. All EMG wereobtained with simultaneous videonasolaryngoscopy.The objective of the utilization of videonasolaryn-goscpy was to observe movements of the vocal cordsduring phonation. The movements were correlatedwith EMG recordings.

The protocol for the EMG recordings included: (1)recording of insertional activity for 100—500 ms; (2)recording of spontaneous activity for 500 ms; (3)recording of voluntary motor activity during phona-tion (sustained/m/, and sustained/e/). For therecordings, software of quantitative EMG was usedas described previously [11—15].

All EMG recordings were analyzed independentlyby two examiners with several years of experiencein these procedures. The observers were asked toanalyze the EMG activity of each of the laryngealmuscles evaluated during phonation. A concordancewas obtained through a Kappa statistic. When theyielded Kappa index was greater than 75%, it wasconsidered as excellent concordance. The meanareas (UV/s) of the quantitative EMG recordingsduring phonation were compared by a one-wayanalysis of variance [10,16]. A blind assessmentprocedure was utilized, whereby both examinersdid not know if the laryngoscopy and EMG recordingsthat were being reviewed, corresponded to one ofthe patients from the active or control groups. The

952 A. Ysunza et al.

examiners were asked to determine if there was anormal or a denervation pattern in each of the EMGrecordings. When there were differences, each casewas jointly discussed until a consensus was reached.The indices of specificity, sensitivity, positive pre-dictive value, negative predictive value for EMG as adiagnostic test of vocal cord paralysis wereobtained. The gold standard was the videonasolar-yngoscopy.

Additionally, nine cases of children in whom thediagnosis of unilateral arytenoid joint dislocationhad been demonstrated using the same gold stan-dard as mentioned herein (videonasolaryngoscopy)were studied with laryngeal EMG. In these cases, thefindings from the CT scan (unilateral arytenoid dis-placement), and clinical history provided additionaldata for supporting the diagnosis.

Five cases of dislocation after orotracheal intu-bation, and four cases of external neck trauma wereincluded. The two examiners were also asked toreview these EMG recordings. They were kept blindwith respect to the diagnosis of these cases. As withthe rest of the EMG, they were asked to determine ifthere was a normal or a denervation pattern in eachof the recordings. A Kappa statistic was used forassessing concordance. When the yielded Kappaindex was greater than 75%, it was considered asexcellent concordance [10].

It should be pointed out that all EMG recordingsfrom the patients included in both groups (activeand control groups), as well as the patients presentwith arytenoid dislocation, were obtained with aminimum of 2 weeks, and a maximum of 5 monthsafter the onset of the disorder.

From the 25 patients diagnosed as vocal foldparalysis, included in the active group, 14 weremale, and 15 were females. Age ranged from 11to 16 years (median age = 12.5 years). The patientsincluded in the control group were matched by sex,their age also ranged from 11 to 16 years (medianage = 13.5 years).

The age of the nine patients with arytenoids jointdislocation ranged from 12 to 14 years.

3. Results

The EMG was completed in all patients withoutincident or complications. All patients experiencedminor discomfort during the study.

A U Mann—Whitney test demonstrated that therewas a non-significant difference between the med-ian age of both groups of patients (p = 0.898).

A Kappa index of 90% was found between the twoexaminers assessing the EMG recordings of thepatients from the active group (vocal fold paralysis),

and the control group (vocal pathologies secondaryto vocal abuse and misuse).

Both examiners agreed that the 25 patients withvocal fold paralysis showed a denervation pattern onthe EMG recordings of the thyroarytenoid muscleson the affected side, including the presence offibrillations potentials at rest, decreased recruit-ment during phonation, and incomplete interfer-ence pattern during phonation.

Both examiners agreed that the 25 patients withvocal cord paralysis showed normal parameters onthe EMG recordings of the thyroarytenoid muscleson the side with adequate vocal fold mobility.

Both examiners coincided that the 25 patientswith vocal cord paralysis showed normal parameterson the EMG recordings of the cricothyroid muscleson both sides.

In contrast, both examiners coincided that 23 ofthe patients from the control group showed normalparameters on the EMG recordings of both muscles(cricothyroid and thyroarytenoid) on both sides.Nonetheless, both examiners reported that twopatients from the control group showed milddecreased recruitment in the recordings of one ofthe thyroarytenoids during phonation. However, theydid not find fibrillation potentials at rest in these twocases. Moreover, the recordings of the cricothyroid inthese two patients showed normal parameters.

A one-way analysis of variance demonstrated asignificant difference (p < 0.01) in the mean area ofthe quantitative EMG recordings between theaffected side and the normal side in the group ofpatients with unilateral vocal cord paralysis.

In contrast, a non-significant difference(p > 0.05) was found when the EMG recordings ofthe patients from the control group were comparedside to side.

A significant difference (p < 0.01) was foundwhen the EMG recordings of the patients from thecontrol group (both sides) were compared with therecordings of the patients from the active group onthe affected side (vocal cord paralysis).

Finally, a non-significant difference (p > 0.05)was found when the EMG recordings of the patientsfrom the control group (both sides) were comparedwith the recordings of the patients from the activegroup on the unaffected side.

When the nine cases of arytenoid dislocation,were evaluated, both examiners coincided thatnone of these patients showed abnormalities onthe EMG recordings of both muscles.

The vocal fold paralysis, present in all patientsfrom the active group, was due to different etiol-ogies (see Table 1). Also, the types of vocal pathol-ogy, secondary to vocal abuse or misuse in patientsfrom the control group were various (see Table 2).

Role of laryngeal EMG in diagnosing vocal fold immobility in children 953

Table 1 The etiology of the vocal cord immobility inthe patients from the active group is displayed

Patient no. Etiology Affected side

1 PO; neck tumor; left Left VC paralysis2 Idiopathic Left VC paralysis3 Idiopathic Left VC paralysis4 Idiopathic Left VC paralysis5 Idiopathic Left VC paralysis6 PO; thyroid Left VC paralysis7 PO; thyroid Left VC paralysis8 PO; thyroid Left VC paralysis9 Idiopathic Right VC paralysis

10 PO; thyroid Left VC paralysis11 Idiopathic Right VC paralysis12 Idiopathic Left VC paralysis13 PO; neck tumor; left Left VC paralysis14 Idiopathic Left VC paralysis15 Idiopathic Left VC paralysis16 PO; thyroid Left VC paralysis17 Idiopathic Right VC paralysis18 Idiopathic Left VC paralysis19 Idiopathic Left VC paralysis20 Idiopathic Left VC paralysis21 Idiopathic Left VC paralysis22 Idiopathic Left VC paralysis23 PO; thorax surg. Left VC paralysis24 Idiopathic Left VC paralysis25 Idiopathic Left VC paralysis

Active, experimental group; PO, postoperative; thyroid, thyr-oid surgery; thorax surg., thorax surgery; VC, vocal cord.

Table 2 The etiology of the vocal pathologies, sec-ondary to vocal abuse and misuse in the patients fromthe control group is displayed

Patient no. Diagnosis

1 Vocal nodules2 Vocal nodules3 Edema4 Polyp (left)5 Vocal nodules6 Polyp (right)7 Polyp (left)8 Vocal nodules9 Vocal nodules

10 Vocal nodules11 Edema12 Vocal nodules13 Vocal nodules14 Polyp (right)15 Vocal nodules16 Vocal nodules17 Vocal nodules18 Edema19 Vocal nodules20 Vocal nodules21 Vocal nodules22 Vocal nodules23 Edema24 Vocal nodules25 Edema

Control group (vocal pathologies secondary to vocal abuse andmisuse).

Table 3 displays the analysis of the EMG as adiagnostic marker for vocal fold paralysis. Thevalues of true positives, true negatives, false posi-tives, false negatives, positive predictive value,negative predictive value, are included.

In Fig. 1, an example of one of the patients fromthe active group is shown. The patient was a 16-year-old male with a mass on the left lateral aspectof the neck. Patient’s voice was normal. A coronalview of a MRI displays the mass on the left side. The

Table 3 The values of sensitivity, specificity, positive pred

Test (EMG) Disorder

(+) Unilateral VC para

(+) (Denervation) 25(�) (Normal EMG parameters) 0

True positivesTrue negativesFalse positivesFalse negativesSensitivity (%)Specificity (%)Positive predictive value (%)Negative predictive value (%)

tumor was removed surgically. During the operation,both laryngeal nerves had to be elongated in orderto completely remove the mass. In the immediatepostoperative period, the patient was severelyhoarse. A videonasolaryngoscopy and a CT scandemonstrated a left vocal fold paralysis. The hys-topathological study of the removed mass, unex-pectedly reported a definitive diagnosis ofteratoma. Fig. 2 shows the EMG recording of the

ictive value and negative predictive value are displayed

lysis (�) Normal VC mobility vocal abuse—misuse

223

252320

1009292

100

954 A. Ysunza et al.

Fig. 1 MRI of one of the patients from the active group.The coronal view shows the mass on the left side of theneck. Patient’s voice was normal preoperatively.

left cricothyroid during phonation, normal biphasicmotor unit action potentials, with normal recruit-ment, and a complete interference pattern can beobserved. These data indicate that the superiorlaryngeal did not show any sign of dysfunction. Incontrast, in Fig. 3, the EMG recording of the leftthyroarytenoid at rest shows constant fibrillationpotentials. Finally, Fig. 4 shows the EMG recordingof the left thyroarytenoid during phonation, inwhich a moderate denervation pattern can be iden-tified. The presence of fibrillations and a partialdenervation pattern on the thyroarytenoid indicatea left recurrent laryngeal nerve incomplete lesion.

Fig. 2 EMG recording of the left cricothyroid during phonanormal recruitment, and a complete interference pattern ca

The patient was treated with vocal rehabilitationfor a period of 7 months with complete recovery.

Fig. 5 shows an example of one of the cases ofarytenoid dislocation. The patient was a 14-year-oldmale who suffered an external neck injury during a‘‘friendly’’ wrestling with a school peer. After thetrauma, the patient showed hoarseness. The CTscanof the larynx shows a lateral displacement of the leftarytenoid. A videonasolaryngoscopy corroboratedthe dislocation of the left arytenoid. In Figs. 6and 7, the EMG recordings of the left cricothyroid,and the left thyroarytenoid during phonation, shownormal parameters in both cases, indicating normalactivity of both, the superior laryngeal and therecurrent laryngeal nerves. The patient was treatedsuccessfully with a closed reduction.

4. Discussion

Vocal fold paralysis continues to be a dominant topicin pediatric laryngology.

Laryngeal EMG provides data concerning theactivity of laryngeal muscles. Denervation and rein-nervation patterns can be efficiently detected withthis procedure.

All the patients with unilateral vocal fold paraly-sis studied in this paper, showed a denervationpattern in the EMG recordings of the thyroarytenoid.In contrast, all these patients showed normal para-meters in the EMG recordings of the cricothyroid.These data indicate that all these patients showed adysfunction of the recurrent laryngeal nerve, andthat the superior laryngeal nerve was spared.

tion, normal biphasic motor unit action potentials, withn be observed.

Role of laryngeal EMG in diagnosing vocal fold immobility in children 955

Fig. 3 EMG recording of the left thyroarytenoid at rest. Constant fibrillation potentials can be identified.

Laryngeal EMG efficiently detected denervationin all cases of unilateral vocal cord paralysis (100%sensitivity). However, two patients with a functionalvoice disorder showed mild abnormalities in theEMG on one side. These two false positives yieldeda specificity value of 92%.

Additionally, laryngeal EMG demonstrated normalmuscle activity in all the cases of arytenoid disloca-tion studied herein.

There can be a possible explanation for the twocases of EMG abnormalities in patients in whom thediagnosis of vocal abuse and misuse had been made.First of all, it should be pointed out that all the casesin the control group (with vocal abuse and misuse)were studied with the same procedures utilized inthe patients from the active group (clinical history,videonasolaryngoscopy and CT scan). These twocases showed a polyp on one of the vocal folds.

Fig. 4 EMG recording of the left thyroarytenoid during phon

Although the polyps were not large, it is possiblethat the motion of one of the cords was somehowlimited by the polyp. This limitation could beresponsible for the abnormalities detected in theEMG. The abnormalities in these two cases were amildly decreased recruitment or motor unit actionpotentials during phonation, with no detection offibrillations or positive sharp waves at rest. In con-trast, fibrillations during the rest condition weredetected in all the cases of unilateral vocal cordparalysis. It is interesting that the statistical analy-sis showed a non-significant difference between theareas of the motor unit potentials during phonationwhen the activities from both sides were comparedin the patients of the control group, including thetwo cases mentioned herein. Nonetheless, bothexaminers agreed that there was a decreasedrecruitment, although mild, on one side, in these

ation. A moderate denervation pattern can be identified.

956 A. Ysunza et al.

Fig. 5 CT scan of the larynx. Lateral displacement anddislocation of the left arytenoid can be observed.

two cases. It is possible that the decrease wassufficient for being detected visually, but notenough to elicit a significant difference in the sta-tistical analysis.

Concerning the age of the patients studied in thispaper; as mentioned herein, we studied only ado-lescents (11—16 years of age). This age group wasselected in order to assure appropriate compliance.However, in our center, we have been able to per-form EMG in patients as young as 5 years of age withadequate compliance. Furthermore, in cases of con-genital vocal fold paralysis, we have performed EMGin patients with just a few months of age. In thesecases, the study is performed with the parent and/or nurse holding the patient firmly. We have not

Fig. 6 EMG recording of the left cricothyroid during phonanormal recruitment, and a complete interference pattern ca

found complications in more than 10 years of usingthis procedure.

In a survey of the American Bronco-Esophagolo-gical Association, the most frequent muscles uti-lized for Laryngeal EMG recordings were thethyroarytenoid (100%) and the cricothyroid (94%)[17].

In our center, in children, we use these twomuscles regularly in all cases since they reflectthe activity of the superior laryngeal nerve andthe recurrent laryngeal nerve.

It has been reported that the gold standard forthe diagnostic study of a vocal cord paralysis is theendoscopy of the larynx. In this paper, we usedflexible videonasolaryngoscopy as the gold stan-dard. Moreover, we also included a CT scan andthe clinical history in the diagnostic battery, sincethere are other reports which suggest that an imagestudy can significantly contribute to the integraldiagnostic process [5,7,18,19].

Laryngeal EMG is performed in an unblinded fash-ion in most of the reports. However, it has beendescribed that a more accurate result can beobtained when clinical information is known [17].For this paper, a blind procedure was utilized inorder to assess concordance between a pair ofexaminers. Also, the blind procedure was usefulfor a more accurate evaluation of the specificity,and sensitivity of the test (laryngeal EMG), using agroup of patients with the disease, and a group ofcontrols with symmetric mobility of vocal cords[20].

It has been reported that laryngeal EMG is aclinically valuable tool in the evaluation of laryngealdysfunction and vocal fold immobility. The basic

tion, normal biphasic motor unit action potentials, withn be observed.

Role of laryngeal EMG in diagnosing vocal fold immobility in children 957

Fig. 7 EMG recording of the left thyroarytenoid during phonation, normal biphasic motor unit action potentials, withnormal recruitment, and a complete interference pattern can be observed.

patterns of laryngeal EMG have been classified intothree different types: normal, neuropathy and myo-pathy [21]. In our paper, we included patients whocould be included in the ‘‘normal’’ and ‘‘neuropathycategories’’. Laryngeal EMG showed normal para-meters in all but two of the control patients. Elec-trophysiologic findings of denervation were found inall patients diagnosed with a laryngeal neuropathy.Furthermore, EMG findings were useful for deter-mining the site of lesion on the recurrent laryngealnerve, and for demonstrating that the superiorlaryngeal nerve had been spared.

Laryngeal joint injury or arytenoids dislocation isnot an uncommon complication resulting from intu-bation trauma or external neck trauma. It has beenreported that this condition can be best evaluatedby laryngeal EMG [21].

Early diagnosis of arytenoids dislocation is impor-tant for appropriate surgical management and bet-ter prognosis. In our group of additional patientspresent with arytenoid dislocation, we found normalEMG recordings in all cases. These findings supportthe statement that laryngeal EMG is a valuablediagnostic test that can help the clinician in thediagnosis and management of laryngeal neuromus-cular disorders [22].

It is debatable whether we should have includedthe nine patients with arytenoid dislocation into thecontrol group, or even, to discuss their results in thispaper. However, our intention was to assess theintegral validity of laryngeal EMG as a diagnostictest for vocal fold immobility.

In conclusion, from the results of this study, it isevident that EMG of laryngeal muscles is a safe andreliable procedure for assessing vocal fold paralysis

in children. Moreover, the EMG study is helpful fordifferentiating vocal fold paralysis from arytenoiddislocation. EMG can provide useful data concerninglaryngeal muscle denervation, and more impor-tantly, reinnervation. Besides its utility for the diag-nosis, serial EMG can be useful for monitoringrecovery and for establishing a reliable prognosis,and hence, an adequate treatment plan.

References

[1] J.J. Ballenger, Neurologic diseases of the larynx, in: J.J.Ballenger (Ed.), Diseases of the Nose, Throat, Ear, Head, andNeck, Lea & Febiger, Philadelphia, 1985.

[2] R. Colton, J.K. Casper, Understanding Voice Problems,Williams & Wilkins, Baltimore, 1996.

[3] P. Brazis, J. Masdeu, J. Biller, Localization in ClinicalNeurology, LippincottWilliams &Wilkins, Philadelphia, 2001.

[4] S.D. Schaefer, Laryngeal electromyography, Otolaryngol.Clin. North Am. 24 (1991) 1053—1057.

[5] P.A. Kimaid, A.N. Crespo, E.M. Quagliato, A. Wolf, M.A.Viana, L.A. Resende, Laryngeal EMG: contribution to vocalfold immobility diagnosis, Electromyogr. Clin. Neurophysiol.44 (2004) 371—414.

[6] S.S. Yin, W.W. Qui, F.J. Stucker, Major patterns of laryngealEMG and their clinical application, Laryngoscope 107 (1997)126—136.

[7] D.M. Simpson, D. Sternman, J. Graves-Wright, I. Sanders,Vocal cord paralysis: clinical and electrophysiologic fea-tures, Muscle Nerve 16 (1993) 952—957.

[8] S.S. Yin, W.W. Quiu, F.J. Stucker, Value of electromyographyin differential diagnosis of laryngeal joint injuries afterintubation, Ann. Otol. Rhinol. Laryngol. 105 (1996) 446—451.

[9] A.A. Nierenberg, A.R. Feinstein, How to evaluate a diag-nostic marker test. Lessons fro the rise and fall of thedexamethasone suppression test, JAMA 259 (1988) 1699—1702.

958 A. Ysunza et al.

[10] B. Rossner, Fundamentals of Biostatistics, Duxbury, Thomp-son Learning, Pacific Grove, California, 2000.

[11] A. Ysunza, Ma.C. Pamplona, F. Molina, E. Chacon, M. Col-lado, Velopharyngeal motion following sphincter pharyngo-plasty. A videonasopharyngoscopic and electromyographicstudy, Plast. Reconstr. Surg. 104 (1999) 905—910.

[12] A. Ysunza, Fisiologıa de musculos farıngeos posterior arestauracion quirurgica del esfınter velofarıngeo, Gac.Med. Mex. 141 (2005) 195—199.

[13] I. Trigos, A. Ysunza, D. Vargas, The Sanvenero Roselli phar-yngoplasty: an electromyographic study, Cleft Palate J. 25(1988) 385—390.

[14] E. Stalberg, Quantitative analysis of individual motorunit potentials, J. Clin. Neurophysiol. 3 (1986) 313—349.

[15] A. Ysunza, Ma.C. Pamplona, Velopharyngeal function aftertwo different types of pharyngoplasty, Int. J. Pediatr. Otor-hinolaryngol. 70 (2006) 1031—1037.

[16] A. Feinstein, Multivariate Analysis, Yale University Press,New Haven, 1997.

[17] S.L. Halum, N. Patel, T.L. Smith, S. Jaradeh, R.J. Toohill, A.L.Merati, Laryngeal electromyography for unilateral vocal foldimmobility:a surveyof theAmericanBronco-Esophagological-Association, Ann. Otol. Rhinol. Laryngol. 114 (2005) 425—428.

[18] Al. Merati, S.L. Halum, T.L. Smith, Diagnostic testing forvocal fold paralysis: a survey of practice and evidence basedmedicine review, Laryngoscope 116 (2006) 1539—1552.

[19] S.L. Ishman, S.L. Halum, N.J. Patel, J.E. Kerschner, A.L.Merati, Management of vocal paralysis: a comparison ofadult and pediatric practices, Otolaryngol. Head Neck Surg.135 (2006) 590—594.

[20] A. Feinstein, Clinical Epidemiology. The Architecture ofClinical Research, W.B. Saunders, Philadelphia, 1985.

[21] S.S. Yin, W.W. Qiu, F.J. Stucker, Major patterns of laryngealelectromyography and their clinical application, Laryngo-scope 107 (1997) 126—136.

[22] J.A. Koufman, G.N. Postma, C.S. Whang, C.J. Rees, M.R.Amin, P.C. Belafsky, et al., Diagnostic laryngeal EMG: thewake Forrest experience, Otolaryngol. Head Neck Surg. 124(2001) 603—606.