0566 DRUG-INDUCED SLEEP ENDOSCOPY AND SURGICAL OUTCOMES: AN INTERNATIONAL, MULTICENTER COHORT STUDY

Abstract

Introduction: Successful surgical management of obstructive sleep apnea (OSA) is based upon determining the pattern of upper airway obstruction and design of a tailored treatment. Drug-Induced sleep endoscopy (DISE) is an evaluation using pharmacologic sedation designed to simulate natural sleep. DISE has shown promise in studies at single institutions or single interventions. The objective of this study was to examine the association between DISE findings and surgical outcomes across multiple institutions and surgical procedures. Methods: This was a retrospective cohort study of adults who underwent surgery for obstructive sleep apnea, with preoperative recorded DISE video and both preoperative and postoperative sleep studies. Patients were excluded if they had a history of any prior hypopharyngeal surgery other than tonsillectomy. The preoperative recorded DISE videos were reviewed by 4 reviewers in a blinded fashion using the VOTE Classification. Surgical outcomes were defined by change in the apnea-hypopnea index and the commonly- used definition of surgical response. Multiple linear and logistic regression evaluated potential associations between preoperative DISE findings and surgical outcomes. These analyses were performed for the entire cohort and for subgroups defined by specific DISE findings and surgical interventions. Statistical adjustment was performed for demographic and specific physical examination findings. Results: Fourteen institutions contributed a total of 628 study participants. Two DISE findings were broadly associated with poorer outcomes: complete concentric collapse related to the velum/palate and complete oropharyngeal lateral wall collapse. Soft palate surgery outcomes were better in cases with isolated soft palate obstruction. Multilevel surgery that matched multilevel obstruction on DISE improved outcomes in selected instances. Conclusion: DISE may characterize specific patient subgroups that may not respond well to surgery. DISE may also improve the design of effective surgical treatment plans.