Lung-protective ventilation has been the main supportive intervention in managing acute respiratory distress syndrome caused by SARS-CoV2 infection (C-ARDS), similar to ARDS from other causes. Due to the persistence of severe hypoxemia during C-ARDS, adjunctive measures such as prone positioning have been frequently used in addition to the limitation of tidal volumes, driving and plateau pressures, and the individual selection of positive end-expiratory pressure (PEEP) [1]. However, prone positioning needs trained personnel and is not without risks; complications such as accidental extubation, endotracheal tube obstruction or displacement, brachial plexus palsy, and facial and thoracic pressure ulcer have been described [2, 3].
Recruitment maneuvers are techniques designed to improve oxygenation by reopening and keeping open nonaerated parts of the lungs. The classical recruitment maneuvers are based on the application of high pressures, usually through a sustained inflation or stepwise increase of inspiratory pressure and/or of PEEP over a sufficient period of time. Clearly, they expose the patient to hemodynamic consequences [4, 5]. A recent randomized clinical trial using the stepwise approach described major complications of high pressures, and the arm using an open lung approach had higher mortality [6].
Lateral positioning does not require the application of higher pressures. However, it has mainly been attempted in unilateral pneumonia to improve oxygenation by improving the ventilation of the sick lung placed up [7, 8]. We reasoned that, in bilateral lung injury like C-ARDS, lateral positioning performed in sequential steps might act as a recruitment maneuver for each lung sequentially provided that sufficient PEEP is provided to prevent derecruitment. The different effects of the gravitational axis on each lung during lateral positioning can modify regional transpulmonary pressure (PL) that may help re-expand collapsed regions [9, 10]. This postural recruitment maneuver (P-RM) of the dependent parts of the lungs can be administered without changes in applied airway pressures or the need to turn the patient prone completely [11, 12].
We hypothesized that P-RM could be a useful adjuvant intervention improving lung aeration, helping to homogenize ventilation distribution without using high airway pressures or prone positioning. The objective of this study was to evaluate the feasibility and short-term physiological effects of the P-RM on pulmonary mechanics, gas exchange, lung aeration, and regional distribution of tidal ventilation and perfusion in patients with COVID-19-associated ARDS.
Methods
A more detailed description of the methods is provided in the Additional file 1.
The study was approved by the Ethical Committee (Rebagliati Hospital, Lima, Perú, N° 1307) and registered at Clinicaltrials.gov NCT04475068. Informed consent was obtained from the legally authorized substitute decision-maker.
Patients
This single-center prospective observational study enrolled consecutively patients from July 2020 through Oct 2020. Inclusion criteria were: (1) patients with positive SARS-CoV-2 infection (confirmed by using real-time quantitative PCR on nasopharyngeal swabs); (2) moderate-to-severe ARDS as per the Berlin definition (PaO2/FiO2 ≦ 200 mmHg) under mechanical ventilation [13]; (3) Age ≧ 18 years old; (4) body mass index ≤ 35 kg/m2. Exclusion criteria were: (1) contraindications for EIT monitoring as (a) unstable spine or pelvic fractures; (b) pacemaker, automatic implantable cardio-defibrillator; (c) skin lesions between the 4th and 5th ribs where the EIT belt is positioned; (2) pregnancy; (3) mechanical ventilation > 1 week; (4) multi-organ failure; (4) hemodynamic instability defined as persistent mean arterial pressure lower than 60 mm Hg despite adequate fluid resuscitation and two vasopressors or increase of vasopressor dose by 30% in the previous 6 h; (5) COPD; (6) pneumothorax; and (7) increased intracranial pressure.
Mechanical ventilation settings
Patients were mechanically ventilated (Servo-I, Maquet), deeply sedated, and paralyzed. Patients were receiving volume-controlled ventilation, FIO2 adjusted to SpO2 92–97%, tidal volume ≤ 6 mL/kg predicted body weight, adjusted to a plateau pressure of ≤ 28 cmH2O and a driving pressure ≤ 15 cmH2O, respiratory rate 20–30 breaths/min (adjusted to pH 7.20–7.40), an inspiratory–expiratory ratio of 1:1.5 to 1:2 (with an inspiratory pause of 10%) [14, 15]. The PEEP level was chosen using the one-breath decremental PEEP maneuver to calculate the recruitment-to-inflation ratio (R/I ratio) [16]: PEEP = 15 cmH2O for R/I ratio > 0.5, and PEEP = 12 cmH2O for R/I ratio ≤ 0.5. We reasoned that maintaining a sufficient PEEP level during the P-RM was important to avoid collapse during lateral position. Ventilatory settings were kept constant throughout the study.
Measurements
Regional ventilation (ΔZ = change in impedance) and aeration (EELI = end-expiratory lung impedance) were obtained with an EIT monitor (Enlight 1800, Timpel, Brazil). The distribution of tidal ventilation was determined as a percentage of regional ΔZ/total ΔZ and used to estimate regional tidal volume (VTr = regional ΔZ/total ΔZ × total VT). Regional lung compliance was calculated as regional VTr/ΔP. The change in lung aeration was estimated by the change in EELI [ΔEELI × (VT/ΔZ)]. Lung perfusion was obtained by injecting a 10-mL bolus of 7.5% hypertonic saline solution into a central venous catheter during an expiratory pause [17].
Ventilation and perfusion maps were segmented into regions of interest (ROI) [18] (Fig. 1). To compare the changes between position steps, the lungs were segmented into two equally sized ROIs: ventral (upper lung or non-dependent half) and dorsal (lower lung or dependent half); we further divided the lungs into four ROIs for the lateral position according to the new situation when lateralized (ventral dependent or non-dependent, dorsal dependent or non-dependent).
Pleural pressure was estimated by esophageal manometry (Cooper Surgical). Pulmonary mechanics was measured during inspiratory and expiratory holds of 0.5 and 4 s, respectively [15].
Lung aeration was assessed by lung ultrasound (MyLab Gold 25, Esaote) using the lung ultrasound aeration score (LUS) calculated by summing regional scores (0–3 points) obtained in 6 regions of each lung [19] and the consolidation score to assess the degree of juxta-pleural consolidation; each explored area was divided into four grades and scored between 0 and 3 [20].
Protocol
Patients were studied in five body positions in sequential order, each maintained during 30 min: Supine-1 (S1), which served as the baseline condition; Lateral-1 (L1-the less ventilated lung evaluated by EIT was positioned up first); Supine-2 (S2-after first lateral position); Lateral-2 (L2-the contralateral lung was positioned up); Supine-3 (S3-after second lateral position) (Fig. 1). Lateral positioning was done with an inclination of 30° using a custom-made support cushion lined with a special foam (see Additional file 1: Fig. S1).
At the end of each 30-min period, arterial blood gas samples, hemodynamics, pulmonary mechanics, and both EIT and lung ultrasound images were recorded. In one patient, it was only possible to obtain ultrasound and perfusion images at S1 and S3, and it was not possible to also insert an esophageal balloon.
Statistical analysis
Due to the lack of previous comparable studies on the subject that allow the calculation of a sample size, we initially chose a convenience sample of 12 patients and increased it to 15 patients after obtaining additional EIT equipment from Timpel. Descriptive statistics were expressed as mean (standard deviation), median (interquartile range), or counts and percentages, as appropriate. Normality was assessed by the Shapiro–Wilk test. Differences between measurements at different body positions were evaluated using a restricted maximum likelihood analysis for the mixed-effects model. When multiple comparisons were made, P values were adjusted through Sidak post hoc correction. The PR-M effects are expressed as mean difference and 95% CI or median difference (interquartile range). A paired t-test or Wilcoxon signed-rank test was used to analyze paired differences between two positions, as appropriate. All tests were 2-tailed, and differences were considered significant when P-value < 0.05. Analysis was performed using Prism version 8 (GraphPad Software).