BackgroundFailure to maintain an adequate airway can lead to brain damage and death. To reduce the risk of difficulty in maintaining an airway during general anesthesia, there are several known predictors of difficult intubation. People with a Malay background have different craniofacial structures in comparison with other individuals. Therefore, different predictors should be used for patients of Malay race.ObjectivesThe aim of this study was to determine the ability to predict difficult visualization of the larynx (DVL) in Malay patients based on several predictors, such as the modified Mallampati test (MMT), thyromental distance (TMD), and hyomental distance ratio (HMDR).Patients and MethodsThis cross-sectional study included 277 consecutive patients requiring general anesthesia. All subjects were evaluated using the MMT, TMD, and HMDR, and the cut-off points for the airway predictors were Mallampati III and IV, < 6.5 cm, and < 1.2, respectively. During direct laryngoscopy, the laryngeal view was graded using the Cormack-Lehane (CL) classification. CL grades III and IV were considered difficult visualization. The area under the curve (AUC), sensitivity, and specificity for each predictor were calculated both as sole and combined predictors. Logistic regression analysis was used to determine independent predictors of DVL.ResultsDifficulty in visualizing the larynx was found in 28 (10.1%) patients. The AUC, sensitivity, and specificity for the three airway predictors were as follows: MMT: 0.614, 10.7%, and 99.2%; HMDR: 0.743, 64.2%, and 74%; and TMD: 0.827, 82.1%, and 64.7%. The combination providing the best prediction in our study involved the MMT, HMDR, and TMD with an AUC, sensitivity, and specificity of 0.835, 60.7%, and 88.8%, respectively. Logistic regression analysis showed that the MMT, HMDR, and TMD were independent predictors of DVL.ConclusionsThe TMD, with a cut-off point of 65 mm, had superior diagnostic value compared with the HMDR and Mallampati score. Therefore, the TMD could be used in Malay patients to predict the difficulty of larynx visualization during laryngoscopy.