Original Contribution

Determination of difficult intubation in the ED^?

Secgin Soyuncu MD?, Cenker Eken MD, Yildiray Cete MD, Firat Bektas MD, Mehmet Akcimen MD

Department of Emergency Medicine, Akdeniz University School of Medicine, 07059 Antalya, Turkey

Received 5 May 2008; revised 30 June 2008; accepted 2 July 2008

Abstract

Objective: The aim of this study is to determine the predictors of difficult intubation in the emergency setting.

Methods: This prospective observational clinical study was conducted in the emergency department (ED) of a University Hospital with an annually census of 50 000 visits from May 2005 to May 2007. All patients requiring intubation in the ED were included into the study. During the study period, same airway management protocol was used all intubations. The study form included patient’s demographic and variables according to intubation such as the Cormack-Lehane grade, modified LEMON score, Glasgow Coma Scale score, success rate, and associated complications.

Results: A total of 366 patients were included in the study. The mean age of the study patients was

46.8 +- 22.8, and 68.6% (n = 251) of them were male. A total of 86 (23.5%) patients were classified in the difficult intubation group and 280 (76.5%) patients in easy intubation group. Logistic regression analysis performed by the variables found to be significant in the univariate analysis revealed thyroid-to- hyoid distance less than 2 fingers (odds ratio, 3.34; 95% confidence interval, 1.35-8.27; P = .009) as an independent factor complicating the intubation. Cormack and Lehane classification was strongly related to difficult intubation. Intubation was more difficult from grade 1 to 4 (11% vs 25.2% vs 34% vs 81.8%, respectively; P = .000).

Conclusions: The thyroid-to-hyoid distance less than 2 fingers is the only independent variable in predicting difficult intubation. Mallampati classification is not a useful tool in classifying the difficult intubation in the ED that the “LEMON” acrostic can be modified to “LEON”.

(C) 2009

Introduction

There is still controversy on the definition of difficult airway in the medical literature. The difficult airway represents a complex interaction between the anatomic features of the patient, the clinical setting, and the procedural

^? This study was supported by Akdeniz University Foundation.

* Corresponding author. Tel.: +90 242 2496183; fax: +90 242 2274490.

E-mail addresses: ssoyuncu@akdeniz.edu.tr (S. Soyuncu), cenkereken@akdeniz.edu.tr (C. Eken), ycete@akdeniz.edu.tr (Y. Cete), fbektas@akdeniz.edu.tr (F. Bektas), akcimen13@yahoo.com (M. Akcimen).

skills of the physician. The American Society of Anesthesiol- ogy (ASA) published the ASA Difficult Airway Algorithm in 2003 [1]. In this guideline, the difficult airway is defined as the clinical situation in which a conventionally trained anesthesiologist experiences difficulty with face mask ventilation of the upper airway, difficulty with tracheal intubation, difficulty with laryngoscopy, and failed intubation.

Because airway management in the operating room and in the emergency department (ED) differs, the ASA algorithm involves a number of characteristics preventing its utilization in the ED practice.

0735-6757/$ – see front matter (C) 2009 doi:10.1016/j.ajem.2008.07.003

The airway management is more difficult in the ED than the operating room because of the compressed time frames, unpredictable events, and the nonideal conditions. To define a difficult airway is a challenging issue. Walls [2] defined a difficult airway as difficult to oxygenate and ventilate, difficult to intubate, and difficult to perform a cricothyroidotomy, which is also similar to ASA. Cormack and Lehane [3] defined difficult intubation as poor glottic visualization during direct laryngoscopy, or a high-grade Laryngeal view with no ability to see the vocal cords or the glottic aperture. These are subjective theoretical definitions of difficult airway, which are not appropriate for a clinician before an intubation.

The aim of this study is to determine the predictors of difficult intubation in the emergency setting.

Methods

Study design

This prospective observational clinical study was con- ducted in the ED of a university hospital with an annual census of 50 000 visits from May 2005 to May 2007.

Setting

All patients who required intubation in the ED were included into the study. All intubations were supervised by attending physicians from 8 AM to midnight. The other intubations in the remaining time were supervised by senior residents. All ED residents were given the theoretical and practical education about airway management regularly during their residency program.

Data collection and measurements

The study data were collected by the attending physicians and residents with the residency program of

5 years. The study forms were filled by the intubating doctor just after the intubation attempt. The study form included patient’s age and sex, indication for intubation, vital signs before and after intubation, drugs used during the intubation, endotracheal tube size, class of the resident and the supervisor, number and details of each attempt at intubation including the Cormack-Lehane grade, modified LEMON score, Glasgow Coma Scale score, success rate, and associated complications.

• • 1. Definitions (according to the ASA difficult airway algorithm)

Difficult laryngoscopy: Unable to visualize any portion of the vocal cords after multiple attempts at conventional laryngoscopy.

Difficult tracheal intubation: Tracheal intubation requires multi- ple attempts in the presence or absence of tracheal pathology.

Fig. 1 LEMON law.

• • 1. Definition of LEMON Law

National Emergency airway management Course has developed the “LEMON law” for identification of the difficult airway. The score with a maximum of 10 points has been calculated by assigning 1 point for each of the LEMON criteria (see Fig. 1).

Reed et al [4] have recommended a modified LEMON score in the emergency settings in their study. Modified LEMON score has been calculated by assigning 1 point for each criterion except Mallampati score, with a maximum of 9 points. In our study, the modified LEMON version was also used.

• • 1. Cormark and Lehane classification

Cormack and Lehane have proposed a classification scheme for views of the laryngeal inlet obtained at laryngoscopy. This 4-grade scheme has become the standard measurement of glottic views and facilitated communication between researcher and practitioners as to the impact of the view obtained on the success of tracheal tube placement. Grade 1 corresponds to a view of all or most of the glottis, grade 2 to a view in which only the posterior portion of the glottis is visible, grade 3 to visualization of only the epiglottis, and grade 4 to inability to see the glottis or epiglottis at all [3].

• • 1. Definitions of unstable vital signs

The parameters determined in Rapid Emergency Medicine Score System were accepted as normal vital signs [5], according to the following parameters:

Mean arterial pressure (mm Hg): 70 to 109 Pulse rate (beats/min): 70 to 109

Respiratory rate (breaths/min): 12 to 24 Peripheral oxygen saturation (%): N89

• • 1. Definition of the standard intubation protocol applied to the patients

During the study period, all patients were intubated consi- dering the same airway management protocol. According

to this protocol, ‘Crash Airway Intubation Algorithm’ was applied to all traumatic or nontraumatic arrest patients. Arrest patients were intubated without giving any medications. All the patients other than arrest were intubated according to ‘Rapid Sequence Intubation algorithm.’ ‘Alternative airway management’ protocols were applied to the patients with whom the intubation failed either during the Rapid Sequence Intubation or crash airway protocol. Laryngeal mask airway and fiberoptic intubation were the alternative airway manage- ment protocols carried out. If all of these methods failed, “Failed Airway Algorithm” was applied to the patients by performing a cricothyrotomy with “Quick-Trach Emergency Cricothyrotomy Device.”

Table 1 The demographic features of study patients

Variable No. of patients %

Age (mean +- SS) 46.8 22.8 Sex Male/Female 251/115 68.6/31.4 facial trauma 48 13.1 Large incisors 18 4.9 Beard 39 10.7 Large tongue 34 9.3 Incisors distance b3 48 13.1 Hyoid/Mental distance b3 43 11.7 Thyroid to hyoid b2 38 10.4 Obstructed airway 29 7.9 Poor neck mobility 144 39.3 Causes of intubation Trauma 134 36.6 Acute respiratory distress 94 25.7 Cerebrovascular accident 52 14.2 Metabolic 50 13.7 Intoxication 32 8.7 Burn 4 1.1

Statistical analysis

Study data were analyzed with SPSS 16.0 for Windows. Continuous data were expressed as mean +- standard deviation and frequent data as percentiles. The comparisons of 2 groups with continuous and ordinal variable were performed by Student t test and Mann-Whitney U test, respectively, and the 2 group comparison of categorical data

by ?2. Kruskal-Wallis test was used for the comparison of 3 or more groups with ordinal variables and ?2 for categorical

variables. The independent factors in predicting difficult intubation were determined by logistic regression analysis. All hypotheses were constructed 2 tailed, and a P value of

.05 or less was accepted as significant.

Results

A total of 366 patients were included in the study. The mean age of the patients was 46.8 +- 22.8, and 68.6% (n = 251) of them were male. Eighty six (23.5%) patients were classified in the difficult intubation group and 280 (76.5%) patients in the easy intubation group. A patient flow chart was displayed in Fig. 2. Trauma (n = 134, 36.6%) was the most recorded indication of intubation followed by acute

Table 2 Intubation outcomes.

Variable n %

Complications Esophageal intubation 27 7.4 Right bronchial intubation 20 5.0 Vocal cord injury 1 0.3 Teeth fracture 1 0.3 The success of the intubation First 280 76.5 Second 66 18.0 Third 14 3.8 Fourth a 6 1.7

a Laryngeal mask airway, fiberoptic intubation, and Quick-Trach Emergency Cricothyrotomy Device.

Fig. 2 Patient flow chart. ?Independent variables in predicting difficult airway in logistic regression analysis. Significant analysis in univariate analysis.

??

Modified Lemon Scores	Easy intubation group, n (%)	Difficult group, n	intubation (%)	P
0	135 (48.2)	22 (25.6)
1	69 (24.6)	23 (26.7)
2	38 (13.6)	14 (16.3)
3	21 (7.5)	13 (15.1)
4	11 8 (3.9)	5 (5.8)	.000
5	5 (1.8)	5 (5.8)
6	1 (0.4)	4 (4.7)
Total	280 (100)	86 (100)
Median	1	1
Interquartile range	0-2	0-3

respiratory distress (n = 94, 25.7%) and cerebrovascular accident (n = 52, 14.2%). The demographic features of study population were shown in Table 1 and Table 2.

Table 3 Comparison of difficult and easy intubation

Variable	Difficult intubation		Easy intubation	P
Age	44	+- 20.7	47.7 +- 23.3	.191
Male	65	(75.6)	186 (66.4)	.110
Female	21	(24.4)	94 (33.6)
Facial trauma	13	(15.1)	35 (12.5)	.530
Large incisors	9	(10.5)	9 (3.2)	.015
Beard	9	(10.5)	30 (10.7)	.948
Large tongue	18	(20.9)	16 (5.7)	.000
Incisors distance b3	15	(17.4)	33 (11.8)	.174
Hyoid/Mental distance b3	18	(20.9)	25 (8.9)	.003
Thyroid to hyoid b2	20	(23.3)	18 (6.4)	.000
Obstructed airway	11	(12.8)	18 (6.4)	.056
Poor neck mobility	45	(52.3)	99 (35.4)	.005
Glasgow Coma Scale (median)	6		6	.789
Unstable vital signs
Blood pressure	40	(46.5)	138 (49.3)	.653
Respiration rate	47	(54.7)	182 (65)	.083
Oksi-Hb saturation	44	(51.2)	175 (62.5)	.061
Pulse rate	40	(20.2)	158 (79.8)	.106
Trauma	41	(47.7)	93 (33.2)	.015
Obesity	14	(16.3)	22 (7.9)	.022
Seniority of the resident (y)	4		2	.000

Of the 4 “LOOK” criteria of the LEMON method, large incisors (10.5% vs 3.2%, respectively; P = .015) and large tongue (20.9% vs 5.7%, respectively; P = .000) were likely related to difficult intubation. There was no difference between 2 groups with facial trauma (15.1% vs 12.5%,

respectively; P = .530) and beard (10.5% vs 10.7%, respectively; P = .948). Of the 3 evaluation criteria; “hyoid-to-mental distance less than 3 fingers” (20.9% vs 8.9%, respectively; P = .003) and “thyroid-to-hyoid distance less than 2 fingers” (23.3% vs 6.4%, respectively; P = .000) were both associated with difficult intubation, although the incisors distance less than 3 fingers (17.4% vs 11.8%, respectively; P = .174) did not have such an association. However the number of patients with decreased incisors

distance was higher in the difficult intubation group despite the lack of statistical significance. Of the 2 other criteria of the LEMON method, poor neck mobility (52.3% vs 35.4%, respectively; P = .005) was associated with difficult intubation; however, obstructed airway (12.8% vs 6.4%, respectively; P = .056) had a borderline P value that might be associated with difficult intubation.

Obese patients (16.3% vs 7.9%, respectively; P = .022) and patients with trauma (47.7% vs 33.2%, respectively; P =

Table 5 Modified Lemon Scores of the 2 intubation groups

.015) were more likely to have a difficult intubation, although these features were not included in the criteria of the LEMON method. The univariate comparison of difficult and easy intubation group was shown in Table 3.

Logistic regression analysis performed by the variables found to be significant in the univariate analysis revealed thyroid-to-hyoid distance less than 2 fingers (odds ratio, 3.34; 95% confidence interval, 1.35-8.27; P = .009) and the

large tongue (odds ratio, 2.3; 95% confidence interval, 0.97- 5.44; P = .057) with a borderline P value as the independent factors complicating the intubation (Table 4). Although the median values of the LEMON scores of the 2 groups (median values, 1 vs 1) were the same, there was a significant difference between 2 groups originating from the different distribution of the LEMON scores of 2 groups, which was displayed in Table 5 (interquartile range, 0-3 vs 0-2).

Cormack and Lehane classification was strongly related to difficult intubation. Intubation was more difficult from grade 1 to 4 (11% vs 25.2% vs 34% vs 81.8%, respectively;

Table 4 Logistic regression analysis for determining difficult intubation

Table 6 Relation between Cormack-Lehane classification and difficult intubation

Variable	Odds ratio	95% confidence interval	P
Large incisors	2.33	0.83-6.58	.108
Large tongue	2.30	0.97-5.44	.057
Hyoid/Mental distance b3	1	0.41-2.44	.992
Thyroid to hyoid b2	3.34	1.35-8.27	.009
Obstructed airway	1.40	0.58-3.4	.446
Poor neck mobility	1.57	0.6-4.17	.365
Obesity	1.16	0.49-2.75	.729
Trauma	1.03	0.39-2.72	.943

Variabl	e	Difficult intubation, n (%)	Easy intubation, n (%)	Modified LEMON score, median (IQR)
Grade	1	18 (11.1)	144 (88.9)	0 (0-1)
Grade	2	34 (25.2)	101 (74.8)	1 (0-2)
Grade	3	16 (34)	31 (66)	1 (0-3)
Grade	4	18 (81.8)	4 (18.2)	3 (1.75-5)
P		.000	.000
IQR indicates interquartile range.

Variable CL-1 CL-2 CL-3 CL-4 (modified LEMON)
Facial trauma 18	(11.1)	15	(11.1) 10	(21.3)	5	(22.7)
Large incisors 5	(3.1)	6	(4.4) 4	(8.5)	3	(13.6)
Beard 13	(8.0)	18	(13.3) 4	(8.5)	4	(18.2)
Large tongue 5	(3.1)	13	(9.6) 8	(17.0)	8	(36.4)
Incisors distance b3 10	(6.2)	20	(14.8) 10	(21.3)	8	(36.4)
Hyoid/Mental 6	(3.7)	18	(13.3) 7	(14.9)	12	(54.5)
distance b3
Thyroid to 3	(1.9)	14	(10.4) 9	(19.1)	12	(54.5)
hyoid b2
Obstructed airway 6	(3.7)	11	(8.1) 9	(19.1)	3	(13.6)

P = .000). In the post hoc analysis, all the groups differed from each other except grades 2 and 3 (P = .241). The relation between Cormack and Lehane classification and difficult intubation was shown in Table 6.

Table 7 Relation between Cormack-Lehane classification and demographic features

The relation between LEMON and Cormack-Lehane classification was displayed in Table 7.

An algorithm constructed by the high-risk variables in predicting difficult intubation was shown in Fig. 3.

Discussion

In a multicenter study, Sagarin et al [6] analyzed more than 6000 endotracheal intubation attempts in EDs. The first attempt was successful in 87% of patients [6]. And difficult intubation incidence according to ASA 2003 guideline based on the study by Sagarin was 13%, and cricothyrotomy was performed in 0.9% of patients.

In a study by Levitan et al [7], 456 patients were intubated by the emergency medicine residents in the ED, and the success rate at the first attempt was 86% with the cricothyrotomy ratio of 0.4%. However in our study, difficult intubation incidence was found in 23.5%, and cricothyr- otomy rate was 0.5%.

Mallampati classification can only be performed in patients in the sitting position, with head in a neutral position, mouth wide open, and tongue protruding to its maximum. Because it requires cooperation with patient and sitting position, ED patients are less suitable for this evaluation. Patients with multitrauma, altered mental status, or those unable to cooperate according to clinical situation cannot be evaluated with Mallampati classification. Reed et al reported that only 57% of patients intubated in the ED had been evaluated with Mallampati score [4]. In this study, only

52 (14.2%) patients were evaluated with Mallampati classification.

Reed et al had reported that Modified LEMON, which meant airway assessment without Mallampati could be an

applicable scoring system for evaluating difficult intubation in the ED patients [4]. In the study by Reed et al, difficult intubation was classified using Cormack and Lehane classification. According to this classification, grade 1 was classified as easy intubation, and grades 2 to 4 were classified as difficult intubation. In our study, the criteria reported in ASA 2003 guideline, “Difficult tracheal intubation: Tracheal intubation requires multiple attempts, in the presence or absence of tracheal pathology,” is accepted as certain difficult intubation criteria. The intubation attempts in our study were grouped as ‘easy or difficult intubation’ according to this criteria, and then, modified LEMON score and airway assessment score were calculated in these groups. The modified LEMON score was significantly higher in the difficult intubation group in our study. Although Reed et al classified easy and difficult intubation according to Cormack and Lehane classification as mentioned before, some authors regarded grades 1 to 2 as easy and grades 3 to 4 as difficult [8]. However, in our study, after scoring the patients with LEMON score, we categorized Cormack and Lehane classification as follows: grade 1 as easy, grades 2 to 3 as intermediate, and grade 4 as difficult intubation.

Although the modified LEMON scores of patients with difficult airway were higher than the patients with easy airway, all of the parameters used in modified LEMON method were not found to be significant in the univariate analysis comparing easy and difficult intubation except thyroid-to-hyoid less than 2 fingers, large incisors, large

Fig. 3 An algorithm constructed by the high-risk variables in predicting difficult intubation.

tongue, hyoid-to-mental less than 3 fingers, obstructed airway (with borderline P value), poor neck mobility, trauma, and obesity. The logistic regression analysis revealed only thyroid-to-hyoid less than 2 fingers as the independent variable predicting difficult intubation.

Facial distortion, secretions, swelling, mandibular injury, and potential cervical spine injury lead difficult intubation in trauma patients [9]. cervical collars and spinal immobiliza- tion also disrupt the laryngosgopic view as 20% [10]. In our study, the factors in trauma patients pointed out above were not found to be significant in patients with difficult intubation. And the clinical indications of intubation and vital signs of patients were not significant in predicting difficult airway.

It is emphasized in the previous studies that the training level of the practitioner can be an effective factor while determining difficult intubation criteria [4,6]. Sagarin et al

[6] reported that the success rates were 72%, 82%, and 88% at the first, second, and third years of the residency program, respectively; however, the success rate fell down again to 82% at fourth and later years. It was because of the fact that the residents were more commonly given supervisory roles in the fourth year of their residency. Thus, the fourth-year residents generally handle with a selected group of patients including the most difficult cases. The rate of rescue cricothyrotomies was highest after initial attempts by fourth-year residents (2.1%), and this supported our hypoth- esis. In this study, we observed that the median rate of the practitioners’ seniority was 4 years for difficult intubation group and 2 years for easy intubation group. That difference in our study had the same underlying basis reported by Sagarin. Comparing the success rates of experienced and inexperienced residents would not reflect the accurate results; instead, it would reflect a potential bias in clinical practice on the seniority of the residents.

Conclusions

Although the thyroid-to-hyoid distance less than 2 fingers is the only independent variable in predicting

difficult intubation, large incisors, large tongue, hyoid-to- mental distance less than 3 fingers, obstructed airway, poor neck mobility, trauma, and obesity also affect the difficulty of the intubation. Higher modified LEMON scores are associated with difficult intubation. The difference between the grade 1 and grade 4 of Cormack and Lehane score is clear in predicting easy and difficult intubation; however, grades 2 and 3 are similar. Mallampati classification is not a useful tool in classifying the difficult intubation in the ED that the LEMON acrostic can be modified to “LEON.”

References

1. Caplan RA, Benumof JL, Berry FA, et al. American Society of Anesthesiologists Task Force on Management of the Difficult Airway. 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. Anesth 2003;98(5):1269-77.
2. Walls RM. The emergency airway algorithms. In: Walls RM, editor. Manual of airway management. Philadelphia: Lippincott, William and Wilkins; 2000. p. 16-25.
3. Cormack RS, Lehane J. Difficult tracheal intubation in obstetrics. Anesth 1984;39:1105-11.
4. Reed MJ, Dunn MJG, McKeown DW. Can an airway assessment score predict difficulty at intubation in the emergency department? Emerg Med J 2005;22:99-102.
5. Olsson T, Terent A, Lind L. Rapid Emergency Medicine Score can predict long-term mortality in nonsurgical emergency department patients. Acad Emerg Med 2004;11(10):1008-13.
6. Sagarin MJ, Barton ED, Chng YM, et al. National Emergency Airway Registry Investigators. Airway management by US and Canadian emergency medicine residents: a multicenter analysis of more than 6,000 endotracheal intubation attempts. Ann Emerg Med 2005;46 (4):328-36.
7. Levitan RM, Rosenblatt B, Meiner EM, et al. Alternating day emergency medicine and anesthesia resident responsibility for management of the trauma airway: a study of laryngoscopy performance and intubation success. Ann Emerg Med 2004;43:48-53.
8. Orebaugh SL. difficult airway management in the emergency department. J Emerg Med 2002;22(1):31-48.
9. Walls RM. Management of difficult airway in the trauma patient. Emerg Med Clin N Am 1998;16:45-61.
10. Nolan JP, Wilson ME. orotracheal intubation in patients with potential c-spine injuries. Anesth 1993;48:630-3.