Abstract
Background
We aimed to compare the efficacy of a new bedside screening test named acromioaxillosuprasternal notch index (AASI) with modified Mallampati (MMP).
Methods
A total of 603 adult patients, who were candidates for tracheal intubation in elective surgery, were enrolled in this prospective study. Preoperative airway assessment was carried out with AASI and MMP. The new AASI score is calculated based on the following measurements: (1) using a ruler, a vertical line is drawn from the top of the acromion process to the superior border of the axilla at the pectoralis major muscle (line A); (2) a second line is drawn perpendicular to line A from the suprasternal notch (line B); and (3) the portion of line A that lies above the point where line B intersects it is line C. AASI is calculated by dividing the length of line C by that of line A (AASI = C/A). After induction of anesthesia, the laryngeal view was recorded according to the Cormack–Lehane grading system. Receiver operating characteristic curve analysis was employed to compare between AASI and MMP.
Results
Difficult visualization of larynx (DVL, Cormack–Lehane III and IV) was observed in 38 (6.3%) patients. The best cutoff point for DVL was defined at AASI > 0.49. AASI had a lower false negative rate and higher predictive values (sensitivity, positive predictive value, and accuracy) in comparison with MMP.
Conclusion
AASI was associated with higher predictive values than MMP and could be used for estimation of DVL.
Keywords
anesthesia; intubation; intratracheal; laryngoscopy;
1. Introduction
Maintaining a patent airway following the induction of general anesthesia is undeniably the most imperative concern for an anesthesiologist. Unanticipated difficult intubation, especially when associated with difficult or lack of ventilation in anesthetized patients, is still the main cause of morbidity or mortality.1
The incidence of difficult laryngoscopy or tracheal intubation was reported to be in the range of 0.1–20.2%; this variation is due to the different patient populations and criteria used.2, 3, 4, 5, 6, 7, 8, 9 Prediction of difficult intubation in preoperative evaluation has been attempted by numerous investigators using simple bedside physical examinations based on anatomical landmarks such as modified Mallampati test (MMP), interincisive distance, thyromental distance (TMD), sternomental distance, upper lip bite test, and hyomental distance ratio,3, 7, 10, 11 all of which have shown different sensitivities and specificities.
In the authors' experience, difficult visualization of larynx (DVL) was observed in individuals whose neck was situated deep in the chest (i.e., with a sloping clavicle); therefore, to consider a bedside test based on surface anatomy is suggested. We observed that the portion of the arm–chest junction above the level of suprasternal notch might be used as an indicator to estimate DVL. This study was aimed to evaluate the predictive validity of a new index (based on the surface anatomy of the upper chest), called the acromioaxillosuprasternal notch index (AASI), and compare it with a previously established test (MMP) for assessing difficult laryngoscopic view in patients who were candidates for general anesthesia.
2. Materials and methods
After obtaining approval from the Institutional Review Board and written informed consent from patients, 603 consecutive adult Asian patients aged 20–65 years with ASA class Ι–ΙΙ, scheduled to undergo elective surgery requiring endotracheal intubation, were enrolled in this prospective observational study during the period February 2011–April 2012. Exclusion criteria were as follows: obvious anatomical abnormality, upper airway abnormality (e.g., tong tumor, maxillofacial tumor, or fracture), recent head and neck surgery, ASA class ΙΙΙ–ΙV, and disability to open the mouth. Each patient underwent physical examination prior to surgery, and AASI and MMP were assessed. A new AASI index was developed by the first author (M.R.K). With the patients lying in a supine position and their upper extremities resting at the sides of the body, AASI was calculated based on the following measurements: (1) using a ruler, a vertical line was drawn from the top of the acromion process to the superior border of the axilla at the pectoralis major muscle (line A; Fig. 1); (2) a second line was drawn perpendicular to line A from the suprasternal notch (line B); and (3) the portion of line A that lay above the point at which line B intersected line A was line C. AASI was calculated by dividing the length of line C by that of line A (AASI = C/A). MMP (the oropharyngeal view) was measured while patients were sitting, with a fully protruded tongue without saying “ah”. MMP classification was as follows: class Ι = soft palate, fauces, uvula, and pillars were visible; class ΙΙ = soft palate, fauces, and uvula were visible; class ΙΙΙ = soft palate and base of uvula were visible; and class ΙV = soft palate was not visible.8
Download full-size image
All patients received premedication with midazolam (0.03 mg/kg) and fentanyl (2 μg/kg). Anesthesia was induced with sodium thiopental (5 mg/kg) and atracurium (0.6 mg/kg). With the head in the sniffing position, laryngoscopy was attempted by an attending anesthesiologist blinded to the measurements following ventilation of the lungs with 100% oxygen. Laryngoscopy was performed after the loss of the fourth twitch in the train of four, with a Mackintosh blade (No. 3) and Cormack–Lehane grading was assessed. The laryngeal view was graded according to the Cormack and Lehane grading system: Grade Ι—full view of the glottis; Grade ΙΙ—glottis partly exposed, anterior commissure not seen; Grade ΙΙΙ—only epiglottis seen; Grade ΙV—epiglottis not seen (12). Grades Ι and ΙΙ were considered as easy visualization of larynx (EVL) and Grades ΙΙΙ and ΙV as DVL.
If the first intubation attempt failed and difficulty was encountered, intubation was attempted with Macintosh blade No. 4, coupled with adjustment of external laryngeal pressure and head position. All preoperative assessments including MMP and AASI were performed by an attending anesthesiologist. SPSS software version 16 (SPSS Inc., Chicago, IL, USA) was used to analyze the data. Predictive values of the abovementioned tests were determined. Sensitivity, specificity, positive and negative predictive values, accuracy, odds ratio, and positive and negative likelihood ratios were calculated. Receiver operative characteristic curves were used to compare AASI with MMP. Considering an incidence of difficult laryngoscopy of 5%, power of 80% with a type 1 error of 5%, to detect an improvement of discriminating power of an absolute value of 7% and using a two sided alternative hypothesis, 603 patients were estimated to suit the study. A comparison of the proportions of patients with DVL and EVL was performed by Chi-square test for diagnostic accuracy and t test for continuous independent variables. Two-sided p < 0.05 were considered statistically significant.
3. Results
A total of 603 patients were enrolled in the study. Patient characteristics are shown in Table 1. In total, 38 patients had a laryngoscopic view of Cormack–Lehane Grades III (30) and IV (8). The prevalence of difficult laryngoscopy was 6.3% [4.5–8.5%, 95% confidence interval (CI)]. Using discrimination analysis, AASI ≤ 0.49 was defined as the best cutoff point for difficult intubation. Predictive values of the two tests used are shown in Table 2. The main finding of this study was the area under the receiver operative characteristic curve for AASI (AUC = 0.89; 95% CI, 0.83–0.97) was higher than that of MMP (AUC = 0.74; 95% CI, 0.62–0.86; Fig. 2). Of the patients who underwent direct laryngoscopy with difficulties, 78.9% (54.4–93.9%, 95% CI) were identified correctly with an ASSI of ≤ 0.49. Among the patients who underwent direct laryngoscopy without difficulties, 89.4% (85.1–92.7%, 95% CI) were correctly predicted with ease. AASI had higher predictive values and a lower false negative rate than MMP (Table 2). Statistically significant differences were observed between sensitivity, positive predictive values, and accuracy of the two mentioned tests (p < 0.05), showing higher levels for AASI. Comparisons of specificity and negative predictive values between two tests did not depict significant differences (p > 0.05).
Download full-size image
4. Discussion
Our study revealed that AASI had higher predictive values (sensitivity, specificity, positive predictive value, negative predictive value, and accuracy) and a lower false negative rate than MMP. Management of a difficult airway is still the most challenging duty of anesthesiologists, and they are faced daily with the difficult obligation of determining the intubation facility.
Although patient's airway can be evaluated preoperatively through taking a detailed history, performing a physical examination, and using various anatomic measurements and features, there are still several reports with questionable true prediction possibility.12, 13, 14 Although prediction of difficult intubation preoperatively would be very useful, there is no single test with adequate sensitivity and a low false positive rate.15
In our study, the incidence of DVL was 6.3%. Cook16 reported that difficult intubation happened in three quarters of difficult laryngoscopic views compared with 3% of easy laryngoscopic views. Difficult intubation could be the final result of difficult laryngoscopy, which depends on the performer's expertise, patient characteristics, and facilities. Shiga and colleagues9 reported that the incidence of difficult intubation ranged between 1.7% and 20.2%. Bilgin and Ozyurt17 observed that most of the cases of difficult intubation had an improved laryngeal view upon application of external laryngeal compression; therefore, these cases were excluded from the difficult intubation group in some studies.
Variability in the incidence of difficult intubation has been attributed to various factors, including anthropomorphic features among populations, intubation protocols, degree of muscle relaxation, different grades of laryngeal view, head position, application of cricoid pressure, and type or size of laryngoscope blade.17, 18
Our findings showed insignificant differences in weight, height, and body mass index (BMI) between EVL and DVL patients (Table 1). These are in contrast to the findings of the studies performed in obstetric patients that have shown an association between patient weight and a reduced laryngoscopic view.19, 20 However, Brodsky et al21 had similar findings to ours, and found that neither obesity nor BMI was associated with difficult intubation. A large cohort study conducted by Lundstrom et al22demonstrated that BMI could be a statistically significant predictor of difficult intubation; although the predictive value of BMI was weak, it might be more appropriate than that of weight.
Our study revealed that all predictive values of the AASI were superior to those of MMP (Table 2). The sensitivity and specificity values of MMP in our study were 52.4% and 85.7%, respectively, which were similar to pooled measurements in a meta-analysis conducted by Shiga et al.9 A recent meta-analysis of the accuracy of the Mallampati tests reported significant differences in sensitivity and specificity values. Sensitivity of the MMP tests ranged from 0.12 to 1.00 and specificity ranged from 44% to 98%; finally, the summary sensitivity and specificity values were estimated to be 55% and 84%, respectively, and the overall accuracy was estimated to vary from poor to good, depending on the used version of related reference tests.23
Cattano and colleagues24 in a study of 1956 patients found that Mallampati classification only is insufficient for predicting difficult intubation. Krobbuaban and colleagues25 determined that Mallampati Classes III and IV had 70% sensitivity and 60% specificity, with a Positive Predictive Value (PPV) of 20.
Although the statistical cutoff point of AASI was 0.49 cm, clinically this can be approximated to 0.5 for simplification. A value of less than 0.5 can be used as a criterion for EVL, and that higher than 0.5 may be considered for DVL.
In our experience, sometimes AASI could be estimated visually and an exact calculation using a ruler was not required. When comparing the predictive values of the AASI test with those of MMP, one should bear in mind that in laryngoscopy and intubation, a correct diagnosis of difficult intubation cases is very important.
Therefore, diagnostic tests should be associated with a low false negative rate. In other words, sensitivity of a diagnostic test in this field should be maximized and associated with a reasonable specificity; this is the case for the AASI test. Sensitivity of the AASI test is higher than that of MMP.
Although the rate of difficult intubation is much lower than that of difficult laryngoscopy, the prediction of difficult laryngoscopic view ushers in estimation of difficult intubation and prevents irreversible damages to patients.
We presented AASI as a new test in this study; however, it should be noted that there is no single reliable test that can predict DVL, so anesthesiologists have to recruit several prevailing tests. Considering this, some models were presented using a combination of several measurements, such as Wilson (weight, head and neck movement, jaw movement, receding mandible, and buck teeth), Arne [previous history of difficult intubation, disease associated with difficult intubation, clinical symptoms of airway pathology, interincisor distance (IID) and mandible subluxation, TMD, maximum range of head and neck movement, and Mallampati], and Naguib [thyrosternal distance (TSD), Mallampati score, TMD, and neck circumference] models.26 In future, further studies are suggested to compare AASI with other tests.
In conclusion, AASI, a new diagnostic test, was shown to be a good predictor of DVL, with higher sensitivity, accuracy, and area under the curve when compared with MMP. Further studies are required to confirm our findings.