0549 Treating OSA With Oral Appliances: A Feedback Controlled Mandibular Positioner Prospectively Identifies Responders

Abstract

Introduction

We have developed a feedback controlled mandibular positioner (FCMP) incorporated into a home sleep test (HST) to predict outcome with oral appliance therapy (OAT). A previous clinical trial with prototype FCMP device demonstrated significant prospective predictive accuracy. The purpose of the present study was to assess prospective predictive accuracy of a final device (MATRx plus).

Methods

Patients with OSA (n=53; mean AHI: 29.5 hr^-1; mean BMI: 30.9 kg/m²) received a 2- to 3-night FCMP test in...

Abstract

Introduction

We have developed a feedback controlled mandibular positioner (FCMP) incorporated into a home sleep test (HST) to predict outcome with oral appliance therapy (OAT). A previous clinical trial with prototype FCMP device demonstrated significant prospective predictive accuracy. The purpose of the present study was to assess prospective predictive accuracy of a final device (MATRx plus).

Methods

Patients with OSA (n=53; mean AHI: 29.5 hr^-1; mean BMI: 30.9 kg/m²) received a 2- to 3-night FCMP test in the home, and all patients received custom OA by a blinded dentist. AHI and ODI values were measured at baseline and outcome (custom OA in place) in the home using the HST component of the device.

Results

Using a success criterion of ODI < 10 hr^-1, agreement analysis between the FCMP prediction and OAT outcome yielded sensitivity, specificity, positive and negative predictive values (%) for the present study of 90.5%, 90.9%, 97.4%, and 71.4%, for the previous study of 85.3%, 92.9%, 96.7%, and 72.2%, and for both studies combined of 88.2%, 90.0%, 97.3% and 71.9%, respectively. The overall accuracy of the two studies was comparable (present: 90.5%; previous: 87.5%). Use of AHI < 10 hr^-1 as a success criterion was available only in the present study, and the values for the above agreement parameters were 88.1%, 90.0%, 97.3%, and 64.3%, respectively.

Conclusion

The equivalence of the agreement parameters derived from these two independent studies suggests that the results of the FCMP test are generalizable to other OSA populations. We conclude that the final product embodying our FCMP technology provides a feasible home test that accurately predicts OAT outcome. These results also indicate that the FCMP test accurately predicts OAT outcome in an AHI framework.

Support (If Any)

Zephyr Sleep Technologies, Prosomnus Sleep Technologies.