Study population
The study was approved by the Institutional Research and Ethics Committee of the Peking Union Medical College Hospital. Informed consent was obtained from all patients or next of kin before data were included into the study. The clinical trial registration number was NCT04081155.
When the research team was available from Jan 2019 to May 2020, patients with ARDS or with high-risk ARDS admitted to the Department of Critical Care Medicine of Peking Union Medical College Hospital, who received mechanical ventilation, were screened for eligibility. Diagnosis of ARDS was based on the Berlin definition [10]. High-risk ARDS was defined as those mechanically ventilated patients who had some high risk factors of ARDS (major operation, massive transfusion and trauma, etc.) and lung collapse in the dependent region but with PaO2/FiO2 > 300 mmHg at the enrollment.
Included patients should have been deeply sedated and a central venous catheter placed for treatment as per clinical decision at the time of enrollment. Patients were excluded from the study in the presence of age < 18 years, pregnancy, ribcage malformation, baseline PEEP > 12 cmH2O and SpO2 < 88%, and any contraindication to the use of EIT (e.g., automatic implantable cardioverter defibrillator, and implantable pumps).
Study protocol
Patient demographics and relevant clinical data were collected at the enrollment day, including age, sex, Acute Physiology and Chronic Health Evaluation II score (APACHE II), heart rate, mean arterial pressure, FiO2, SpO2, and as outcome the 28-day mortality.
Patients were ventilated under pressure control mode. Ventilator settings were tidal volume 6–8 ml/kg of ideal body weight. PEEP, FiO2 was set to maintain SpO2 > 90%, and respiratory rate was set to obtain arterial pH of 7.30–7.45 based on the ARDS-Net suggestions [11]. All patients were deeply sedated (Richmond Agitation-Sedation Scale at − 4) and kept in the supine position. The following PEEP adjustment was performed for each patient:
-
1.
PEEP was switched to a zero end-expiratory pressure (ZEEP) for 10 min, and FiO2 was titrated to obtain peripheral oxygen saturation (SpO2) > 90%.
-
2.
PEEP was increased to a high PEEP level (preferably 15 cmH2O) for another 10 min within a single step. If the patient was not able to tolerate 15 cmH2O as assessed by the physician (e.g., due to impaired circulation), PEEP of 12 cmH2O was used instead.
Regional ventilation and perfusion measured by EIT
EIT measurements were performed with PulmoVista 500 (Dräger Medical, Lübeck, Germany) throughout the PEEP adjustment. A silicone EIT belt with 16 surface electrodes was placed around the patient’s thorax at the 4th intercostal space level. All patients received standard care. EIT measurements were continuously recorded at 20 Hz. At the end of each PEEP level (i.e., ZEEP and high PEEP), a bolus of 10 ml 10% NaCl was injected during a respiratory pause (at least for 8 s) through the central venous catheter. The EIT data were digitally filtered using a low-pass filter with a cut-off frequency of 0.67 Hz to eliminate periodic cardiac-related impedance changes (for evaluation of both ventilation and perfusion). Perfusion evaluated via saline bolus injection corresponded to non-periodic impedance drop that was not influenced by the low-pass filtering. Further, the data were analyzed offline using customized software programmed with MATLAB R2015 (the MathWorks Inc., Natick, MA).
Ventilation map was equally divided into two non-overlapping horizontal anterior-to-posterior regions of interest, which were denoted as the ventral and dorsal regions. Regional ventilation map was calculated by subtracting the end-expiration from the end-inspiration image, which represents the variation during tidal breathing. The tidal images before the apnea period (2-min period) were averaged to increase the signal-to-noise ratio.
$$ {V}_i=\frac{1}{N}{\sum}_{n=1}^N\left(\varDelta {Z}_{i, Ins,n}-\varDelta {Z}_{i, Exp,n}\right) $$
(1)
where Vi is the pixel i in the ventilation image, N is the number of breaths within the analyzed period, and ΔZi,Ins and ΔZi,Exp are the pixel values in the raw EIT image at the end-inspiration and end-expiration, respectively.
The ventilation gain and loss via PEEP increase were assessed as follows:
$$ {\Delta V}_i={V}_{i\_\mathrm{PEEP}}-{V}_{i\_\mathrm{ZEEP}} $$
(2)
ΔVi > 0 is associated with ventilation gain whereas ΔVi < 0 with ventilation loss. To improve signal-to-noise ratio, we defined a threshold of 20% of maximum Vi. Recruited pixels were defined as pixels r that exhibited ventilation gains higher than the threshold:
$$ {\Delta V}_r>20\%\times \max \left({V}_i\right),i\in \left[1,1024\right] $$
(3)
Similarly, overdistended pixels were defined as pixels o with ventilation loss higher than the threshold:
$$ {\Delta V}_o<-20\%\times \max \left({V}_i\right),\kern1.50em i\in \left[1,1024\right] $$
(4)
The number of overdistended pixels over the number of recruited pixels (O/R ratio) was subsequently calculated. With the O/R ratio, we tried to summarize the degrees of overdistension and recruitment with one single index. The patients were divided into low O/R (O/R ratio < 15%) and high O/R groups (O/R ratio > 15%).
Changes of end-expiratory lung impedance (ΔEELI) were determined relative to the reference time point during device calibration. The global inhomogeneity (GI) index [12] was calculated offline.
Due to its high conductivity, 10% NaCl acts as an EIT contrast agent, passes through the pulmonary circulation thereby producing a dilution curve after bolus injection during the apnea period based on the first pass kinetics theory [13, 14]. Regional perfusion map was calculated as the slope of regional impedance–time curves after saline bolus injection [15, 16]. The detailed calculation was described in previous studies [6, 7]. In brief, the regional impedance–time curves during the descending phase were fitted with linear regression:
$$ \varDelta {Z}_i(t)={a}_it+b $$
(5)
where t is the time starting from one cardiac cycle after the initial descent in the global impedance curve caused by saline injection, and ending at the trough of the global curve during the apnea period. Pi, the perfusion value of pixel i in the perfusion image, was equaled to –ai.
Further, ventilated and perfused regions were defined as follows: Region k is ventilated if:
$$ {V}_k>20\%\times \max \left({V}_K\right),\kern1.50em K\in \left[1,1024\right] $$
(6)
Similarly, region g was perfused if:
$$ {P}_g>20\%\times \max \left({P}_G\right),\kern1.50em G\in \left[1,1024\right] $$
(7)
Subsequently, the following three regions were identified: the area that was only ventilated (AV), the area that was only perfused (AP), and the area that was both ventilated and perfused (AV+P). To correlate with clinical events, the following EIT-derived parameters were calculated according to their physiological definitions:
$$ \mathrm{DeadSpac}{\mathrm{e}}_{\%}={A}_V/\left({A}_V+{A}_P+{A}_{V+P}\right)\times 100\% $$
(8)
$$ \mathrm{Shun}{\mathrm{t}}_{\%}={A}_P/\left({A}_V+{A}_P+{A}_{V+P}\right)\times 100\% $$
(9)
$$ \mathrm{VQMatc}{\mathrm{h}}_{\%}={A}_{V+P}/\left({A}_V+{A}_P+{A}_{V+P}\right)\times 100\% $$
(10)
Figure 1 illustrates the analysis with patient data. The tidal impedance variation during normal tidal breathing before apnea was used for the calculation of ventilation-related parameters. The impedance–time curve caused by saline bolus during the apnea period was used for the perfusion-related parameters. The regional recruited and overdistended pixel distribution image was derived from the difference of ZEEP and high PEEP ventilation tidal images (Fig. 1 top). The regional V-Q images (Fig. 1 bottom) were derived from the difference of ventilation and perfusion images at the same PEEP level.
Statistical analysis
A descriptive analysis was performed. Normal distribution was assessed with the Kolmogorov–Smirnov normality test. Normally distributed results were presented as mean ± SD whereas non-normally distributed results were presented as median (25th–75th percentile). The Mann–Whitney test was used to compare groups on continuous variables, and chi-square and Fisher’s exact tests were used to compare categorical variables. Paired t test or Wilcoxon’s signed-rank test was performed to compare values at ZEEP and high PEEP, as appropriate. Comparisons of two continuous variables were performed using Spearman’s correlation and linear regression. All comparisons were two-tailed, and P < 0.05 was required to exclude the null hypothesis. The areas under the receiver operating characteristic (AUC) curves were compared using a Hanley–McNeil test. The statistical analysis was performed by using the software package SPSS 24.0 (SPSS Inc. Chicago, IL) and MedCalc 11.4.3.0 Software (Mariakerke, Belgium).