A Prospective Cohort Study on the Respiratory Effect on Modified Mallampati Scoring

Background Mallampati scoring is a common exam method for evaluating the oropharynx as a part of the airway assessment and for anticipation of difficult intubation. It partitions the oropharynx into 4 categories with scores of 1, 2, 3, and 4. Even though its reliability is known to be limited by confounding factors such as patient positioning, patient phonation, tongue protrusion, and examiner variability, the effect of respiration, i.e., inspiration and expiration, has not yet been formally studied. Methods Mallampati scores were collected from 100 surgical patients during both inspiration and expiration and later compared to the score obtained in the medical record, determined by a board certified anesthesiologist. Results Score deviations from the medical record reference were compared for both inspiration and expiration showing that respiration affects Mallampati scores; for some patients, the scores improved (i.e., decreased), while in others they worsened (i.e., increased). The respiratory change effect was quantified and visualized by plotting the area under the curve of the histogram of the deviations. 42% of the patients had a worsening of scores by 1 or 2 points with inspiration while 36% of the patients had a worsening of scores by 1 or 2 points with expiration. Conclusions Mallampati scoring is commonly used in evaluating the oropharynx as a part of the airway assessment and as a screening tool for difficult intubations. However, as this study points out, the respiratory cycle substantially affects the Mallampati scoring system, with significant deviations of 1 or 2 points. In a scoring system of 4 score categories, these deviations are remarkable.


Introduction
In 1983, Rao Mallampati frst published a brief observational statement that large tongue volumes, which concealed the uvula, faucial pillars, and soft palate on examination of the oropharynx, were more likely to have a small angle between the tongue and the larynx rendering it difcult to access the larynx during intubation [1].Mallampati stated in part "in the great majority of orotracheal intubation difculties I have encountered . . .this clinical sign (concealment of faucial pillars and uvula by the tongue) was helpful in predicting intubation difculty."[1] Two years later, Mallampati published his fndings concluding that the visibility of three pharyngeal structures, faucial pillars, soft palate, and uvula, could be used preoperatively as a proxy for laryngeal visualization during intubation [2].It was then that Mallampati's three diferent classes were established.Class I represented visualization of all three pharyngeal structures; Class II represented visualization of only the faucial pillars and soft palate, and Class III only showing the soft palate.In 1987, Mallampati's class system was further updated by Samson and Young to include Class IV, a class which shows no visualization of any pharyngeal structures [3].It is this modifed classifcation that is commonly used for preoperative bedside airway evaluations in anesthesiology today.
Te modifed Mallampati class system has since become entrenched in the clinical airway assessment.Te Mallampati score serves as a proxy in assessing how much physical space is available in the mouth for instrumentation during laryngoscopy and displacement of the tongue anteriorly.Given that an array of other clinical parameters afect the intubation process, it is not surprising that the reported sensitivity of the Mallampati score in identifying difcult intubations varies widely (from 11% to 84%) [4][5][6][7].Furthermore, studies have even questioned its efcacy at predicting difcult intubation [8,9].Additionally, there are several variables that can afect Mallampati scoring that need to be taken into account (Table 1).Examiner variability is a consideration when using Mallampati [14].Another is patient positioning, i.e., supine vs. sitting, which has been shown to infuence Mallampati scores in predicting difcult tracheal intubations [10].Tongue protrusion can also modify Mallampati scores [11].Finally, patient phonation also infuences Mallampati scoring by improving the predictability of difcult intubation [12,15].Because of its limitations, the Mallampati score has poor accuracy at identifying difcult intubations when used as the sole metric [9].Nonetheless, Mallampati scores continue to be a very commonly used metric so there is value in identifying further limitations.
Te pharyngeal airway is dynamic and changes shape and size during the respiratory cycle through inspiration and expiration [16,17] such that the pharyngeal cross-sectional area increases during early expiration and decreases during late expiration with a net result of pharyngeal collapse throughout expiration [17,18].Tis is explained by increased activity of upper airway dilator muscles, particularly, the genioglossus, during inspiration and less activation of these muscles during expiration [17].Te activation of these muscles prevents airway collapse by countering the negative pressure generated in the pharynx during inspiration [17][18][19][20].
For that reason, in the current investigation, we investigated the efect of respirations on Mallampati scores.To test the hypothesis, we analyzed and compared Mallampati scores in the same patients during inspiration and expiration to determine the efect of the respiratory cycle on this scoring system.

Ethics Approval.
A trained member of the anesthesia team conducted the airway assessments on surgical patients at University Hospital in Newark, New Jersey, USA.University Hospital is a 519-bed academic medical center with an active medical staf of more than 785.It is a Level I Trauma center in the state of NJ and a tertiary care center, often acting as a referral center from neighboring hospitals.It mainly cares for its local community of Newark, NJ, an underserved, racially diverse, impoverished community.Te hospital performs about 12,000-13,000 surgeries per year.Since this study posed no invasive risk to any of the patients, the study was classifed by the Newark Health Sciences Institutional Review Board as "exempt" from review.Verbal informed consent was obtained from each patient included in this study.

Conduct of the Study.
A total of 100 surgical patients were included in this study.All patients undergoing surgery at University Hospital in Newark, New Jersey, were included in this study.Te majority of the patients included in this study underwent orthopedic, neurosurgical, endoscopic, plastic, or cardiovascular surgeries/procedures. Children, pregnant women, prisoners, and non-English speakers were excluded from the study.Airways evaluations were based on the Mallampati methodology described in the literature and outlined briefy below [1,3].Patients were instructed to sit upright in the hospital bed or stretcher with their torso at a 90-degree angle.With the patient's head in the neutral position, patients were instructed to open their mouth as wide as possible and maximally protrude their tongue.While the patient's tongue remained protruded, patients were instructed to inhale and hold their breath while the inspiration Mallampati score was obtained.Ten, in the same view, patients were instructed to exhale and the expiration Mallampati score was obtained.Te examiner was positioned opposite the patient so that their eye level was at the level of the patient's oropharynx which was then visualized with a penlight.Te airway was classifed based on the modifed four-class Mallampati scoring system, and the results were documented.Te baseline Mallampati score, the score documented by the anesthesia team, was used as a reference to which both the inspiration and expiration scores of the same patient were compared to.Score deviations from the reference were considered positive when clinically more reassuring and closer to a score of 1, or negative when less reassuring and closer to a score of 4. For example, if a patient's reference Mallampati score was 2, but was 3 during inspiration, we would score that change as −1, as the score went from 2 to a worse score of 3. Scores were then partitioned into the following seven categories: worse: −3, −2, and −1, same: 0, or better: 1, 2, and 3. A histogram was then generated for both inspiratory and expiratory scores, and the area under the curve (AUC) was quantifed.A similar analysis was conducted for the obese patients' subset (BMI > 30; n � 36).

Statistical Analysis.
Demographic data were stratifed for age, race, sex, BMI, and insurance status, and Chi-square analysis was used to determine diferences in the population data.

Sample Demographics.
Sampled patients were mainly middle-aged with the majority falling within the age range of 40-59 years old (Table 2; p < 0.001).Biological sex was equally split, with 50% male and 50% female patients (p � 0.68).Race was predominantly Hispanic and Black (72%, p < 0.001).Te majority of patients had Medicare insurance and had a BMI of 25-29.9.A signifcant portion of the population was obese (36%) with roughly 8% being morbidly obese (BMI > 40).

Respiratory Efect on Mallampati Scores of Obese Patients.
From the sample, 36 patients were obese.Similar to the total study population, the change from the overall score for inspiration (Figure 3

Discussion
To our knowledge, this is the frst study to demonstrate a change in Mallampati scores through the respiratory cycle.
A scoring system of 4 scores with deviations of 1 or 2 score points can only be of limited value.Te portion of deviations is remarkable; 42% of the patients had worsening of scores by 1 or 2 points with inspiration, while 36% of patients had a worsening of scores by 1 or 2 points with expiration.
When plotting the Mallampati scores during inspiration or expiration in comparison to those scores in the medical record, we noticed a signifcant change (Figures 2(a) and 2(b)).Tis is made especially apparent when the comparison is made to the null hypothesis (blue line) in which respiration does not afect Mallampati scoring.Te following data helps appreciate this efect: 44% of the patient population had a worsening of the Mallampati score with inspiration and 37% of the patient population had a worsening of the Mallampati with expiration.15% of the population had improvement of Mallampati score with inspiration, while 17% had improvement of the Mallampati score with expiration.
Worsening (increase) Mallampati score with expiration, as observed in this study, may be explained by pharyngeal changes that occur during expiration [17].Under normal physiology, the pharyngeal airway is up to three times more collapsible during expiration as compared to inspiration [17].Furthermore, there is a complete closure of the upper airway during mid-expiration [17,21,22].Tis increased collapsibility of the upper airway is due to the physiological decrease in genioglossus muscle tone during expiration [17,23] which, in turn, may lead to worsening of Mallampati scores.Tis muscle plays a key role in maintaining upper airway muscle tone and patency by countering two forces, the intraluminal negative pressure generated from inspiration and the extraluminal tissue pressure from surrounding structure [22,24].With regards to worsening of Mallampati scores with inspiration, it is possible that the weakening of this muscle could lead to airway collapse and manifest as a worse Mallampati score.
Obstructive sleep apnea (OSA) is a particular pathophysiology accompanied by dysfunction of upper airway muscles [25].OSA patients have increased resistance to airfow in the oro-nasopharynx even when awake and exhibit disordered breathing patterns during sleep [25].Interestingly, this increased resistance is even seen in OSA patients with normal, when compared to age controlled subjects, pharyngeal cross-sectional areas [25].Te etiology of this increased resistance is the dysfunction of the  upper airway dilator muscles [25].Tis dysfunction can lead to airway collapse during REM sleep which thereby further contributes to apneic episodes throughout the night.Not surprisingly, these patients who exhibit airway collapse with sleep-disordered breathing during REM also tend to have airway collapse during mask general anesthesia [26].Patients who display worsening Mallampati score with inspiration may be more likely to have underlying OSA.In fact, it was even suggested that Mallampati scoring could be used to predict OSA presence and severity [13].However, any such approach would have to take in to consideration the signifcant respiratory efect on Mallampati scoring.In our studied obese patients, the respiratory change in Mallampati scores was more dramatic for both inspiration and expiration (Figures 3(a It is well established that obesity and the associated anatomical and physiological changes can alter proper air movement through the upper airway [27,28].Te Mallampati score improves after bariatric surgery, suggesting that anatomical diferences in the obese population afect respiration [28].In this present study, the obese patient group had a higher percentage of worsening of Mallampati scores with expiration than the general study population.Tere are several anatomical factors in these patients that could account for this diference.Obese patients tend to have more soft tissue in the cervical neck and hypopharyngeal region, and this can contribute to the collapse of the pharyngeal walls [28].It can lead to gradual increases in the concavity of the posterior epiglottis which can be used as a qualitative measure to determine the chronicity of airway collapse in obese patients [29].Additionally, the hyoid bone in obese patients is positioned lower than that of healthy controls [28,30].A lower position of the hyoid bone pushes the tongue vertically which obscures the view of several oropharyngeal structures [28].All of these anatomical ferences can contribute to increased airway collapse during expiration resulting in score worsening.Te obese patients in this study also had a higher percentage of worsening of Mallampati scoring with inspiration.Since obesity is a risk factor for OSA, it is likely that at least some obese patients exhibit similar dysfunction of upper airway dilator muscles that OSA patients experience.Tis in turn can lead to weakening of this musculature and to distortion of the view of oropharyngeal structures and the airway.Te current study is not without limitations, most notably, the small sample size.While a larger sample size would have reduced any signifcant deviation, we believe that our sample size was large enough to prove our hypothesis, i.e., that there is a diference in Mallampati scoring with respiration.Another limitation was the demographics of our study population, which consisted of mainly middle-aged Black and Hispanic patients.While it is known that differences in scores exist between certain races, e.g., Asians have a higher Mallampati score than caucasians [31], it does not afect the integrity of the study because we strictly focused on change in score from baseline during inspiration and expiration.Terefore, baseline diferences between races are irrelevant as we focused on absolute change from each patient's individual baseline.

Conclusions
To our knowledge, this is the frst study to demonstrate variations in the modifed Mallampati scoring system with inspiration and expiration.We demonstrated that most patients had a change in Mallampati score during the respiratory cycle.Specifcally, most patients exhibited a worsening (increase) of Mallampati score with inspiration and expiration.Te staggering variations in Mallampati scores in the obese patient population when compared to the general study population could potentially be due to the anatomical changes that can alter air movement in obese individuals.Overall, our study demonstrated that the modifed Mallampati scoring system is greatly afected by respiration, leading to a worsening of scores in a signifcant proportion of the population.

2
Anesthesiology Research and Practice 3.2.Te Efect of Respiration on Mallampati Scoring.A review of the Mallampati score classes is shown in Figure 1, giving a brief description of each class (I-IV).Te change from each patient's reference medical record score to the score either under inspiration (Figure 2(a)) or expiration (Figure 2(b)) was plotted using a histogram and indicated by the black line.If Mallampati scores were independent of respiration, the expected change between the inspiratory, expiratory, or the medical record score would be minimal as approximated by the blue line.However, 42% of the patients had a worsening of scores by 1 or 2 points with inspiration while 36% of the patients had a worsening of scores by 1 or 2 points with expiration.Te observed respiratory efect on Mallampati scores can also be appreciated when measuring the AUC of both respiration plots (Figure2(c)).Te AUC decreases from 141, during inhalation, to 135.5 during expiration.
(a)) and expiration (Figure 3(b)) were plotted (black line) showing signifcant deviation from the ideal curve (blue line).Areas under the curve for both inspiration and expiration were measured.AUC increased from 50.5 during inhalation to 54.5 during expiration (Figure 3(c)).

Figure 1 :Figure 2 :
Figure 1: An illustrative summary of the diferent Mallampati scoring classes and how they are determined.
) and 3(b)) than that seen with the total sample of patients in the study.Specifcally, 44% and 56% of obese individuals demonstrated worsening of Mallampati scoring during inspiration and expiration, respectively, and 22% and 16% showed improvement of Mallampati scores during inspiration and expiration, respectively.Te diference between the entire patient sample and the obese subset is further exemplifed when comparing the change in the AUC when transitioning from inspiration to expiration.While the AUC decreased from inspiration to expiration in the total patient population (Figure2(c)), it actually increased in the obese group (Figure3(c)), further suggesting the variability of Mallampati scoring with respiration and the added factor of obesity.

Figure 3 :
Figure 3: A subset of obese patients from the total study population was examined.Te change in score between the overall score and the inspiration (a) and subsequently the expiration (b) scores was plotted (black line) and compared to the ideal curve in which Mallampati is unafected by respiratory status (blue line).Te AUC was calculated, showing an increase in the AUC when transitioning from inspiration to expiration (c).

Table 1 :
Summary of variables that afect the Mallampati score.

Table 2 :
Demographic information with chi-squared analysis.