Prediction of oral appliance treatment outcomes in obstructive sleep apnea: A systematic review

Summary

While oral appliances (OA) have demonstrated good efficacy in patients ranging from mild to severe levels of obstructive sleep apnea (OSA), this form of treatment is not completely effective in all patients. As a successful treatment response is not dependent solely on apnea hypopnea index severity, the prediction of OA treatment efficacy is of key importance for efficient disease management. This systematic review aims to investigate the accuracy of a variety of clinical and experimental tests for predicting OA treatment outcomes in OSA. A systematic literature review was conducted and the quality of the selected studies was assessed using the quality assessment of diagnostic accuracy studies (QUADAS-2) tool. Some 17 studies involving various prediction methods were included in this review. The predictive accuracy varied depending on the definitions of treatment success used as well as the type of index test. The studies with the best predictive accuracy and lowest risk of bias and concerns of applicability used a multisensor catheter. While a remotely controlled mandibular positioner study showed high accuracy, there was a high risk of bias. The available information on the validity of predictive index tests is very useful in clinical practice and allows for greater disease management efficiency.

Introduction

Obstructive sleep apnea (OSA) is a common syndrome that is characterized by recurrent episodes of partial or complete upper airway obstruction during sleep, resulting in sleep fragmentation and oxygen desaturation. OSA is associated with reduced quality of life, decreased cardiovascular health, and increased healthcare utilization and mortality [1], [2]. Continuous positive airway pressure (CPAP) is an efficient treatment for OSA and has been demonstrated to improve daytime symptoms and to reduce cardiovascular disease [3]. Although CPAP is highly efficacious in preventing upper airway collapse, patient acceptance, tolerance, and adherence are often low, consequently reducing effectiveness [4].

Treatment with an oral appliance (OA) is an alternative to CPAP for OSA and although less efficacious, it is more acceptable by patients. An American Academy of Sleep Medicine and American Academy of Dental Sleep Medicine clinical practice guideline recommends OA treatment for adult patients with OSA who prefer OA therapy or are intolerant of CPAP therapy [5]. A recent comprehensive review of OA treatment showed that a complete response occurred in around 48% of patients, with a range of 29%–71% among studies [6]. At present, patient selection for OA therapy is largely based on the apnea hypopnea index (AHI) severity alone. However, patients with severe OSA who successfully respond to OA therapy have also been reported [7], [8], [9]. Treatment recommendations based solely on AHI restrict a potentially preferred treatment option to a small portion of OSA patients. As the efficacy of OAs varies greatly in patients with OSA, the prediction of OA treatment response is of key importance for efficient disease management.

A number of studies have reported predictors of OA treatment outcomes using polysomnographic parameters ∗[10], [11], ∗[12], ∗[13], cephalogram [14], ∗[15], CPAP pressure [16], ∗[17], spirometer [18], drug-induced sleep endoscopy [19], remotely controlled mandibular positioner ∗[20], [21], and multisensor catheter parameters [22]. However all these studies are derivation studies rather than validation studies, which are lacking in the existing literature. While these methods still have some clinical importance, they vary greatly in terms of technical complexity, prediction accuracy, and clinical applicability and have not been systematically reviewed, which makes comparisons difficult.

This systematic review aims to investigate the accuracy of a variety of clinical and experimental tests in predicting OA treatment outcomes in OSA using the quality assessment of diagnostic accuracy studies (QUADAS-2) tool.