Open Journal of Anesthesiology, 2013, 3, 320-325
http://dx.doi.org/10.4236/ojanes.2013.37070 Published Online September 2013 (http://www.scirp.org/journal/ojanes)
Clinical Criteria for Airway Assessment: Correlations with
Laryngoscopy and Endotracheal Intubation Conditions
Gustavo Henr iqu e S. Wa nderley1, Luciana Cavalcanti Lima2,3,4, Tânia C u r sino de Menezes Couceiro2,5,6,
Waston Vieira Silva2, Raquel Queiroz G. A. Coelho2, Andrea Cavalcanti C. Lucena7,
Anne Danielle Santos Soares2
1Agreste Regional Hospital, Caruaru, Brazil; 2Instituto de Medicina Integral Professor Fernando Figueira (IMIP), Recife, Brazil;
3State University of São Paulo (UNESP), São Paulo, Brazil; 4Faculdade Pernambucana de Saúde, Recife, Brazil; 5Medical Residency
Program in Anesthesiology, IMIP, Recife, Brazil; 6Neuropsychiatry and Behavioral Science, Federal University of Pernambuco,
Recife, Brazil; 7Brazilian Society of Anesthesiology Accredited Teaching and Training Center, IMIP, Recife, Brazil.
Received June 27th, 2013; revised July 27th, 2013; accepted August 15th, 2013
Copyright © 2013 Gustavo Henrique S. Wanderley et al. This is an open access article distributed under the Creative Commons At-
tribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is prop-
erly cited.
ABSTRACT
Difficult intubation, inadequate ventilation and esophageal intubation are the principal causes of death or brain damage
related to airway manipulation. The objective of this cross-sectional study was to correlate a preanesthetic evaluation
that may be capable of predicting a difficult intubation with the conditions encountered at laryngoscopy and endotra-
cheal intubation. Eighty-one patients submitted to general anesthesia were evaluated at a preanesthetic consultation ac-
cording to the modified Mallampati classification, the Wilson score and the American Society of Anesthesiologists
(ASA) difficult airway algorithm. Findings were then correlated with the Cormack-Lehane classification and with the
number of attempts at endotracheal intubation. No statistically significant correlations were found between the patients’
Mallampati classification and their Cormack-Lehane grade or between the Mallampati classification and the number of
attempts required to achieve endotracheal intubation. Laryngoscopy proved difficult in four patients and in all of these
cases the Wilson score had been indicative of a possibly difficult airway, highlighting its good predicting sensitivity.
However, the specificity of this test was low, since another 24 patients had the same Wilson score but were classified as
Cormack-Lehane I/II. Moreover, two patients who had a Wilson score ≥ 4 were also classified as Cormack-Lehane
grade I/II. The study concluded that the Wilson score, although seldom used in clinical practice, is a highly sensitive
predictor of a difficult airway; its specificity, however, is low.
Keywords: Preanesthetic Evaluation; Respiratory System; Airway; Endotracheal Intubation; Measurement Techniques;
Mallampati; Wilson; Cormack-Lehane
1. Introduction
Difficult airway management is one of the principal
challenges faced by anesthesiologists in their routine
practice. Data published by the American Society of An-
esthesiologists (ASA) show that, despite the decline reg-
istered over recent decades, adverse respiratory events
were involved in 32% of all lawsuits raised against anes-
thesiologists in the 1990s. Difficult intubation, inade-
quate ventilation and esophageal intubation were the
principal factors responsible for death or brain damage
[1].
Different clinical parameters have been proposed for
preoperative airway assessment. In 1985, Mallampati et
al. introduced a scoring system based on the visibility of
the oropharyngeal structures, which was later modified
into four classes by Samsoon and Young in 1987 [2].
Wilson developed a scoring system based on the sum of
constitutional and anatomical characteristics, and other
authors evaluated indexes calculated according to the
distances between anatomical structures (thyromental
distance, sternomental distance and interincisor distance)
[3,4]. The American Society of Anesthesiologists pub-
lished an algorithm for a difficult airway and listed 11

Clinical Criteria for Airway Assessment: Correlations with Laryngoscopy and Endotracheal Intubation Conditions 321
routine preoperative tests with their respective undesir-
able results (possible predictors of a difficult airway) [5].
Nevertheless, the diagnostic accuracy of airway assess-
ment tests has varied significantly in the different studies,
probably as a function of variations in the incidence of
difficult intubation, which may be explained by the con-
stitutional differences in the individual patients and in the
populations evaluated [3].
The objective of the present study was to correlate the
findings of the modified Mallampati classification, the
Wilson score and the ASA difficult airway algorithm
with the conditions encountered at conventional direct
laryngoscopy (Cormack-Lehane classification) and at
endotracheal intubation.
2. Methods
Following approval by the institution’s internal review
board, a cross-sectional study was conducted at the In-
stituto de Medicina Integral Professor Fernando Figueira
(IMIP), Recife, Pernambuco, Brazil, between December
2010 and November 2011.
The study inclusion criteria consisted of patients of ei-
ther sex in the 18 - 65 year age group, with an ASA
physical status classification of P1, P2 or P3, who re-
quired endotracheal intubation for general anesthesia.
Pregnant women, patients with cognitive deficiencies or
with any pathology that could alter the anatomy of the
face or neck were excluded.
Following selection, and after all the patients had
signed an informed consent form, a preanesthetic evalua-
tion was performed by the anesthesiologist and/or a resi-
dent anesthesiologist.
At the time of airway assessment, three data sets were
obtained: the patient’s modified Mallampati classifica-
tion (I, II, III or IV) (Figure 1); the Wilson score, which
takes five factors into consideration: weight, head and
neck mobility, jaw movement, retrognathia and buck
teeth (Table 1); and finally, the ASA difficult airway
algorithm, in which each undesirable result is awarded
one point (Table 2).
In the operating theater, the patient was placed in the
dorsal decubitus position and in the sniffing position,
Ⅰ Ⅱ Ⅲ Ⅳ
Figure 1. Modified mallampati classification.
which consists of flexing the neck and then extending the
head. A pillow is used to support the head sufficiently
high so as to ensure that the external auditory meatus and
the sternal notch are aligned horizontally [6].
With the patient duly in position, general anesthesia
was induced with the standard drugs used in this institute.
A neuromuscular blocking drug, atracurium, was then
given to all patients at a dose of 0.5 mg/kg. Once the 5-
minute latency period was over, laryngoscopy was per-
formed using a conventional laryngoscope with a Mac-
intosh blade # 3, 4 or 5, and the patient was then classi-
fied as Cormack-Lehane grade I, II, III or IV (Figure 2).
Due to the fact that the IMIP is a teaching institute,
laryngoscopy was performed by the resident anesthesi-
ologist scheduled for that particular procedure. This
resident may have been in the first, second or third year
of the residency program. Classification was then con-
firmed by the chief anesthesiologist. In cases of discor-
dance, the classification suggested by the chief anesthe-
siologist prevailed. Endotracheal intubation was then
performed, with the number of attempts required until
successful intubation, or the impossibility of endotra-
cheal intubation, being recorded.
Data concerning sex, weight, height, body mass index
(BMI) and ASA physical status classification were re-
corded. The variables analyzed were: the modified Mal-
lampati class, the Wilson score, the ASA difficult airway
algorithm, the Cormack-Lehane grade and the number of
endotracheal intubation attempts.
For the purposes of data analysis, the patients were
subdivided into groups for each index: Mallampati I/II or
Mallampati III/IV; a Wilson score of 0/1 or 2/3 or ≥4;
ASA difficult airway algorithm ≤6 or ≥7; Cormack-Le-
hane grade I/II or III/IV; number of attempts at intuba-
tion: ≤2 or >2 or failed intubation. The Epi Info software
program, version 3.5.3 was used and Fisher’s exact test
and the chi-square test were applied in the statistical
analysis. Results with p-values <0.05 were considered
statistically significant.
3. Results
The sample consisted of 81 patients, more than half of
whom were female. Mean age was 44 years and mean
BMI was 27. The predominant ASA physical status clas-
sification was I/II (Table 3).
A Mallampati classification of I/II was found in 52 pa-
Ⅰ Ⅱ Ⅲ Ⅳ
Figure 2. Cormack-lehane classification.

Clinical Criteria for Airway Assessment: Correlations with Laryngoscopy and Endotracheal Intubation Conditions 322
Table 1. The Wilson score method.
Risk Factors Score Points*
Weight
<90 kg
90 - 110 kg
>110 kg
0
1
2
Mobility of the head and neck
(Angle formed between the positions of greatest extension and greatest flexion of the neck)
>90˚
~90˚
<90˚
0
1
2
Jaw movement
IO: maximum interincisal opening
SLux: Jaw subluxation and maximum forward protrusion of the lower incisors beyond the upper incisors.
IO > 5 cm or SLux > 0
IO < 5 cm or SLux = 0
IO < 5 cm or SLux < 0
0
1
2
Retrognathia
Absent
Moderate
Severe
0
1
2
Buck teeth
Absent
Moderate
Severe
0
1
2
Scores ≥ 2 and ≤4 = a possibly difficult intubation; >4 = often difficult intubation.
Table 2. Airway assessment and undesir a ble re sults: ASA pr e dic tor s.
Parameters Undesirable Results
1. Length of upper incisors Relatively long
2. Relation of maxillary and mandibular incisors
during normal jaw closure
Prominent “overbite”
(maxillary incisors anterior to mandibular incisors)
3. Relation of maxillary and mandibular incisors during voluntary
protrusion of jaw Patient cannot bring mandibular incisors anterior to maxillary incisors
4. Interincisor distance Less than 3 cm
5. Visibility of uvula Not visible when tongue is protruded with patient in sitting position
(e.g. Mallampati class greater than II)
6. Shape of palate Highly arched or very narrow
7. Compliance of mandibular space Stiff, indurated, occupied by mass, or nonresilient
8. Thyromental distance Less than three ordinary finger breadths
9. Length of neck Short
10. Thickness of neck Thick
11. Range of motion of head and neck Patient cannot touch tip of chin to chest or cannot extend neck
tients (64.2%), while 29 (35.8%) were class III/IV. Fifty-
one patients (63%) had a Wilson score of 0/1, while 28
(34.6%) had a score of 2/3 and 2 patients (2.5%) scored
≥ 4. All the patients had an ASA algorithm < 6 (Table
4).
Seventy-seven patients (95.1%) were classified as
Cormack-Lehane grade I/II, while 4 patients (4.9%) were
considered grade III/IV. Overall, 97.5% of the patients (n
= 79) were intubated at the first or second attempt; how-
ver, in one patient (1.2%) more than two attempts were e

Clinical Criteria for Airway Assessment: Correlations with Laryngoscopy and Endotracheal Intubation Conditions 323
Table 3. Demographic characteristics of the study sample.
Sex ASA Physical Status
Male Female
Mean age1 Mean Height2 Mean Weight3 Mean BMI4
P1 P2 P3
23 58 44 1.61 68 27 38 40 3
(28.4%) (71.6%) (21 - 65) (1.47 - 1.86) (45 - 122) (18 - 44) (46.9%) (49.4%) (3.7%)
Table 4. Number of patients in the study sample according
to their modified Mallampati classification and Wilson
score.
Mallampati Class Wilson Score
I or II III or IV 0 or 1 2 or 3 ≥4
n = 52 n = 29 n = 51 n = 28 ≥4
(64.2%) (35.8%) (63%) (34.6%) (2.5%)
required and in another (1.2%) intubation proved impos-
sible with the use of a conventional laryngoscope, re-
quiring the use of a gum elastic bougie (i.e. a straight,
semi-rigid stylette-like device), as a guide through which
intubation was achieved.
When the modified Mallampati and Cormack-Lehane
classifications were correlated, it was found that, of the
52 patients who were Mallampati class I/II, 50 (96.2%)
were Cormack-Lehane grade I/II, while 2 (3.8%) were
grade III/IV. Of the 29 patients classified as Mallampati
III/IV, 27 (93.1%) were Cormack-Lehane grade I/II,
while 2 (6.9%) were grade III/IV. No statistically sig-
nificant correlations were found (p = 0.54) (Table 5).
All 51 patients with a Wilson score of 0/1 were classi-
fied as Cormack-Lehane I/II. Of the 28 patients with a
Wilson score of 2/3, 24 (85.7%) were classified as Cor-
mack-Lehane I/II, while 4 (14.3%) were grade III/IV.
The two patients who received a Wilson score ≥ 4 were
Cormack-Lehane grade I/II. This correlation was statis-
tically significant (p = 0.01) (Table 6).
When the modified Mallampati classification was cor-
related with the number of attempts at intubation, it was
found that of the 52 patients classified as Mallampati I/II,
50 (96.2%) were intubated at the first or second attempt,
while in one case (1.9%) more than two attempts were
required, and in another case (1.9%) intubation proved
impossible using the conventional laryngoscope. All 29
patients classified as Mallampati III/IV were successfully
intubated at the first or second attempt; with no statisti-
cally significant correlations being established (p = 0.56)
(Table 7).
All the patients with a Wilson score of 0/1 were intu-
bated at the first or second attempt. Of the 28 patients
with a Wilson score of 2/3, 26 (92.9%) were intubated at
the first or second attempt, while in one case (3.6%)
Table 5. Association between the modified Mallampati
classification and the Cormack-Lehane grade
Mallampati Cormack-Lehane TOTAL
I or II III or IV
I or II 50 (96.2%) 2 (3.8%) 52
III or IV 27 (93.1%) 2 (6.9%) 29
p = 0.54.
Table 6. Association between the Wilson score and the
Cormack-Lehane grade
Wilson score Cormack-Lehane TOTAL
I or II III or IV
0 or 1
2 or 3
51 (100%)
24 (85.7%)
0
4 (14.3%)
51
28
≥ 4 2 (100%) 0 2
p = 0.01.
Table 7. Association between the modified Mallampati
classification and the number of endotracheal intubation
attempts.
Intubation attempts TOTAL
Mallampati
1 or 2 >2 Failed intubation
I or II 50 (96.2%)1 (1.9%) 1 (1.9%) 52
III or IV 29 (100%)0 0 29
p = 0.56.
more than two attempts were required and in another
case (3.6%) intubation proved impossible with the con-
ventional laryngoscope. The two patients with a Wilson
score ≥ 4 were intubated at the first or second attempt.
Therefore, no statistically significant correlations were
found (p = 0.42) (Table 8).
The correlation between the Cormack-Lehane classi-
fication and the number of endotracheal intubation at-
tempts showed that of the 77 patients classified as Cor-
mack-Lehane I/II, 76 (98.7%) were intubated at the first
or second attempt, while in one case (1.3%) more than
two attempts were required to achieve successful intuba-
tion. Of the four patients classified as Cormack-Lehane
III/IV, 3 (85%) were intubated at the first or second at-

Clinical Criteria for Airway Assessment: Correlations with Laryngoscopy and Endotracheal Intubation Conditions 324
tempt, while in one case (25%) intubation proved impos-
sible. This correlation was statistically significant (p =
0.0001) (Table 9).
4. Discussion
In anesthesiology, airway assessment at the preanesthetic
consultation has been found to constitute a moment of
extreme importance, and investigators in this field are
constantly searching for better predictors of a difficult
airway. The most commonly used tests for predicting
difficult intubation include the Mallampati score, modi-
fied by Samsoom and Young [2], measurement of the
sternomental and thyromental distances, the mouth
opening, and the mobility of the neck and the jaw. In-
dexes that are less commonly used in practice, such as
the Wilson score [7], and even the ASA difficult airway
algorithm [5], have been studied by some authors, with
conflicting results.
One of the important characteristics of the present
study is that the exclusion criteria were few, resulting in
the inclusion of a broad range of patients. This is impor-
tant since, in clinical practice, the anesthesiologist will be
confronted with a wide diversity of patients and their
respective physical constitutions and anatomical varia-
tions.
In addition to standardizing the patient’s position (the
sniffing position), it was also important to standardize the
neuromuscular blocking drug (0.5 mg·kg−1 of atracurium,
with a latency period of 3 - 5 minutes), thus guaranteeing
optimal conditions for endotracheal intubation in all the
Table 8. Association between the Wilson score and the
number of endotracheal intubation atte mpts.
Intubation attempts TOTAL
Wilson score
1 or 2 >2 Failed intubation
0 or 1 51 (100%) 0 0 51
2 or 3 26 (92.9%) 1 (3.6%)1 (3.6%) 28
≥4 2 (100%) 0 0 2
p = 0.42.
Table 9. Association between the Cormack-Lehane grade
and the number of endotracheal intubation attempts.
Intubation attempts
Cormack-Lehane
1 or 2 >2 Failed intubation
TOTAL
I or II 76
98.7%
1
1.3% 0 77
III or IV 3
75%
0 1
25%
4
p = 0.0001.
patients. Although the laryngoscopies were performed by
residents, they were always confirmed by the chief anes-
thesiologist, who determined the Cormack-Lehane clas-
sification.
Shiga et al. [3] published a meta-analysis in 2005
showing that specificity and sensitivity were not high
with any of the tests used alone to predict a difficult air-
way and that they may result in poor positive and nega-
tive predictive values. Combining these tests leads to
slightly better indexes. Lundstrom et al. [4] reported
similar results in a meta-analysis published in 2011 in-
volving 177,088 patients in which only 35% of the pa-
tients in whom endotracheal intubation proved difficult
had been identified as Mallampati III or IV. Adamus et al.
[8] reported a sensitivity of 64.6% for the modified Mal-
lampati classification in predicting cases of a difficult
airway. In the present study, 50% of the patients (n = 2)
in whom laryngoscopy was predicted to be difficult (Cor-
mack-Lehane III/IV) were classified as Mallampati III/
IV, whereas those in whom intubation indeed proved
difficult (1) or impossible (1) had been classified as Mal-
lampati I/II. Although these results were not statistically
significant in the present study (p = 0.54 and p = 0.56,
respectively), they show a tendency towards agreement
with the previously mentioned studies.
The historical importance of the Mallampati test is in-
disputable. Prior to 1985 when it was created [1], inves-
tigators were already concerned with studying the airway
and were aware of the need to identify reliable predictors
of a difficult airway. Nevertheless, with the publication
of new studies, this index has undergone criticism. Fac-
tors such as the positioning of the patient during the
exam, the patient’s ability to understand, the presence or
absence of phonation, and pregnancy [9] may alter the
patient’s Mallampati class, which may explain the dif-
ferent incidence of Mallampati classes in the different
studies. On the other hand, when studies use similar
methodologies and have similar objectives, results tend
to be concordant.
In the present study, the Wilson score successfully pre-
dicted 100% (n = 4) of the patients in whom laryngo-
scopy proved difficult (Wilson 2/3) (p = 0.01). This re-
flects the good sensitivity of this test. Specificity, how-
ever, was poor given that another 24 patients had the
same Wilson score but were classified as Cormack-Le-
hane I/II. Furthermore, laryngoscopy proved simple in
another two patients who had a Wilson score of ≥4.
These results appear to be in line with the findings of
Domi [10], published in 2009, in which the Wilson score
successfully predicted 82.5% of cases of difficult airway,
a better result than the 22.5% found when the Mallampati
classification was used in conjunction with the thy-
romental and sternomental distances.

Clinical Criteria for Airway Assessment: Correlations with Laryngoscopy and Endotracheal Intubation Conditions 325
Since the Wilson score takes various factors and ana-
tomical characteristics into consideration rather than just
one as in the case of the Mallampati classification, sensi-
tivity and specificity tend to be higher. Moreover, the
characteristics evaluated are well defined and well de-
scribed, leaving less margin for subjectivity during the
exam. This brings the various studies closer methodol-
ogically, ensuring that the results are in general compa-
rable.
Some limitations of the present study must be men-
tioned. The total number of patients in the sample (81) is
considered small for a study involving events that are
relatively rare in the general population such as the case
of a difficult airway. This may have been responsible for
the lack of statistical significance in some of our results.
Since this is a teaching and training institute for anesthe-
siologists, most of the data were collected by trainee
physicians. Nevertheless, although data were collected
by different individuals throughout the study period, all
were duly trained for this function by the authors. Be-
cause the intubations were performed by resident physi-
cians, the number of attempts recorded may have been
less in some cases if the procedure had been performed
by an experienced anesthesiologist. It should be empha-
sized that in the patient in whom intubation was unsuc-
cessful by conventional laryngoscopy, the anesthesiolo-
gist responsible for the case also failed to intubate the
patient and was obliged to resort to the use of a gum
elastic bougie to successfully conclude the procedure.
The present study concludes that the Wilson score,
despite being seldom used in clinical practice, is a highly
sensitive predictor of a difficult airway, although its
specificity is low. Further studies with larger sample
sizes are required to confirm these findings and to obtain
statistically significant results for the other indexes eva-
luated.
REFERENCES
[1] S. R. Mallampati, S. P. Gatt, L. D. Gugino, S. P. Desai, B.
Waraksa, D. Freiberger and P. L. Liu, “A Clinical Sign to
Predict Difficult Endotracheal Intubation: A Prospective
Study,” Canadian Anaesthetists’ Society Journal, Vol. 32,
No. 4, 1985, pp. 429-434.
[2] G. L. Samsoon and J. R. Young, “Difficult Endotracheal
Intubation: A Retrospective Study,” Anaesthesia, Vol. 42,
No. 5, 1987, pp. 487-490.
doi:10.1111/j.1365-2044.1987.tb04039.x
[3] T. Shiga, Z. Wajima, T. Inoue and A. Sakamoto, “Pre-
dicting Difficult Intubation in Apparently Normal Pa-
tients. A Meta-Analysis of Bedside Screening Test Per-
formance,” Anesthesiology, Vol. 103, No. 2, 2005, pp.
429-437. doi:10.1097/00000542-200508000-00027
[4] L. H. Lundstrøm, M. Vester-Andersen, A. M. Møller, S.
Charuluxananan, J. L’hermite and J. Wetterslev, “Poor
Prognostic Value of the Modified Mallampati Score: A
Meta-Analysis Involving 177 088 Patients,” British Jour-
nal of Anaesthesia, Vol. 107, No. 5, 2011, pp. 659-667.
doi:10.1093/bja/aer292
[5] J. L. Apfelbaum, C. A. Hagberg, R. A. Caplan, C. D. Blitt,
R. T. Connis, D. G. Nickinovich, C. A. Hagberg, R. A.
Caplan, J. L. Benumof, F. A. Berry, C. D. Blitt, R. H.
Bode, F. W. Cheney, R. T. Connis, O. F. Guidry, D. G.
Nickinovich and A. Ovassapian, “Practice Guidelines for
Management of the Difficult Airway: An Updated Report
by the American Society of Anesthesiologists Task Force
on Management of the Difficult Airway,” Anesthesiology,
Vol. 118, No. 2, 2013, pp. 251-270.
doi:10.1097/ALN.0b013e31827773b2
[6] K. B. Greenland, M. J. Edwards and N. J. Hutton, “Ex-
ternal Auditory Meatus-Sternal Notch Relationship in
Adults in the Sniffing Position: A Magnetic Resonance
Imaging Study,” British Journal of Anaesthesia, Vol. 104,
No. 2, 2010, pp. 268-269. doi:10.1093/bja/aep390
[7] M. E. Wilson, D. Spiegelhalter, J. A. Robertson and P.
Lesser, “Predicting Difficult Intubation,” British Journal
of Anaesthesia, Vol. 61, No. 2, 1988, pp. 211-216.
doi:10.1093/bja/61.2.211
[8] M. Adamus, S. Fritscherova, L. Hrabalek, T. Gabrhelik, J.
Zapletalova and V. Janout, “Mallampati Test as a Predic-
tor of Laryngoscopic View,” Biomedical Papers of the
Medical Faculty of the University Palacký, Vol. 154, No.
4, 2010, pp. 339-344. doi:10.5507/bp.2010.051
[9] A. Lee, L. T. Fan, T. Gin, M. K. Karmakar and W. D.
Ngan Kee, “A Systematic Review (Meta-Analysis) of the
Accuracy of the Mallampati Tests to Predict the Difficult
Airway,” Anesthesia and Analgesia, Vol. 102, No. 6, 2006,
pp. 1867-1878. doi:10.1213/01.ane.0000217211.12232.55
[10] R. Domi, “A Comparison of Wilson Sum Score and
Combination Mallampati, Tiromental and Sternomental
Distances for Predicting Difficult Intubation,” Macedo-
nian Journal of Medical Sciences, Vol. 2, No. 2, 2009, pp.
141-144. doi:10.3889/MJMS.1857-5773.2009.0045