Nasal CPAP is the current therapy of choice in sleep related breathing disorders. The CPAP pressure is usually a fixed pressure for inspiration and expiration. However, nCPAP therapy has a limited compliance even in patients who require this therapy due to a high quantity of apneas and hypopneas. The recently introduced PAP system with C-flex (REMstar Pro) intends to improve this situation by lowering the expiratory pressure proportional to the patient's airflow. Three different gain values for the proportional reduction of pressure during expiration correspond to three different C-flex modes. Since the pressure is lowered adaptively we wanted to investigate the occurrence of flow limitations during inspiration in different sleep stages using the three possible C-flex modes.
Subjects with an AHI > 20/hour were included in this study. Patients underwent cardiorespiratory polysomnography with added recording of respiratory flow (pneumotachograph and pressure transducer by Hans Rudolph) and respiratory effort (esophageal pressure transducer by Gaeltec). Effective pressure was titrated using the CPAP mode. Subsequently the CPAP mode and the three C-flex modes were used in a randomized order.
The four conditions were applied during NREM sleep for continuous episodes of 15 minutes each. The four conditions were also applied during REM sleep. Since REM sleep episodes tend to be shorter, five continuous minutes were taken as one episode. Then three episodes of these five minutes of REM sleep were recorded for each of the four ventilation conditions in order to balance the number of breaths in REM and NREM sleep. A total of 120 minutes were used for subsequent evaluation. Breath by breath evaluation was performed with the Windaq software (Dataq Instruments, OH, USA ). Inspiratory flow limitation was counted for all breaths if the amplitude of esophageal pressure exceeded a fixed value, calculated for each subject during quiet wakefulness as the mean esophageal pressure amplitude plus two-fold standard deviation.
Four subjects were recorded with full 120 minutes of valid and artifact free signals for esophageal pressure, airflow and polysomnography and with all four conditions for NREM and REM sleep. A total of 7116 breaths were evaluated. 223 breaths (3.1%) had inspiratory flow limitation according to our criteria. 117 of these breaths (52%) occurred during CPAP mode and 106 (48%) occurred during the three C-flex modes. During NREM sleep we found again 117 breaths (52%) and during REM sleep 106 breaths (48%) were detected. One of the four subjects had most flow limited breaths (162). In this subject 93 breaths occurred during CPAP mode and 69 during C-flex mode.
The results confirm that most flow limited breaths are found in subjects which do snore even when having effective CPAP therapy. During the sum of all three C-flex modes we do see less flow limited breaths than during the CPAP mode. Therefore we summarize that there are no signs of an increased number of flow limitations during Cflex compared to CPAP mode. However, the number of subjects under this protocol has to be increased in order to obtain considerable statistical strength for our results.
This research was supported by an unrestricted grant from Respironics.