Detection of respiratory events under non-invasive ventilation (NIV) depending on the type of flow measurement: Comparison of pressure-based versus ventilator-based flow measurement

Abstract

Background: Nocturnal non-invasive ventilation (NIV) is a standard therapy in chronic hypercapnic respiratory failure. Adjusting the ventilator, however, requires a high level of professional experience. Only polysomnography (PSG) in the sleep laboratory allows for thorough analysis of residual obstructions of the upper airways, central apneas, trigger failures, and mask leaks. The flow measurement under NIV is not easy from a technical point of view, especially since standardized interfaces are missing to incorporate the flow data into the PSG recording. Objectives: Comparison of a standard combined pressure/flow monitor with the signals derived from the respirator interfaces regarding the detection of apneas/hypopneas as well as NIV-specific events in the PSG. Methods: In this retrospective study, 20 recordings of patients under NIV who had previously been recorded in a standardized protocol were re-evaluated in a blinded manner. We compared three different methods of flow assessment (effort-based, pressure/flow monitor, and device-interface) with regards to the incidence of respiratory events. Results: By means of an easy to connect dynamic pressure measurement using a pressure/flow monitor, significantly fewer events could be diagnosed compared to the effort-based or device-interface flow determination. Also, NIV-induced events like patient ventilator asynchrony and unintentional leaks could only be detected reliably when using the interface signal. Conclusions: Measuring flow during PSG via a standard pressure/flow monitor did not provide sufficiently reliable results. The implementation of the respirator signals into the measurement via a device interface is therefore required.