PT - JOURNAL ARTICLE AU - Moerer, Onnen AU - Harnisch, Lars-Olav AU - Herrmann, Peter AU - Zippel, Carsten AU - Quintel, Michael TI - Patient-Ventilator Interaction During Noninvasive Ventilation in Simulated COPD AID - 10.4187/respcare.04141 DP - 2015 Nov 10 TA - Respiratory Care PG - respcare.04141 4099 - http://rc.rcjournal.com/content/early/2015/11/10/respcare.04141.short 4100 - http://rc.rcjournal.com/content/early/2015/11/10/respcare.04141.full AB - BACKGROUND: During noninvasive ventilation (NIV) of COPD patients, delayed off-cycling of pressure support can cause patient ventilator mismatch and NIV failure. This systematic experimental study analyzes the effects of varying cycling criteria on patient-ventilator interaction.METHODS: A lung simulator with COPD settings was connected to an ICU ventilator via helmet or face mask. Cycling was varied between 10 and 70% of peak inspiratory flow at different breathing frequencies (15 and 30 breaths/min) and pressure support levels (5 and 15 cm H2O) using the ventilator's invasive and NIV mode with and without an applied leakage.RESULTS: Low cycling criteria led to severe expiratory cycle latency. Augmenting off-cycling reduced expiratory cycle latency (P < .001), decreased intrinsic PEEP, and avoided non-supported breaths. Setting cycling to 50% of peak inspiratory flow achieved best synchronization. Overall, using the helmet interface increased expiratory cycle latency in almost all settings (P < .001). Augmenting cycling from 10 to 40% progressively decreased expiratory pressure load (P < .001). NIV mode decreased expiratory cycle latency compared with the invasive mode (P < .001).CONCLUSION: Augmenting the cycling criterion above the default setting (20–30% peak inspiratory flow) improved patient ventilator synchrony in a simulated COPD model. This suggests that an individual approach to cycling should be considered, since interface, level of pressure support, breathing frequency, and leakage influence patient-ventilator interaction and thus need to be considered.