Respiratory patternduring neurally adjusted ventilatory assistin acute respiratory failure patients.

Elettronic Supplementary Material

Nicolò Patroniti MD, Giacomo Bellani MD,Erica Saccavino MD, Alberto Zanella MD, Giacomo Grasselli MD, Stefano Isgrò MD, Manuela Milan MD, Giuseppe Foti MD, Antonio Pesenti MD.

Additional Methods and Results

NAVA principles

The study was performed with a Servo-I ventilator integrating the NAVA module[3,13]. The EAdi was registered with a nasogastric tube carrying a multiple array of esophageal electrodes (EAdi catheter, Maquet Critical Care, Solna, Sweden) [13]. During NAVA, the processed EAdi signal is used to control the Servo-I ventilator; the pressure assist starts when the EAdi reaches an inspiratory trigger threshold manually set above the baseline EAdi level. The ventilator is set to cycle off when the EAdi decreases to 70% of the peak inspiratory activity.The instantaneous airway pressure is adjusted automatically every 16 milliseconds in proportion to the processed EAdi signal, expressed in µV, multiplied by the user-set NAVA level expressed as cmH2O/µV.

PSV settings

During PSV, The the flow-trigger sensitivity was adjusted to the lowest possible level without auto-triggering, while the expiratory trigger sensitivity was set to 30% of the peak inspiratory flow.

Setting of PEEP and FiO2

PEEP and FiO2 levels set by the attending physician were left unchanged throughout the study. In our ICU PEEP and FiO2 are selected by the physicians according to specific patient’s response with a preference for low FiO2 and higher PEEP. PEEP is usually set based on the response to a PEEP trial (evaluating PaO2 and respiratory system compliance). In general if a patient does not respond to PEEP he/she may have a low PEEP level and PaO2 is corrected by increasing FiO2.

Patient-ventilator synchrony

The extent of patient-ventilator dyssynchrony was assessed by computing the asynchrony index (AI) which is the ratio between the number of asynchrony events (ineffective effort and double triggering) and the total RR (the sum of ventilator cycles and ineffective efforts). The AI was computed by visual inspection. For each patient we selected the maximal and minimal average VT,

During NAVA, the differences between Tiflow and Tineural and Teflow and Teneural did not change with increasing NAVA levels. During PSV, Tineural and Tiflow diverged at increasing PS levels(Table 2). No wasted efforts or double triggering were observed during NAVA. The AI exceeded 10% in 4, 6, and 8 patients during PSV at PS levels of 8,12, and 16 cmH2O, respectively. Median and interquartile ranges of the AI were 0% (0-5), 4.5% (0.5-12.4), 6.1%(1-22.3), and 14% (7.3-34.3)at PS levels of 4, 8, 12, and 16 cmH2O, respectively.

The increase of PS level was associated with increase of VT and worsening of patient-ventilator synchrony. On the contrary, during NAVA, we always found an excellent phase synchrony between diaphragm activity and ventilator response that was independent from the level of assistance, even in presence of extremely variable patterns or high RR. Overall, even considering the potential pitfalls at high NAVA levels, we confirm the superiority of NAVA over PSV, in terms of patient ventilator synchrony, reported by previous clinical studies [3-7].
Sedative drugs

Infusion of sedative drugs was maintained constant throughout the study.Four patients received only fentanyl 90 ± 47 μg/hour; 8 patients received propofol 1.2± 0.37 mg/kg/hour + fentanyl 100 ± 36 μg/hour or remifentanyl 20 ± 7 μg/hour; 2 patients received midazolam 0.1 ± 0.05 mg/kg/hour + fentanyl 100 μg/hour.

Additional figures

Figure E1. Effect of the level of assistance during NAVA and PSV on the peak of diaphragm electrical activity (EAdipeak) of each patient. * p< 0.05 versus lowest level of assistance of the same ventilatory mode; # p< 0.05 versus highest level of assistance of the same ventilatory mode.

Figure E2. Effect of the level of assistance during NAVA and PSV on the respiratory rate (RR) of each patient. # p< 0.05 versus highest level of assistance of the same ventilatory mode.


Figure E3. Effect of the level of assistance during NAVA and PSV on the tidal volume (VT) of each patient.*p< 0.05 versus lowest level of assistance of the same ventilatory mode; # p< 0.05 versus highest level of assistance of the same ventilatory mode.

Figure E4. Effect of the level of assistance during NAVA and PSV on the peak of airway pressure (Pawpeak) of each patient.respiratory rate (RR). * p< 0.05 versus lowest level of assistance of the same ventilatory mode; # p< 0.05 versus highest level of assistance of the same ventilatory mode.

Figure E5. Effect of the level of assistance during NAVA and PSV on the coefficient of variation of peak of diaphragm electrical activity (EAdipeak) of each patient.* p< 0.05 versus lowest level of assistance of the same ventilatory mode; # p< 0.05 versus highest level of assistance of the same ventilatory mode.

Figure E6. Effect of the level of assistance during NAVA and PSV on the coefficient of variation of respiratory rate (CV RR) of each patient.* p< 0.05 versus lowest level of assistance of the same ventilatory mode; # p< 0.05 versus highest level of assistance of the same ventilatory mode.

Figure E7. Effect of the level of assistance during NAVA and PSV on the coefficient of variation of tidal volume (CVVT)of each patient.* p< 0.05 versus lowest level of assistance of the same ventilatory mode; # p< 0.05 versus highest level of assistance of the same ventilatory mode.

Figure E8. Effect of the level of assistance during NAVA and PSV on the coefficient of variation of peak of airway pressure (CV Pawpeak)of each patient.* p< 0.05 versus lowest level of assistance of the same ventilatory mode; # p< 0.05 versus highest level of assistance of the same ventilatory mode.

Figure E9. Proportion of tidal volume (VT) greater than 8 ml/kg (top panel) and greater than 10 ml/kg (bottom panel) at different levels of assistance during NAVA and PSV.

Figure E10. Proportion of peak airway pressure (Pawpeak greater than 35 cnH2O at different levels of assistance during NAVA and PSV).