Airway assessment is essential to predict difficult intubation so, necessary arrangements to secure the airway could be done. Approximately 30% of the deaths occurred due to hypoxic brain damage secondary to inability to maintain a patent airway (Benumof 1991).

The incidence of difficult laryngoscopy and intubation who required general anaesthesia is 1.3% to 13%, while a higher incidence of up to 20% has been reported (Etezadi et al. 2013; Panjiar et al. 2019; Palczynski et al. 2018). In our study incidence of DVL ( difficult visualization of larynx) was found in 26 (17.3%) patients and no failed intubation was found. BURP was used in 28(18.6%) patients while bougie /styletwas used in 18(12%) patients. Similar incidence of DVL was found by Palczynski et al. (2018) (16.5%) and Allahayari et al. (2008) (18.2%) while Etazadi F et al. (7.3%), Jain N et al. (9.3%), Panjiar P et al. (10%), Rao KV et al. (8.2%), Yabuki S et al. (12%) found lesser incidence of DVL (Etezadi et al. 2013; Jain et al. 2017; Panjiar et al. 2019; Rao et al. 2018; Yabuki et al. 2019). This variation in incidence of DVL can be due to ethnic differences among populations, head position during laryngoscopy, external laryngeal manoeuvre, and varied standards used to define difficult laryngoscopy and intubation.

Although various airway tests have been devised to improve diagnostic accuracy but none of the tests individually have proven to be adequate. However, combining these tests have been proposed to improve their predictive value.

Etezadi et al. 2013 found that TMHT was a more accurate predictor of difficult laryngoscopy, rather than the pre-existing anatomical measurements. The area under ROC curve was used to find out the cut-off point for TMHT which was found to be 5 cm (Etezadi et al. 2013).

The cutoff value for TMHT in our study was found to be 5.15 cm by ROC curve. We performed the statistical analysis of TMHT by taking cutoff value of TMHT as 5 cm which was similar to the study done by Etezadi et al. In our study TMHT at a 5 cm cutoff had the highest sensitivity (76.92%), specificity (98.38%), PPV (90.90%), NPV (95.31%) and accuracy (94.67%) and was found the most accurate airway predictor test when compared with MMPG, IIG, ULBT, TMD and SMD. Similarly the sensitivity, specificity, PPV and NPV was found to be 82.6%, 99.31%, 90.47%, and 98.63% respectively by Etezadi et al. in a study done on 314 Iranian population (Etezadi et al. 2013).

Jain N et al. assessed the TMHT in 345 Indian patients undergoing coronary artery bypass grafting under general anaesthesia and found sensitivity, specificity, PPV, NPV and accuracy of 75%, 97%, 73%, 97% and 97% respectively which was similar to this study (Jain et al. 2017). Similar findings by Panjiar P. et al. revealed that the TMHT is the most accurate and best predictive test with sensitivity, specificity, PPV, NPV and accuracy 78.18%, 93.94%, 58.90%, 97.48%, 92.36% respectively in 550 Indian patients (Panjiar et al. 2019). Panjiar et al. and Jain N et al. also found similar cutoff point for TMHT (5.1 cm and 5.2 cm). Rao KV et al. reported sensitivity, specificity, PPV, NPV and accuracy 84.62%, 98.97%, 88%, 98.63%, 97.7% respectively in 340 Indian patients that also suggest TMHT is the most sensitive and accurate test for predicting difficult laryngoscopy (Rao et al. 2018).

However Yabuki et al. (2019) did a study on 609 Japanese patients and found a cut-off value for TMHT 5.4 cm without BURP and 5.0 cm with BURP. They also found different cut-off value of TMHT for men and women (5.5 cm and 5.6 cm). They found less sensitivity (49.3%), specificity (70.5%), PPV (18.6%), NPV (91.1%) and accuracy (68%) of TMHT at 5 cm cutoff value without BURP so they concluded TMHT to be a poor predictor of DVL when used alone which was contradictory to our results. Palczynski P et al. evaluated the TMHT in 237 patients from Poland and found less sensitivity (70%), specificity (70%), PPV (17%) and NPV (85%) (Palczynski et al. 2018). Selvi et al. also re-evaluated TMHT in Turkish population and found sensitivity (64.8%), specificity (78%), PPV (20.87%), NPV (96.1%) at 4.3 cm cutoff value of TMHT (Selvi et al. 2017). These studies also indicate that TMHT is not a good predictor of difficult laryngoscopy when used alone. This may be due to important factors like racial differences, different incidence of DVL and different cut-off value of TMHT.

MMPG being one of the most widely reported methods used for prediction of difficult laryngoscopy. Although this approach has a low predictive value when used alone, but it can be useful in a multivariate model to predict difficult laryngoscopy. The grading of MMPG is prone to error with phonation, susceptible for incorrect evaluation and interobserver variability. In this study MMPG had sensitivity (46.15%), PPV (66.66%), specificity (95.16%), NPV (89.39%) and good accuracy (86.67%) for prediction of difficult intubation which was less than TMHT. Panjiar et al. (2019) also found similar results while Yabuki et al. (2019) found less sensitivity, specificity and accuracy.

We observed high specificity (98.38%) and NPV (88.40%) of ULBT which predicts easy intubation more confirmatory while sensitivity (38.4%) and positive predictive value (83.33%) was nearly moderate which shows that this test is not much significant in prediction of difficult visualization of larynx. Badheka JP et al. found high sensitivity (96.44%), PPV (92.74%), specificity (82.35%), NPV (91.3%) and concluded ULBT is good predictor for difficult laryngoscopy (Badhekha et al. 2016). ULBT score of predicting difficult laryngoscopy has also some limitations. It is not appropriate for edentulous patients. The craniofacial structure of populations varies by ethnicity. The predictive value of ULBT must be calculated in each population independently for prediction of difficult laryngoscopy (Safavi et al. 2014).

IIG in our study found to be an inadequate predictor of difficult intubation due to low sensitivity but it could be a good predictor for easy laryngoscopy as it had high specificity. Similarly Rao KV et al. also found less sensitivity (69.23%), PPV (17.65%) and high specificity (71.03%), NPV (96.26%) for IIG (Rao et al. 2018).

TMD has been used as a predictor of difficult intubation from earlier days but was found to vary with the patient’s size and body proportion. In our study we observed that TMD has very high specificity (91.13%) and NPV (85.61%), though it had very low sensitivity (26.92%) and low PPV (38.89%). The accuracy of TMD in our study was found to be 80% which is lower as compared to other predictors for difficult intubation. Similarly Panjiar P et al. also found low sensitivity (20%) and low PPV (28.95%) while specificity (94.55%) and NPV (91.41%) for TMD was found to be high (Panjiar et al. 2019). Etezadi F et al. also found low sensitivity (21.73%) and low PPV (8.06%) and high specificity (80.41%) and NPV (92.65%) for TMD (Etezadi et al. 2013).

SMD had very low sensitivity (38.46%) and PPV (25%) but had higher specificity (75.81%) and NPV (85.45%) and the accuracy of SMD was found to be 69.33% which is very low as compared to other predictors of difficult intubation in our study. Similarly other studies (Etezadi et al. 2013; Jain et al. 2017) also found poor sensitivity and PPV while Palczynski P et al. and Basunia SR et al. found high sensitivity (60%, 60% respectively) which was contrary to our study (Palczynski et al. 2018; Basunia et al. 2013).

Limitations of the study: Study group included only patients scheduled for elective surgery with no history of significant difficult intubation. The validity indices of tests were measured only individually and hence combinations of tests were not measured for validity index. Cut-off value of TMHT for males and females was not calculated separately.

Future study should be done to define ethinicity, race, age groups and gender specific cut off values of TMHT for accurate validation of prediction in difficult laryngoscopy. Pilot testing for interobserver variability should also be planned.