Self-optimising respiratory therapy system

A method for a therapy system for operation of respiratory pressure therapy devices for treating respiratory disorders of the plurality of patients, the method comprising:

• generating communications between (a) a plurality of respiratory pressure therapy devices, each respiratory pressure therapy device comprising a controller comprising one or more processors, a motor operated blower coupled to the controller, and a communications interface coupled to the controller, the controller configured to control operation of the motor operated blower to generate a respiratory therapy at a patient interface coupled with the motor operated blower for a patient of the plurality of patients (b) one or more servers, the communications comprising;
• receiving, at the one or more servers, communications relating to respiratory therapy delivered to a patient via at least one respiratory pressure therapy device of the plurality of respiratory pressure therapy devices;
• generating, by the one or more servers, one or more responsive communications in response to an automated evaluation that assesses, with a plurality of programmed conditions, a compliance prediction about a progress of the respiratory therapy by the at least one respiratory pressure therapy device, wherein the compliance prediction is a score indicating a probability that the patient will be compliant with a predetermined compliance rule;
• and
• transmitting, by the one or more servers, the one or more responsive communications for control of the at least one respiratory pressure device according to a selected action to improve the respiratory therapy, wherein, according to the selected action, a communication of the one or more responsive communications, from the one or more servers, sends a control command to the at least one respiratory pressure therapy device that controls an operation of the at least one respiratory pressure therapy device, and wherein the selected action to improve the respiratory therapy is based on the compliance prediction.