Data on respiratory variables in critically ill patients with acute respiratory failure placed on proportional assist ventilation with load adjustable gain factors (PAV+)

The data show respiratory variables in 108 critically ill patients with acute respiratory failure placed on proportional assist ventilation with load adjustable gain factors (PAV+) after at least 36 h on passive mechanical ventilation. PAV+ was continued for 48 h until the patients met pre-defined criteria either for switching to controlled modes or for breathing without ventilator assistance. Data during passive mechanical ventilation and during PAV+ are reported. Data are acquired from the whole population, as well as from patients with and without acute respiratory distress syndrome. The reported variables are tidal volume, driving pressure (ΔP, the difference between static end-inspiratory plateau pressure and positive end-expiratory airway pressure), respiratory system compliance and resistance, and arterial blood gasses. The data are supplemental to our original research article, which described individual ΔP in these patients and examined how it related to ΔP when the same patients were ventilated with passive mechanical ventilation using the currently accepted lung-protective strategy “Driving pressure during assisted mechanical ventilation. Is it controlled by patient brain?” [1].


a b s t r a c t
The data show respiratory variables in 108 critically ill patients with acute respiratory failure placed on proportional assist ventilation with load adjustable gain factors (PAVþ) after at least 36 h on passive mechanical ventilation. PAVþ was continued for 48 h until the patients met pre-defined criteria either for switching to controlled modes or for breathing without ventilator assistance. Data during passive mechanical ventilation and during PAVþ are reported. Data are acquired from the whole population, as well as Abbreviations: PAV þ, Proportional assist ventilation with load adjustable gain factors; CMV, Controlled mechanical ventilation (Passive mechanical ventilation); Rmin, End-inspiratory airway resistance during controlled mechanical ventilation; PEEPi, Intrinsic positive end-expiratory airway pressure; V T , Tidal volume; Crs, Respiratory system compliance; ΔP, Driving pressure; VT CMV , Tidal volume during controlled mechanical ventilation; Crs CMV , Respiratory system compliance during controlled mechanical ventilation; ΔP CMV , Driving pressure during controlled mechanical ventilation; VT PAV þ aver , Average tidal volume during the first 8-h period of proportional assist ventilation with load adjustable gain factors; Crs PAV þ aver , Average respiratory system compliance during the first 8-hour period of proportional assist ventilation with load adjustable gain factors; ΔP PAV þ aver , Average driving pressure during the first 8-h period of proportional assist ventilation with load adjustable gain factors; VT PAV þ , Tidal volume during proportional assist ventilation with load adjustable gain factors; Crs PAV þ , Respiratory system compliance during proportional assist ventilation with load adjustable gain factors; ΔP PAV þ , Driving pressure during proportional assist ventilation with load adjustable gain factors; PaCO 2 , Partial pressure of arterial CO 2 ; ARDS, Acute respiratory distress syndrome n Corresponding author.
E-mail address: georgop@med.uoc.gr (D. Georgopoulos). from patients with and without acute respiratory distress syndrome. The reported variables are tidal volume, driving pressure (ΔP, the difference between static end-inspiratory plateau pressure and positive end-expiratory airway pressure), respiratory system compliance and resistance, and arterial blood gasses. The data are supplemental to our original research article, which described individual ΔP in these patients and examined how it related to ΔP when the same patients were ventilated with passive mechanical ventilation using the currently accepted lung-protective strategy "Driving pressure during assisted mechanical ventilation. Is it controlled by patient brain?" [1]. May stimulate further research in critically ill patients on the ability of the feedback systems of regulation of breathing to protect the lungs from ventilator induced lung injury.
May facilitate new approaches for titrating ventilator settings in critically ill patients.

Data
The data show V T , Crs and ΔP in critically ill patients during PAV þ and CMV, and the changes in these variables when patients were switched from CMV to PAV þ. Data in patients with and without ARDS are presented, as well as Rmin and PEEPi during CMV in these patients. The relationship between PaCO 2 and ΔP during PAV þ is also shown (Figs. 1-11).  Individual relationships between the change in V T and that of Crs when the patients were switched from CMV to PAV þ in ARDS (n¼64) and non-ARDS (n ¼44) patients. Continuous lines; Regression lines. During CMV, the measurements of V T and Crs were obtained within 8 h before switching to PAV þ when criteria for passive mechanical ventilation were met (VT CMV , Crs CMV ). V T and Crs during PAV þ were obtained by averaging these variables during the first 8-h PAV þ period (VT aver , Crs PAV þ aver ). Therefore, each patient was characterized by a single data point. The percentage of patients in whom VT PAV þ aver increased while Crs PAV þ aver decreased did not differ between ARDS and non-ARDS patients (15.6% vs. 6.8%, p ¼0.23). VT PAV þ aver À VT CMV ; difference in tidal volume between PAV þ (average data) and CMV. Crs PAV þ aver À Crs CMV ; difference in respiratory system compliance between PAV þ (average data) and CMV.

Patients
Patients under mechanical ventilation for at least 36 h and ventilated with a controlled mode (CMV, volume or pressure control) were screened for eligibility. Enrollment criteria required absence of the following [2]: a do-not-resuscitate order, mechanical ventilation with assisted modes (independent of the duration), expected poor short-term prognosis ( o3 months), neuromuscular disease with respiratory muscle involvement that could permanently impair the ability to breathe spontaneously, and age o18 and 485 years. Inclusion criteria were the ability to trigger the ventilator at a satisfactory rate (4 10 breaths/min); PaO 2 460 mmHg, with fractional concentration of inspired O 2 (FIO 2 ) of o65%; total [extrinsic (PEEP) and intrinsic (PEEPi)] positive end-expiratory airway pressure Individual relationships (all patients) between the change in V T (difference in V T between PAV þ and CMV, VT PAV þ À VT CMV ) and that of Crs (difference in Crs between PAV þ and CMV, Crs PAV þ À Crs CMV ) when the patients were switched from CMV to PAV þ . All measurements (n¼ 744). In each patient during CMV, one measurement was performed, while during PAV þ , multiple measurements at different time points were obtained [8 (4-10) measurements per patient, median (interquartile range)]. Therefore, each patient was characterized by a number of data points equal to the number of measurements during PAV þ .

Measurements during CMV and PAVþ
During CMV and the 48-h PAV þ period, the following parameters were measured at specific time intervals. 1) Gas exchange data: PaO 2 , PaCO 2 , PaO 2 /F I O 2 , and pH 2) Respiratory data: V T (calculated as the ratio of minute ventilation to ventilator rate, which were measured by averaging data over 1 min), end-inspiratory alveolar pressure during CMV and PAV þ (Pplat CMV , Pplat PAV þ , respectively), respiratory system compliance (Crs CMV , Crs PAV þ ), PEEP, PEEPi and PEEP TOT (see below).

Respiratory system mechanics during controlled mechanical ventilation (CMV)
During CMV, respiratory system mechanics were assessed within 8 h before switching to PAV þ (when criteria for passive ventilation were met). If the patients were ventilated on volume control mode, respiratory system mechanics were measured at settings in which the patients had previously been ventilated. If the patients were ventilated on pressure control mode, respiratory system mechanics were measured by placing them on volume control. The ventilator rate remained constant and the ventilator was set to deliver a V T similar to that achieved with the pressure control. A square wave inspiratory flow-time profile was used. The mechanical properties of the respiratory system were determined using the occlusion technique [3,4]. Briefly, the airways were occluded at endinspiration for 3 s; there was an immediate drop in airway pressure from a peak (Ppeak) to a lower value (P 1 ), followed by a gradual decay to a plateau (Pplat). In each patient, at least 3 breaths with satisfactory plateau were analyzed and the mean values were reported. Intrinsic PEEP (PEEPi) was measured by occluding the airways at the end of a tidal expiration for 3 seconds and observing the airway pressure. Again, 3 breaths were analyzed. Respiratory system static inflation end-inspiratory compliance (Crs CMV ), end-inspiratory airway resistance (R min , the "ohmic" component of airway resistance), end-inspiratory total resistance (R max ) and the resistance due to time constant inequalities and/or viscolelastic properties (ΔR ¼R max À R min ) of respiratory system were computed according to standard formulas [3,4]. The endotracheal tube resistance was not taken into account.    A software program is built into the ventilator which, when proportional assist ventilation mode (PAV þ) is activated, estimates the compliance (Crs PAV þ ) of respiratory system, based on methods previously described [5]. Briefly, at random intervals of 4-10 breaths, a 300 ms pause maneuver at the end of inspiration is applied and the Paw at end-inspiratory pause time (Pplat PAV þ ) is measured. Given that Pplat PAV þ (1) is equal to end-inspiratory alveolar pressure (Palv) and (2) during the interval of obstruction inspiratory muscle pressure returns to zero [5], Crs PAV þ is calculated as follows: PEEPi is estimated by the ventilator software using the following technique [2]. Since Crs PAV þ has been measured, the software, assuming that expiration is passive, estimates Palv continuously from the beginning to the end of expiration. If expiratory flow continues until shortly before the next trigger, PEEPi is calculated as the difference between estimated Palv and PEEP 100 ms before the next trigger. If expiratory flow becomes zero before a breath is triggered, then Palv¼ Paw ¼PEEP and thus Palv-PEEP ¼ 0 (i.e. PEEPi ¼0).
In an automated system in which interventions are applied randomly under unsupervised conditions, safeguards need to be included to ensure that data obtained under unfavorable conditions are filtered out. Thus, all raw data are subjected to checks, and the estimates of Crs PAV þ are discarded if any of the rejection pre-defined criteria are met [5]. Valid estimates of Crs PAV þ are required for breath delivery, and are constantly updated by averaging new values with previous values. If new values for Crs PAV þ are rejected, the previous values remain active until valid new values are obtained. The ventilator software monitors the update process and generates an escalating alarm condition if the old values do not refresh.
The driving pressure during PAV þ (ΔP PAV þ ) is calculated as VT/Crs PAV þ . ΔP PAV þ is calculated with and without taking PEEPi into consideration. Crs PAV þ without taking PEEPi into consideration, is estimated as follows: It follows that ΔP PAV þ without taking PEEPi into consideration, is the difference between Pplat PAV þ À PEEP.

Statistical analysis
Data are given as median (25th-75th interquartile range), unless stated otherwise. Proportions were compared using the Fisher exact test. Continuous variables were compared with Wilcoxon and Man-Whitney tests, as appropriate. Regression analysis was performed using the least square method. Linear mixed effect models on parameters of repeated measurements were used to investigate changes in various variables over time during PAV þ. The values of the first four serial measurements, corresponding to an 8-h PAV þ period, were included in the model in order to compare with the corresponding variables obtained within the 8-h CMV period. P o0.05 was considered as significant.