Appendix 3:
Detailed Principle of Operation of an OA Study
79
The purpose of this section is to provide a basis for understanding the purpose and operation of the MATRx plus
OA Study. The section should allow the Healthcare Provider (HCP) to develop a mental model of the device and its
operation.
Rationale and Strategy
MATRx plus is an automated system that provides information useful in guiding oral appliance therapy (OAT)
for patients with obstructive sleep apnea (OSA), a treatment that is efficacious in 50-70% of cases. The OA Study
does not diagnose OSA nor does it treat OSA. Rather, using a multi-night test it can be used to identify patients
who will be favourable candidates for OAT and, in those patients, determine a therapeutic mandibular protrusive
distance for a custom-fitted oral appliance. The test studies the patient during sleep while positioning the
mandible in a way that mimics the mechanical action of the therapeutic, custom OA.
The test strategy is to observe respiratory events that are the hallmarks of OSA and to protrude the mandible to
test positions where these disturbances may be eliminated or reduced to an acceptable rate. It is important to
understand that the MATRx plus control system does not attempt to continuously control airflow and oxygen
saturation at normal or reference values. Rather, the system waits for respiratory disturbances to happen, namely:
a cessation of breathing or severe changes in airflow. The system then protrudes the mandible to a next test
position and monitors recurrence of these transient respiratory disturbances. The continued occurrence of
respiratory events despite mandibular protrusion is a reflection of the patient’s ability to respond to the therapy.
However, the outcome is not simply based on the number of respiratory events at the final test position of
the mandibular positioner. An important feature of OAT for OSA is that efficacy varies through the night. This
principally arises from changes in body position and changes in sleep staging, most notably the transient
occurrence of rapid eye movement (REM) sleep which strongly determines the expression of OSA as respiratory
events. To counter this variability, MATRx plus gathers a comprehensive set of variables at the various protrusion
levels of the mandible across the multi-night test and calculates the reported outcomes using a machine learning
classifier.
GENERAL DESCRIPTION
By way of orientation to the MATRx plus system, we provide a general description of the physiologic closed-loop
control system, its application in three test conditions, and the machine learning classifier.
Closed-loop Control System
The common understanding of a closed-loop control system envisions a physiologic variable that is maintained
at a desired level by relating the variable to a reference variable in a closed-loop configuration. The MATRx plus,
however, differs in that the patient’s physiologic variables (O2 saturation and airflow) are not being controlled at
a normal value. Rather, these two physiologic variables are monitored continuously to assess the occurrence of
respiratory events and once a certain threshold has been reached, the controller protrudes the mandible. The
