Evaluation of the association between voice parameters and Mallampati classification

Abstract

Purpose: Difficult airway management still lacks an accurate predictor and subjective parameters such as modified Mallampati test may hamper its prediction leading to misclassification. Objective parameters, on the other hand, such as voice frequencies and formants (Figure 1) may improve the preoperative airway assessment by better evaluating the upper airway anatomy. Therefore, we aimed at observing whether or not objective voice parameters would be associated with Mallampati classification so that they could improve the assessment of upper airway anatomy. Methods: A prospective study with 453 patients scheduled for elective surgery under general anesthesia was performed. At transitional waiting hall before transport to the operating room, we collected data on sex, age, weight, height, ASA physical status, Body Mass Index, modified Mallampati test, and acoustic parameters from voice. Uni and multivariable analyses were conducted. Two logistic regression models and their predictive performances were determined. Results: The main result was the association found between formant if5 (p=0.00) and modified Mallampati test. ASA physical status was also related to Mallampati classification as well as age. Obesity, sex, weight, height, and BMI did not show any association with the Mallampati test. The AUC for the regression model containing all formants together was 65.0%, whereas the AUC for the model containing only three formants (af2, af4, and if5) after a stepwise was 60.2%. Conclusions: Voice formants are objective measurements associated with both upper airway anatomy and Mallampati classification. Such as, voice analysis may improve the upper airway assessment performed by the conventional modified Mallampati test.