7/27/2019 00000539-201312000-00014 (1)

1/5

December 2013 Volume 117 Number 6 www.anesthesia-analgesia.org 1347

Copyright 2013 International Anesthesia Research Society

DOI: 10.1213/ANE.0b013e3182a8c734

Suboptimal management of the difficult airway is one ofthe main causes of anesthesia-related mortality.1Thus,accurate prediction of difficult laryngoscopy (Cormack

and Lehane [CL] grades III and IV) is very important inanesthesia practice.2The incidence of difficult laryngoscopyis reported in the range of 1.5% to 20%.37Prediction of dif-ficult laryngoscopy has been the subject of studies usingsingle anatomical landmarks and multifactorial indexes.813

While multifactorial indexes slightly outperform singlemeasures, no anatomical landmark alone has been reportedto have acceptable accuracy for prediction of difficult laryn-goscopy. Hence, identifying a single reliable predictor of dif-ficult intubation would be valuable. We hypothesized thatthe height between the anterior borders of the mentum andthyroid cartilage, measured while the patient is lying supinewith her/his mouth closed (which we termed thyromentalheight test [TMHT]), could be regarded as a predictor fordifficult laryngoscopy. We assumed that it could be a surro-gate for the following frequently cited anthropometric mea-sures: (1) amount of mandibular protrusion; (2) dimensionsof submandibular space; and (3) anterior position of the lar-ynx. A pilot study (data not shown) suggested that there is

a close association between small thyromental height andoccurrence of difficult laryngoscopy. Thus, we designed this

observational prospective study to determine whether wecould rely on it as a predictor of difficult laryngoscopy.

METHODSAfter obtaining the Ethics Committee approval, 314 con-secutive male and female patients aged 16 years requir-ing general anesthesia and elective direct laryngoscopy

between September 2011 and February 2012 were invited

to participate. Patients who needed awake intubationand patients who underwent emergency operations wereexcluded.

Considering all investigations and measurements wereperformed in a totally noninvasive manner, there was noneed for informed written patient consent. Oral patientconsent was obtained during routine preoperative physicalexaminations.

The following 4 predictive test measurements wereobtained in the preoperative clinic by a member of theresearch team who was not involved in assessment oflaryngoscopy:

1. Modified Mallampati test (MMT): the oropharyngeal

view was assessed using a modified Mallampati clas-sification while patients were seated, with mouthmaximally opened, tongue protruded, and withoutphonation.14Classes 1 and 2 were considered predic-tive of easy laryngoscopy, and classes 3 and 4 wereconsidered predictive of difficult laryngoscopy.

2. Thyromental distance (TMD): the distance in centi-meters between the thyroid prominence and the mostanterior part of the mental prominence of the mandi-

ble, with the head fully extended, was measured witha ruler.7A TMD 6.5 cm was considered predictive ofdifficult laryngoscopy.15

3. Sternomental distance (SMD): the distance in centime-

ters between the superior border of the manubrium

BACKGROUND:The incidence of difficult laryngoscopy is reported in the range of 1.5% to 20%.We hypothesized that there is a close association between the occurrence of difficult laryngos-copy and the height between the anterior borders of the mentum and thyroid cartilage, whilethe patient lies supine with her/his mouth closed. We have termed this the thyromental heighttest (TMHT). Our aim in this study was to determine its utility in predicting difficult laryngoscopy.METHODS: Three hundred fourteen consecutive male and female patients aged 16 yearsscheduled to undergo general anesthesia were invited to participate. Airway assessments wereperformed with the modified Mallampati test, thyromental distance and sternomental distance,and TMHT in the preoperative clinic. Afterward, Cormack and Lehane grade of laryngoscopyviews was assessed during intubation. The laryngoscopist was unaware of airway assess-ments. As a primary end point, the validity and prediction indexes for the TMHT were calculated.Calculation of validity indexes for the 3 other methods of airway assessment was a secondaryobjective of this study.RESULTS:The optimal sensitivity and specificity values were in the range of 47.46 to 51.02mm. To facilitate clinical application, a cutoff value equal to 50 mm was chosen. TMHT was

more accurate than the other tests (all P< 0.0001).CONCLUSIONS:The TMHT appears to be a more accurate predictor of difficult laryngoscopythan the existing anatomical measurements. (Anesth Analg 2013;117:134751)

Thyromental Height: A New Clinical Test for Predictionof Difficult Laryngoscopy

Farhad Etezadi, MD, Aylar Ahangari, Hajar Shokri, Atabak Najafi, MD, Mohammad Reza Khajavi, MD,Mahtab Daghigh, MA, and Reza Shariat Moharari, MD

From the Department of Anesthesiology, Tehran University of MedicalSciences (TUMS), Tehran, Iran.

Accepted for publication July 31, 2013.

Funding: Not funded.

The authors declare no conflicts of interest.

Reprints will not be available from the authors.

Address correspondence to Reza Shariat Moharari, MD, Department of An-esthesiology, Tehran University of Medical Sciences (TUMS), Sina Hospital,Hassan Abad Sq., Tehran, Iran. Address e-mail to moharari@tums.ac.ir.

mailto:moharari@tums.ac.irmailto:moharari@tums.ac.ir

7/27/2019 00000539-201312000-00014 (1)

2/5

1348 www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA

Thyromental Height Test

sterni and the bony point of the mentum, with thehead in full extension and the mouth closed. AnSMD 13.5 cm was considered predictive of difficultlaryngoscopy.16

4. The new clinical test (TMHT) was performed as fol-lows: we measured the height between the anterior

border of the thyroid cartilage (on the thyroid notchjust between the 2 thyroid laminae) and the anterior

border of the mentum (on the mental protuberanceof the mandible), with the patient lying supine withher/his mouth closed (Fig. 1A). This height was mea-sured with a depth gauge (INSIZEElectronic DepthGage, INSIZE Co LTD., Suzhou New District, China;Fig. 1B). We have termed this the TMHT. Furthermore,age, sex, height, body weight, body mass index, andASA physical status of all patients were recorded.

General anesthesia was induced via a combination ofmidazolam (24 mg), fentanyl (23 g/kg), and sodiumthiopental (24 mg/kg). Muscle relaxation was achievedwith atracurium (0.5 mg/kg).

The patients head was placed in a sniffing position,and laryngoscopy was performed 3 minutes after atracu-rium injection using a Macintosh #4 blade. If no laryngeal

view was achieved, a second attempt was made using aMiller #4 blade combined with external, backward laryn-geal pressure and, occasionally, adjustment of head posi-tion. All 3 intubation attempts were performed by thesame resident in all 314 patients under the supervisionof an attending anesthesiologist. The laryngoscopist wasunaware of airway assessments. The best view that ledto successful intubation was assigned a grade of I to IV

according to CL criteria.17CL grades I and II were clas-sified as easy, whereas CL grades III and IV were des-ignated as difficult.

The primary end point was calculation of the predic-tion index (cutoff value) and validity indexes (sensitivity,specificity, accuracy, positive predictive value, negativepredictive value, likelihood ratio, odds ratio, index) forthe TMHT. Calculation of validity indexes for the 3 othermethods of airway assessment was a secondary objectiveof this study.

Statistical AnalysisWe used preoperative assessment data and findings dur-

ing intubation for evaluation of the above-mentioned testsin predicting difficult laryngoscopy. Data are presented asnumber/percent or mean SD. Area under the receiveroperating characteristic (ROC) curve for the TMHT wasused to show the predictive values of various thyromentalheight distances. Area under the curve and area under theROC curve were used to calculate the ideal cutoff point forTMHT. In addition, the validity indexes (sensitivity, specific-ity, accuracy, positive and negative predictive values, oddsratio, likelihood ratios), index, and Pvalues were calcu-lated for each test. Estimating a 10% incidence of difficultlaryngoscopy, a sample size of 180 was calculated to haveat least 90% power to detect the agreement between the CL

test and the predictors. Since a dramatic imbalance in thenumber of patients with and without outcome was antici-pated, power attenuation had been expected. Therefore,for power maintenance, we decided to increase sample sizeabout 2-fold (314). Sensitivity, specificity, accuracy, positivepredictive value, negative predictive value, and confidenceinterval (CI) were calculated based on normality assump-tion as the prerequisite condition (np and nq >5)* was metin most of the indexes. In cases that the prerequisites werenot met, our emphasis was on point estimate rather than CIand significance. Therefore, we did not compute CI by boot-straps or other suitable statistical approaches. Normalitywas also applied to them roughly.

2test (Fisher exact test) was used for statistical compari-son, and measure was used for agreement. Through a pilotstudy, intraclass correlation coefficient was calculated by 2examiners (blinded to each others measurements) and thesame depth gauge. All statistical analyses were performedusing SPSS version 16 (SPSS, Chicago, IL).

RESULTSThree hundred fourteen ASA physical status and IIpatients (149 men [47.5%] and 165 women [52.5%]) with amean age of 44.5 15 years participated in the study. Two

Figure 1.A, Definition of thyromental height; while the patient lies

supine with her/his mouth closed, the line AD is equal to thyro-mental height. B, Measurement of thyromental height using a depth

gauge. * n= (number, sample size), p= prevalence, q= (1- prevalence)

7/27/2019 00000539-201312000-00014 (1)

3/5

December 2013 Volume 117 Number 6 www.anesthesia-analgesia.org 1349

patients were excluded from the study because of compro-mised airway diagnosis in the preoperative clinic. Twenty-three patients (7.3%) had a CL grade III or IV laryngealview. There were no failed intubations. Descriptive andquantitative data of the patients and the airway assessmentmethods are shown in Tables 1 and 2. According to the ROCcurve, the optimal sensitivity and specificity values were inthe range of 47.46 to 51.02 mm and are shown in Table 3. To

facilitate clinical application, a cutoff value equal to 50 mmwas chosen. True-positive, false-positive, true-negative, andfalse-negative (FN) results together with sensitivity, speci-ficity, positive predictive value, negative predictive value,accuracy, odds ratio, likelihood ratio, index, and Pvaluescalculated for MMT, TMD, SMD, and TMHT are shown inTables 2 and 4. index was significant for the TMHT (0.853[0.740.95], P< 0.0001). However, it was not significant forTMD, MMT, and SMD tests. In a pilot study (15 eligible sub-

jects) performed by 2 blinded examiners, the means of thy-romental height measurements were 59.17 10.71 mm and58.98 10.94 mm (intraclass correlation coefficient = 0.93).

DISCUSSIONThe TMHT was found to be a more accurate predictor of dif-ficult laryngoscopy than other single anatomical measures.MMT is one of the most widely reported methods used forprediction of difficult laryngoscopy. Although when usedalone this method has a poor predictive value, it may bevaluable as part of a multivariate model for prediction ofa difficult laryngoscopy.1820Schmitt et al.21made a modi-fication in the TMD test: ratio of patients height to thyro-mental distance (RHTMD). The sensitivity of the TMD andRHTMD test was the same, while the RHTMD test showeda little better specificity.21Al Ramadhani et al.16found thatthe SMD test by itself may not be an adequate sole predictor

of subsequent difficult laryngoscopy. They suggested thatthe SMD test is better when used as part of a series of air-way assessment tests.16

The consequence of an FN prediction (i.e., laryngoscopyand intubation would be easy) may prove to be catastrophic.Thus, decreasing FN predictions (increased sensitivity)is more important than falsely predicting difficulty (falsepositive) for patients in whom laryngoscopy and intubationare accomplished easily.15,22The results of the current studysupport previous studies regarding poor sensitivity andpositive predictive values of MMT, TMD, and SMD tests.

Also, specificity and negative predictive values for the 3above-mentioned tests in the current study are comparablewith that of the previous studies.16,1824The typical advan-tage of the TMHT is high sensitivity and positive predictivevalues in comparison with the 3 other methods of airwayassessment. Naguib et al.11 reported the sensitivity of 3multivariate clinical models (Wilson, Arn, and Naguib)as 40.2%, 54.6 %, and 81.4%, respectively. The sensitivity ofthe TMHT, 82.6% (CI, 74%88%), is approximately equal toNaguib et al.s multivariate clinical model. Also, the speci-ficity value of the TMHT test, 99.31% (CI, 96%99.98%), iscomparable with that of the above-mentioned multivariateclinical tests.

In an analytical observational study that focused exclu-sively on TMD, Qudaisat and Al-Ghanem24found that theTMD method is a surrogate for inadequate head exten-sion rather than dimensions of submandibular space. Thecalculated sensitivity of TMD (21.7%) in the current studyis close to that of Qudaisat and Al-Ghanems study (19%).The TMHT on its own is not dependent on active headextension. However, it must be noted that patients lyingsupine inevitably would place 98% of them in a relativehead extension.25Both SMD and TMD tests must be mea-sured in full head extension; thus, they are dependent onthe patients cooperation, adequate cervical spine mobil-ity, and having no contraindication for full head extension.

A more caudal or anterior larynx is associated with dif-ficult laryngoscopy,26 and it can be expected to correlatewith a shorter thyromental height. To compensate for this,

backward, upward, and rightward pressure can be used toimprove the laryngoscopic view.5,2729 This posterior dis-placement increases the TMD, effectively increasing thethyromental height. One advantage of the TMHT is the useof an inexpensive, easily applicable instrument for measure-ment of an objective quantity (distance between anteriorsurfaces of the mentum and thyroid cartilage). It must bementioned that the anatomical difference and measurementerrors may affect the test results. For example, the assurance

Table 1. Demographic Data of the PatientsGender Male (47.5%) Female (52.5%)

Variable Min Max Mean SD

Age (y) 16 84 44.5 15.3

Height (cm) 110 188 166.1 10.6

Weight (kg) 40 124 72 14.1

BMI (kg/m2) 16.2 39.4 25.8 4.3

BMI = body mass index.

Table 2. Comparison Between CormackLehane Grades and 4 Preoperative Predictors

Model

CL grade

Total (95% CI)a

Significance

Fisher

exact testEasy (n= 291) Difficult (n= 23)

TMH Easy 289 4 2930.853 (0.740.95)

7/27/2019 00000539-201312000-00014 (1)

4/5

1350 www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA

Thyromental Height Test

of a proper horizontal alignment may require the use of abubble device such as those on carpentry levels. However,this study found a small interobserver variability for TMHTmeasurement (this was performed on a small subgroup ofpatients using only 2 observers). In contrast, large interob-server variability is a major problem with the MMT.4

Our study has some limitations. First, the study pop-ulation was limited to patients scheduled for elective,nonemergent surgeries including orthopedic (trauma,spine), thoracic, abdominal, and vascular (central, periph-eral). Thus, the results are only applicable to this groupof patients. Second, determination of the best cutoff pointfor use of the TMHT, as a difficult laryngoscopy predic-tor, and its analysis, as a measure of prediction, have both

been performed on the same population; this may explainthe good statistical results calculated for the TMHT incomparison with the other tests. Third, we calculated theideal cutoff value for the TMHT from the data obtainedin the study, whereas the cutoff value for the other testswas obtained from the literature. Although this statisti-cal approach may lead to a weaker or a better result forthe competing parameters, it should be remembered thatthe primary end point of this study was the TMHT, not anevaluation of existing methods. In addition, the validity

indexes calculated for the competing tests were in concor-dance with many other studies.

In summary, the TMHT appears promising as a singleanatomical measure to predict the risk of difficult laryn-goscopy, but validation will require further studies in morediverse patient populations.E

DISCLOSURES

Name:Farhad Etezadi, MD.Contribution:This author helped design the study, analyze thedata, and write the manuscript.Attestation:Farhad Etezadi has seen the original study data,reviewed the analysis of the data, approved the final manu-

script, and is the author responsible for archiving the studyfiles.Name:Aylar Ahangari.Contribution:This author helped conduct the study.Attestation: Aylar Ahangari has seen the original study dataand approved the final manuscript.Name:Hajar Shokri.Contribution:This author helped conduct the study.Attestation:Hajar Shokri has seen the original study data andapproved the final manuscript.Name:Atabak Najafi, MD.Contribution:This author helped design the study.Attestation:Atabak Najafi has seen the original study data andapproved the final manuscript.

Table 3. Calculated Cutoff Values That Show theBest Range of Sensitivity and Specificity for theThyromental Height TestCutoff point Sensitivity 1 specificity

47.4650 0.826 0.000

48.8750 0.826 0.003

49.6350 0.826 0.007

50.6550 0.826 0.010

50.9900 0.870 0.010

51.0250 0.913 0.010

Ta

ble4

.Statis

tica

lResu

lts

fo

rthe

4Me

tho

ds

toPre

dicttheO

ccurrenceo

fGra

de

3or

4Accor

ding

toCormac

kan

dLe

hane

Mo

de

l

TP

TN

FP

FN

Sens

itivity

P

PV

Spec

ific

ity

NPV

Accuracy

Oddsra

tio

Like

lihoo

d

ratio

(+)

TMH(50mm)

19

289

2

4

82.6

%

(74%

88%)a

90.4

7%

(83%

95%)a

99.3

1%

96%

99.9

8%a

98.63%

(94%

99%)a

98.0

8%

(94%

99%)a

686.3

118.0

01

TMD

(