Predictive criteria for difficult tracheal intubation in a Moroccan population: A prospective observational study

Abstract

Background

Difficult tracheal intubation (DTI) may be a major cause of morbidity and mortality if not anticipated. Pre-anesthesia airway risk assessment is a must for safety reasons. The aim of this study is to analyze the clinical predictive criteria for DTI with particular focus on the Mallampati score in the supine and sitting positions.

Methods

This is a 6-month prospective observational study including all surgical adult patients requiring oro-tracheal intubation and not relevant to an indication for fibroscopic intubation nor urgent surgery (N = 500). DTI predictive criteria, composite scores, mask ventilation, laryngoscopy, and glottic catheterization were collected. Primary outcome was the ability of the collected criteria to predict DTI. Secondary outcome was the comparison of the Mallampatti score in the sitting vs. supine position to predict DTI using Cohen’s kappa non-parametric test.

Results

Mean age was 46.5 years. 10 % had diabetes. The incidence of DTI was 26 %. Six predictive factors for DTI were identified in multivariate analysis ( p < 0.05): body mass index greater than 30 kg/m ² , mouth opening <3.5 cm, thyromental distance <6.5 cm, grades III and IV of Mallampati sitting and supine scores and Cormack and Lehane score IV. Mallampati supine score outperformed Mallampati sitting score in predicting the risk of difficult intubation, with respective ROCs at 0.898 (IC95 % 0.857–0.938) vs 0.751 (IC95 % 0.678–0.824).

Conclusions

Incidence of DTI was higher in our population, probably related to our specific context of care. Predictive factors of DTI are consistent with previous studies while composite scores are not conclusive. The supine Mallampati performed better than sitting Mallampati in predicting DTI and may be reliable pre-operatively to assess the airway in patients unable to sit.

1 Introduction

Airway assessment and management are fundamental skills to an anesthesiologist whose main priority remains oxygenation. Tracheal intubation is an everyday, yet a risky anesthetic procedure. Anticipation through local or international algorithms (Difficult Airway Society, American Society of Anesthesiologists…) contributed to a decrease in airway management-related accidents. Ranging from 1.5 % to 18 %, difficult laryngoscopy and tracheal intubation may be fatal, particularly when urgent and unanticipated. Given its incidence and consequences, it is important to establish effective predictive criteria for difficult tracheal intubation (DTI). However, assessing the intubation difficulty raises several challenges. Definitions vary and are not standardized across studies comparing DTI predictive factors. They do not apply to inexperienced operators and optimal intubation conditions are not often specified (head position, laryngeal mobilization, depth of anesthesia or curarization). Mallampati’s classification, when considered alone, is not predictive of DTI. It is operator dependent and has poor correlation with Cormack and Lehane’s grades 2 and 3. ^, Even the DTI scores combining several anatomical criteria show poor predictive power. Several studies investigated the predictive assessment of difficult intubation in anesthesia while comparing Mallampati score in sitting and supine positions. The few available studies concluding that Mallampati in supine position was reliable and even superior to sitting position are mostly monocentric and the results cannot be extrapolated to other populations and settings. Some concluded that the supine position is an equally good predictor of DTI as the positive predictive value of both positions is similar, while the sitting gives a more accurate negative predictive value. Others add to the ongoing debate with some contrasting or non-conclusive results. ^, Being the result of a complex combination of patient’s anatomical characteristics, clinical context (position, optimal anesthesia, etc.…), anesthesiologist’s expertise (anticipation, planning, skills…) and available airway equipment resources, DTI should be addressed according to a contextualized procedure specific to each population and context of care. Moroccan studies focusing on difficult intubation are scarce if not inexistent. The aim of this study is to analyze the clinical predictive criteria for DTI with particular focus on the Mallampati score in the supine and sitting positions.

2 Patients and methods

Study design: this is a 6-month prospective observational study (October 2019 – March 2020) on predictive criteria for difficult intubation in surgical adults conducted in a Moroccan university hospital (local institutional review board number 07/20). The report of the present study followed the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines for observational studies . Adult patients scheduled for elective surgery in our department (thoracic, visceral, urological or vascular surgery) and requiring oro-tracheal intubation, were included. This study excluded pediatric, pregnant, emergency surgical patients and those with planned fiberscope intubation. Difficult intubation and ventilation definitions were adapted from the French Guidelines. A difficult intubation was defined as requiring more than two laryngoscopies and/or the use of an alternative technique, after optimization of head position, with or without external laryngeal manipulation. The French guideline excludes operators in the learning phase from the definition without further precision. In our study, we did include anesthesiology residents with >2 years’ training experience. A difficult mask ventilation was defined after optimal position (jaw thrust and oropharyngeal airway insertion), if one of the following was present: insufficient chest ampliation, an insufflation pressure of at least 25 cm H ₂ O, a need for repeated oxygen bypass, impossibility of maintaining SpO ₂ above 92 %, intervention of a second operator. Difficult laryngoscopy was identified as grades III and IV of Cormack-Lehane . Mallampati score in sitting position was assessed in a sitting patient with a head in neutral position, a mouth fully open, a tongue pulled forward and without phonation. Mallampati score in supine position was assessed in a patient lying on the operating table, neck in neutral position, mouth in maximum opening, tongue pulled forward and without phonation. The observer was placed at the head of the table with eyes placed vertically in the center of the open mouth. The two scores were assessed by the same operator not involved in the study. Induction of anesthesia was carried out according to a protocol tailored to the patient’s condition and surgical procedure. Cases of DTI were managed according to our local protocol (accessible checklist and DTI box). DTI was reported to the patient and/or family and recorded on both patient anesthesia file and computerized medical record, specifying the cause, management and subsequent incidents.

Data collection: Classic difficult intubation predictive criteria (Mallampati score in a sitting position, history of sleep apnea, presence of beard, dentition, cervical mobility, thyro-chin distance, retrognathism, etc.) and composite scores (Naquib, adnet, Arné, El Ganzouri, Wilson) were collected during the pre-anesthetic consultation. The Mallampati score in supine position was recorded in the operating room prior to anesthetic induction. Explanatory diagrams were included in the evaluation sheet to guide assessment. Mask ventilation, laryngoscopy, and glottic catheterization during the anesthetic induction were also recorded among other variables. The primary outcome of this study was the ability of the collected criteria to predict DTI. The secondary outcome was the comparison of the Mallampatti score in the sitting vs. supine position to predict DTI.

Data analysis: Data input was performed on Excel files, and statistical analysis was conducted using Epi info v7 software. Descriptive analysis was carried out to calculate percentages for qualitative variables and means ± standard deviation for quantitative variables. The univariate analysis used the chi2 test to compare percentages and the z-test to compare means. Multivariate analysis was performed using conditional logistic regression, with adjusted odds ratios and their 95 % confidence intervals calculated to determine factors predictive of difficult tracheal intubation. The significance threshold was set at 0.05. Correlation between the two Mallampati scales was assessed using Cohen’s kappa non-parametric test.

3 Results

Study population characteristics. Out of the 1200 patients admitted for elective surgery during the study period, 649 patients (54 %) underwent regional anesthesia and/or sedation without intubation, 50 patients underwent rapid sequence induction for urgent surgeries, and 1 patient underwent a planned fiberscope intubation ( Fig. 1 ). Thus, 500 eligible patients were included in the study, accounting for 42 % of the total admissions. The mean age in this study was 46.5 ± 17.2 years, with extremes ranging from 17 to 98 years. The Male/Female sex ratio was 1.08. 52 % of patients had an ASA score of 1 and 48 % had a 3–4 ASA score. Most of patients underwent visceral (44 %) or thoracic (31 %) surgeries. The main comorbidities reported were obesity (20 %), diabetes (10 %), obstructive sleep apnea syndrome (OSA: 6 %) and otorhinolaryngological pathology (6 %). No history of DTI, cervical trauma or burn was reported. Preanesthetic assessment for predictive criteria of DTI is presented in table 1 . No case of “cannot ventilate” was recorded. Difficult laryngoscopy, defined as grades III and IV of Cormack-Lehane, was reported in 28 % of our patients (140 patients) ( Fig. 1 ). According to Adnet score, mild and moderate to major difficulty were observed in 14 % and 20 % of the patients respectively ( Table 1 ). Difficult intubation was observed in 26 % of our patients (130 cases) and was unexpected in 70 % of those cases. 86 % of cases with difficult laryngoscopy (N = 120) had difficult tracheal intubation ( Fig. 1 ). The features of DTI were the following: external laryngeal manipulation in all 130 patients (100 %), more than two laryngoscopy attempts (of which duration was not reported) in 90 patients (69 %), use of a gum elastic bougie (aka Eschmann tracheal tube introducer) in 40 patients (31 %), involvement of a more experienced operator in 30 cases (23 %), the use of a video laryngoscope in three unexpected cases (2 %). There was no case of “cannot ventilate” nor “a failure to intubate” ( Fig. 1 ). All incidents reported during DTI management in our series were non-fatal and were made up of 14 cases of esophageal intubation, 10 cases of dental trauma and 7 cases of desaturation. No neurological consequences were reported during the follow-up.

Table 1

Airway conditions assessment: frequency of clinical indicators in our patients.

Predictive criteria for DTI		Number of patients (%)
Mouth opening	Good (3.5 cm)	490 (98 %)
Mouth opening	Limited (< 3.5 cm)	10 (2 %)
Thyroid-chin distance	Good (6.5 cm)	428 (86 %)
Thyroid-chin distance	Limited (< 6.5 cm)	72 (14 %)
Oral condition	Good	348 (77 %)
Oral condition	Toothless	84 (8 %)
Cervical spine mobility	Good	499 (99,8 %)
Cervical spine mobility	Limited	1 (0.2 %)
Beard	No	224 (94 %)
Beard	Yes	16 (6 %)
Retrognathism	No	500 (100 %)
Retrognathism	Yes	0 (0 %)
Protrusion of the upper incisors	No	432 (86 %)
Protrusion of the upper incisors	Yes	68 (14 %)
Diabetic prayer sign	Negative	500 (100 %)
Diabetic prayer sign	Positive	0 (0 %)
Mallampati score
In a sitting position	I	36 (7.2 %)
	II	232 (46.4 %)
	III	156 (31.2 %)
	IV	72 (14.4 %)
In a supine position	I	90 (18 %)
	II	316 (63.2 %)
	III	92 (18.4 %)
	IV	2 (0.4 %)
Composite scores
Naguib score ^a	< 0 = Easy	211 (42 %)
Naguib score ^a	> 0 = Difficult	289 (58 %)
Arné score ^b	11 = Easy	360 (72 %)
Arné score ^b	> 11= Difficult	140 (28 %)
El-Ganzouri score ^c	< 4 = Easy	310 (62 %)
El-Ganzouri score ^c	≥ 4 = Difficult	190 (38 %)
Wilson score ^d	< 2 = Easy	318 (64 %)
Wilson score ^d	≥ 2 = Difficult	182 (36 %)
Adnet score for laryngoscopy ^e	0 = Easy	330 (66 %)
	0 – 5 = Mild difficulty	70 (14 %)
	> 5 = Moderate to major difficulty	100 (20 %)

a Based on thyroid-chin distance, mallampati, mouth opening, heigh;.

b Based on history of DTI and contributing disease, respiratory symptoms, mouth opening, throid-chin distance, cervical spine mobility, mallampati;.

c Based on weight, head and cervical spine mobility, mouth opening, subluxation, thyroid-chin distance, mallampati, history of DTI;.

d Based on weight, head and cervical spine mobility, mouth opening, subluxation, retrognathism, upper incisor prominence;.

e Based on: number of attempts>1, number of operators > 1, Cormack and Lehane grading, traction force, laryngeal pressure, abducted or adducted vocal cords.

Univariate and multivariate analysis of risk factors for difficult tracheal intubation. Univariate analysis identified 19 variables that were significantly ( p < 0.05 ) associated with difficult tracheal intubation in our study ( Table 2 ). Multivariate analysis showed that DTI in our population was significantly associated with body mass index > 30kg/m ² (OR = 3.465), mouth opening <3.5 cm (OR = 2.166), thyroid-chin distance < 6.5 cm (OR = 0.813), Mallampati score III and IV in sitting position (OR = 2.467 and OR=1.339 respectively), Mallampati score III and IV in supine position (OR = 1.394 and OR=2.866 respectively), and Cormack and Lehane grade IV (OR = 1.867) ( Table 3 ). Multivariate analysis of the different predictive factors for difficult intubation according to Mallampati position (sitting or supine) identified three significant factors in both groups: body mass index > 30 kg/m2, thyroid-chin distance < 6.5 cm and Mallampati grade III ( Table 4 ).

Table 2

Univariate analysis of risk factors for DTI in our series.

Factors	Non DTI (N = 370)	DTI (N = 130)	p value
Age, mean ± SD	40.8 ± 17.7	54.2 ± 12.9	0.0001
Male gender, n (%)	196 (53 %)	64 (49.2 %)	0.603
Comorbidities , n (%)
Diabetes	30 (8.1 %)	20 (15.4 %)	0.093
Body mass index > 30 kg/m ²	39 (10.5 %)	61 (46.9 %)	0.001
Obstructive sleep apnea syndrome	4 (1.1 %)	24 (18.4 %)	0.0001
Rheumatological disease	18 (4.8 %)	0 (0)	–
Otorhinolaryngology disease	12 (3.2 %)	18 (13.8 %)	0.001
ASA I	208 (56.2 %)	50 (38.5 %)	0.9
ASA II	148 (40 %)	54 (41.5 %)	0.67
ASA III	14 (3.7 %)	24 (18.5 %)	0.0001
ASA IV	0 (0)	2 (1.5 %)	0.06
Anatomo-clinical criteria, n (%)
Mouth opening < 3.5 cm	0 (0)	10 (7.6 %)	0,0001
Thyroid-chin distance < 6.5 cm	24 (6.48 %)	48 (37 %)	0.001
Touthless oral condition	54 (14.5 %)	30 (23 %)	0.49
Limited cervical mobility	0 (0 %)	1 (0.8)	–
Beard	4 (1.08 %)	12 (9.2 %)	0.64
Retrognathism	500	0	–
Protrusion of upper incisors	20 (5.5 %)	48 (37 %)	0.002
Diabetic prayer test	50 (13.5 %)	0 (0)	–
Mallampati sitting position, n (%)
I	84 (22.7 %)	6 (4.6 %)	0.7
II	258 (67.7 %)	58 (44.6 %)	0.64
III	28 (7.6 %)	64 (49.2 %)	0.0001
IV	0 (0)	2 (1.5 %)	0.0001
Mallampati supine position, n (%)
I	36 (9.7 %)	0 (0)	0.61
II	228 (6.6 %)	4 (2.3 %)	0.08
III	94 (25.4 %)	62 (47.7 %)	0.001
IV	12 (3.2 %)	64 (49.2 %)	0.001
Cormack & Lohane score, n (%)
I	304 (82.1 %)	0 (0)	0.07
II	46 (12.4 %)	10 (7.8 %)	0.1
III	12 (3.2 %)	90 (69.2 %)	0.001
IV	8 (2.1 %)	30 (23 %)	0.001
Composite scores , n (%)
Naguib score > 0	174 (47 %)	115 (88.5 %)	0.001
Arné score > 11	30 (8 %)	110 (84.6 %)	0.001
El-Ganzouri score ≥ 4	100 (27 %)	90 (69.3 %)	0.001
Wilson score ≥ 2	80 (22 %)	102 (76.9 %)	0.001
Adnet score ≥ 5	70 (18.9 %)	100 (77 %)	0.001