Interobserver Consistency of Drug-Induced Sleep Endoscopy in Diagnosing Obstructive Sleep Apnea Using a VOTE Classification System.

Objectives: To explore the interobserver consistency of drug-induced sleep endoscopy (DISE) for patients with obstructive sleep apnea syndrome (OSAS) and review the current literature.

Methods: In total, 55 patients with an apnea-hypopnea index >5, as determined by on overnight sleep study, were included in this study, 45 males and 10 females, with an average age of 46.87 ± 10.06 years old (range, 19-71). For all OSAS patients, DISE was performed by the same surgeon, which was recorded digitally. The video recordings of DISE were evaluated independently by 3 experienced surgeons who were asked to note his or her decisions as the pattern, site, and degree of upper airway collapse using a VOTE (velum, oropharynx lateral wall, tongue base, and the epiglottis) classification system.

Results: Interobserver consistency in the diagnosis of velum-related obstruction in anteroposterior, lateral, and concentric configurations ranged from poor to good. Only significant interobserver consistency among observers A and B was obtained in the diagnosis of oropharynx-related obstruction in the lateral configuration (concordance 60.0%, kappa: 0.365, P < 0.05). Interobserver consistency in the diagnosis of the tongue-related collapse in an anteroposterior configuration, the epiglottis-related collapse in an anteroposterior and lateral configuration ranged from fair to moderate (all kappa values >0.20, all P values < 0.05).

Conclusion: Our data suggested that the interobserver consistency of DISE ranged from poor to good. Therefore, further studies with larger numbers of patients are needed to standardize DISE procedures, training, and interpretation.