The Velum, Oropharynx, Tongue base and Epiglottis (VOTE) classification on drug-induced sleep endoscopy (DISE) is used widely for obstructive sleep apnea (OSA) syndrome, though research into comparative physical examinations with VOTE on DISE is still limited. The aim of this study was to evaluate the relationship between the findings of physical examinations and DISE in patients with OSA. Fifty-five patients with OSA were enrolled in this retrospective study. All of the patients received clinical explorations including a Brodsky classification, a modified Mallampati score (MMS), a modified Friedman’s staging system, and a Muller’s test. Drug-induced sleep endoscopy was further evaluated in the operating room. There were significant relationships between Brodsky classification, modified Friedman’s staging system, Muller’s test and oropharynx collapse during DISE (p < 0.05). Brodsky classification, MMS, modified Friedman’s staging system and retropalatal lateral-to-lateral (L–L) collapse of Muller’s test were significantly correlated with VOTE count (p < 0.05). The concordance between VOTE under DISE and Brodsky classification or modified Friedman’s staging system was moderate. In contrast, the concordance between VOTE under DISE and MMS or Muller’s test was slight. The study revealed that Brodsky classification and Friedman staging had a significant relationship with DISE on the velum and oropharynx, but the level of tongue base is uncertain between DISE and MMS. Correlation of awake evaluation of tongue base is still not correlated to the DISE findings. Pre-treatment evaluation of DISE is still warranted.

Background: In recent population-based case-control studies, sleep apnea was significantly associated with a higher incidence (hazard ratio, 1.71) of vertigo and the risk of tinnitus was found to increase 1.36 times in patients with sleep apnea. The possibility that obstructive sleep apnea (OSA) might affect neurotological consequences was not noticed, until studies using polysomnography (PSG) for these patients. Objectives: The purpose of this study was to investigate the relationship between vertigo and OSA. Methods: The collected data among patients from May 1st, 2018 to October 31th, 2018 at Shuang Ho Hospital. Eligibility criteria included an age older than 20 years, a diagnosis of obstructive sleep apnea. The diagnosis of OSA was defined as an oxygen desaturation index of at least 5, was established with the use of polysomnographic examination at hospital. Patients were excluded from the study if they had head injury, brain tumour, headache history and hearing loss. Patients who had vertigo were labeled as Vertigo group. In the other hand, patients who had no dizziness were labeled as control group. 58 patients were in the Vertigo group, and 113 were in the control group. Results: After PSG examination, 58 patients who had vertigo, were diagnosed OSA (29 males, average age = 57.07 years old, BMI = 26.64, RDI = 24.69, ESS = 8.65), and 24 patients of them (41.3%) were REM-related OSA. Meanwhile, in the control group, 113 patients had OSA (92male, average age = 49.66 years old, BMI = 26.06, RDI = 35.19, ESS = 11.43), and 18 patients of them (15.9%) were REM-related OSA (Table 1). Therefore, patient who had vertigo, would have higher proportion of REM OSA (P<0.001). Conclusions: The vertigo patients have a higher rate of REM-related OSA, and the acceptance rate to CPAP use is low. Further research is needed to explore novel therapeutic approaches, or combination of currently available non-CPAP therapies, in patients with REM OSA.

Objectives: Obstructive sleep apnea (OSA) is typically diagnosed by polysomnography (PSG). However, PSG is time-consuming and has some clinical limitations. This study thus aimed to establish machine learning models to screen for the risk of having moderate-to-severe and severe OSA based on easily acquired features. Methods: We collected PSG data on 3529 patients from Taiwan and further derived the number of snoring events. Their baseline characteristics and anthropometric measures were obtained, and correlations among the collected variables were investigated. Next, six common supervised machine learning techniques were utilized, including random forest (RF), extreme gradient boosting (XGBoost), k-nearest neighbor (kNN), support vector machine (SVM), logistic regression (LR), and naïve Bayes (NB). First, data were independently separated into a training and validation dataset (80%) and a test dataset (20%). The approach with the highest accuracy in the training and validation phase was employed to classify the test dataset. Next, feature importance was investigated by calculating the Shapley value of every factor, which represented the impact on OSA risk screening. Results: The RF produced the highest accuracy (of >70%) in the training and validation phase in screening for both OSA severities. Hence, we employed the RF to classify the test dataset, and results showed a 79.32% accuracy for moderate-to-severe OSA and 74.37% accuracy for severe OSA. Snoring events and the visceral fat level were the most and second most essential features of screening for OSA risk. Conclusions: The established model can be considered for screening for the risk of having moderate-to-severe or severe OSA.

Background: Continuous positive airway pressure (CPAP) therapy is the first-line treatment for obstructive sleep apnea (OSA). However, the low acceptance rate of CPAP remains a challenging clinical issue. This study aimed to determine the factors that influence the acceptance rate of CPAP. Methods: This retrospective cohort study was conducted at the sleep center of Shuang-Ho Hospital. Initially, 1186 OSA patients who received CPAP therapy between December 2013 and December 2017 were selected, and finally, 1016 patients were analyzed. All patients with OSA received CPAP therapy for at least 1 week, and their acceptance to treatment was subsequently recorded. Outcome measures included patients’ demographic and clinical characteristics (sex, age, BMI, comorbidities, history of smoking, and the medical specialist who prescribed CPAP treatment), polysomnography (PSG) results, and OSA surgical records. Results: Patients with a lower CPAP acceptance rate were referred from otolaryngologists (acceptance rate of otolaryngology vs. others: 49.6% vs. 56.6%, p = .015), in addition to having a lower apnea–hypopnea index (AHI) (acceptance vs. non-acceptance: 55.83 vs. 40.79, p = .003), rapid eye movement AHI (REM-AHI) (acceptance vs. non-acceptance: 51.21 vs. 44.92, p = .014), and arousal index (acceptance vs. non-acceptance: 36.80 vs. 28.75, p = .011). The multiple logistic regression model showed that patients referred from otolaryngology had a lower CPAP acceptance rate (odds ratio 0.707, p = .0216) even after adjusting for age, sex, BMI, AHI, REM-AHI, arousal index, comorbidities, and smoking status. Conclusions: Before their initial consultation, patients may already have their preferred treatment of choice, which is strongly linked to the type of medical specialists they visit, and consequently, affects their rate of acceptance to CPAP therapy. Therefore, physicians should provide personalized care to patients by exploring and abiding by their preferred treatment choices.

Objective: Current adjustments of continuous positive airway pressure (CPAP) may expose patients to risks of respiratory episodes or oxygen desaturation. Therefore, this study developed machine learning models using leading indicators, such as heart rate variability (HRV) and oximetry-related metrics, to proactively predict optimal adjustment timings. Methods: CPAP titration data were first collected from a sleep center in northern Taiwan. Subsequently, Continuous HRV and oximetry-related metrics were retrieved (60-s window with 1-s stride) and then labeled based on the presence of pressure adjustments. The dataset, comprising seven HRV and two oximetry-related parameters, was independently divided into training/validation (80%) and test (20%) datasets based on patient information. Five cross-sectional and two time-series models were established. The model with the highest accuracy and area under the receiver operating characteristic curve (AUROC) in the training/validation dataset was applied to the test dataset to investigate feature importance through permutation analysis. Results: A dataset comprising 14,629 time-series cases from 374 patients undergoing CPAP therapy was obtained. The InceptionTime model outperformed others during the training/validation phase with accuracies of 78.27% and AUROC of 78.07%, and achieved 80.92% accuracy and 76.52% AUROC in the test phase. Feature importance analysis identified peripheral arterial oxygen saturation, its standard deviation over a 60-s window, and normalized power in the very-low-frequency band of HRV as the most impactful predictors. Conclusions: The findings demonstrated the feasibility of incorporating HRV and oximetry-related metrics, as leading indicators, to proactively predict CPAP adjustment timings. Further research should consider integrating these metrics to improve CPAP therapy.