2 9
New OSA Patient Simulator for Testing PAP Devices
To design the new OSA simulator model we developed a novel approach aimed at realisti- cally replicating a typical night of sleep for a female patient. With this aim, we considerably expanded our library of disturbed breathing patterns anonymously extracted from polysomno- graphy recordings obtained from real OSA patients and we incorporated several new adjustable features into the simulator. Specifically, the new patient model can be set to react to the pres- sure delivered by the APAP device (PAP-responsive mode) or to reproduce a fixed scenario of disturbed breathing events (Steady mode), depending on the device characteristics being tested. Moreover, the severity of the simulated OSA profile is now fully modifiable by changing the frequency and duration of each breathing event. Various artefacts were introduced into the event spectrum, such as changes in tidal volume and breath rate, to replicate typical events dur- ing wake such as irregular breathing, swallowing, moving and talking. By combining these new features, we aimed to create a new OSA model concept model that can realistically replicate a whole night of sleep, including phases of wake, rapid eye movement (REM) and non-REM sleep, and change in body position, each one designed to mimic different characteristics in terms of upper airway collapsibility. For this study specifically, as an example of an entire night of sleep-disordered breathing (SDB), the bench test model was set to simulate the disturbed patterns of a female OSA patient with the following characteristics: long sleep latency (45 min), low positive airway pressures (PAPs) required to overcome obstructive events, high proportion of flow limitation events ver- sus apneas, higher apnea-hypopnea index (AHI) during REM sleep, and only minor positional effects on upper airway collapsibility. The features and structure of this female-specific OSA patient simulation are detailed in Table 1. The breathing pattern of the simulated patient depended on the PAP applied by the device under test, with a total duration of 4 hours and 15 minutes. APAP pressure values required to normalize breathing during each stage of the simu- lation are shown in Fig 1. The simulated night consisted of programming the different stages described in Table 1, starting with 45 minutes of simulated awake stage (sleep onset) followed by a succession of different sleep stages with the features detailed in Table 1 (e.g. breathing fre- quency, number and types of respiratory events) and a final awake short period. In this way we were able to model a patient exhibiting different sleep breathing characteristics throughout consecutive sleep stages. Ten different commercially available APAP devices were tested using the new bench test model and the female-specific simulation described above: AirSense 10 (A) and AirSense 10 AutoSet for Her (B) by ResMed; Dreamstar by Sefam (C); Icon by Fisher & Paykel (D); Resmart by BMC (E); Somnobalance (F) and Prisma 20A (G) by Weinmann; System One by Respironics (H); iCH (I) and XT-Auto by Apex (J). Each APAP device was connected with its own tube to the bench model. Default APAP settings were used (minimum pressure 4 cmH 2 O, maximum pressure 20 cmH 2 O). Each device was tested twice and the results averaged to obtain the final values. Results The new OSA patient simulator could effectively combine a great variety of SDB elements to mimic the response of the predefined patient type. The responses of the assessed APAP devices to the new female-specific bench test model are summarized in Table 2. There was considerable variation among devices, particularly with respect to the mean and maximum nasal pressures applied, and the ability to overcome obstructive events and flow limitation, The residual AHI was calculated as the number of residual obstructive events per hour and the residual flow limi- tation was measured as the portion of the test in minutes (excluding the initial 45-minute wake period) that the simulated patient remained on flow limitation.
PLOS ONE | DOI:10.1371/journal.pone.0151530 March 15, 2016
3 / 11
next generation sourcing
Made with FlippingBook Online newsletter