Study setting {9}
The study will run in ± 40 academic and non-academic centers in- and outside the European Union. A current list of participating centers can be found on https://clinicaltrials.gov/ (ID: NCT05492344).
Eligibility criteria {10}
Patients will be included when they meet all of the following inclusion criteria: admitted to a participating ICU, invasively ventilated and fulfil the Berlin criteria for moderate or severe ARDS(6), and none of the following exclusion criteria are present: age under 18, participation in other interventional studies with conflicting endpoints, conditions in which LUS is not feasible or possible (e.g. subcutaneous emphysema, wounds), mechanical ventilation for longer than 7 consecutive days in the past 30 days, diagnosis of ARDS for longer than 12 hours, history of ARDS in the previous month, body-mass index higher than 40 kg/m², intracranial hypertension, broncho-pleural fistula, chronic respiratory diseases requiring long-term oxygen therapy or respiratory support, pulmonary fibrosis with a vital capacity < 50% (severe or very severe), patients who are moribund or facing end of life, receiving or planned to receive ECMO, patients who receive invasive ventilation in home setting due to a neurological disease, previously randomized in this study, and if no informed or deferred consent could be obtained.
Hospitals with an intensive care unit are able to participate in the PEGASUS study when ethical approval is obtained and all the facilities needed for the PEGASUS study are present (Ultrasound, arterial blood gas analysis) and staff are trained for the interventions (e.g. lung recruitment, defining morphology subphenotype, prone positioning).
Who will take informed consent? {26a}
The local principal investigator, who is trained by a member of the steering committee, or her/his delegate, will obtain consent. Patients admitted for ventilator support to the ICU are, without exception, not able to give informed consent. In the participating countries, local regulations will determine when and how informed consent can be obtained from a proxy (legal representative) and when deferred informed consent can be used. The most liberal approach mandated by the study protocol is deferred proxy consent. In this situation, patients are randomized and informed consent from a legal representative will be obtained as soon as possible, but always within 72 hours after randomization. If informed consent is not obtained within 72 hours or if a legal representative declines participation, the patient will be excluded and data will not be used. When a patient dies before informed consent can be obtained from the legal representative, the data will be used and representative will be informed about the study. This approach has been approved by the ethical board in the Netherlands, Italy, Denmark, Belgium, Bangladesh, Greece, and Ireland. If the ethical board of a participating site does not approve deferred consent due to local regulations, informed consent before randomization by proxy will be accepted as an alternative. When possible, informed consent will also be obtained from the patient when the patient is recovered.
Additional consent provisions for collection and use of participant data and biological specimens {26b}
n/a. The PEGASUS study does not collect extra personal data or biological specimens that fall outside of the scope of a standard informed consent procedure and therefore additional consent provisions are not required.
Interventions
Explanation for the choice of comparators {6b}
Patients who meet all the inclusion criteria and none of the exclusion criteria will receive a 12-region LUS exam (supplement 1) to determine lung morphology using the algorithm presented in Figure 1. Patients will be randomly assigned within 12 hours after ARDS diagnosis to the intervention group, with personalized mechanical ventilation, or to the standard care, where patients will receive standard care. The standard care group will be ventilated according to the current guidelines advised by the European Society of Intensive Care Medicine (ESICM) (Table 1). In these patients, the PEEP level will be selected according to the low PEEP/ high FiO2 table from the ALVEOLI study maintaining an end-inspiratory plateau pressure (Pplat) below 30 cmH2O (Table 2).(18) Prone position is encouraged if PaO2/FiO2 ratio is ≤ 150 and recruitment maneuvers will be used as rescue therapy.
Figure 1. This logarithm is used to determine the lung morphology by ultrasound of eligible patients. The scores are LUS aeration scores explained in supplement 1 based on a 12-region LUS exam. LUS = Lung Ultrasound.
Intervention description {11a}
Patients assigned to the intervention group will have their ventilator settings adjusted based on the lung morphology (Table 1). A detailed handbook for mechanical ventilation in study participants can be found in supplement 2. In “focal” patients randomized to the personalized ventilation group, a LUS exam will be repeated every 48-72 hours to assess whether they have developed “non-focal” ARDS during their ICU stay. When a patient classified as “focal” develops “non-focal” ARDS, the settings are adjusted to the “non-focal” group (figure 2, table 1).
Table 1. Ventilation strategy for randomization and lung morphology groups.
|
|
Standard of care
|
Personalized group
|
|
“Focal”
|
“Non-focal”
|
Mode of ventilation
|
Pressure controlled, volume controlled or pressure support
|
Pressure controlled, volume controlled or pressure support
|
Pressure controlled, volume controlled or pressure support
|
Tidal volume
|
6 mL/kg PBW
|
6 to 8 mL/kg PBW
|
4 to 6 mL/kg PBW
|
PEEP
|
Table 2
|
≤ 9 cm H2O
|
≥ 15 cm H2O
|
Recruitment maneuver
|
Only for rescue
|
Only for rescue
|
Daily*
|
Prone positioning
|
PaO2/FiO2 < 150
|
PaO2/FiO2 < 200
|
PaO2/FiO2 < 150
|
Ventilation strategies according to our study protocol per randomization arm and lung morphology. Formula for calculating the tidal volume size with PBW are 50 + 0.91 x (centimeters of height – 152.4) for males and 45.5 + 0.91 x (centimeters of height – 152.4) for females.(19) * A more detailed description of the ventilation strategies can be found in supplement 2, including the recruitment maneuver. PBW = predicted body weight, PEEP = positive end-expiratory pressure, PaO2 = Partial Pressure of Oxygen in arterial blood, FiO2 = fraction of inspired oxygen.
|
Table 2. FiO2 and PEEP strategy for the standard care group.
|
FiO2
|
0.3
|
0.4
|
0.4
|
0.5
|
0.5
|
0.6
|
0.7
|
0.7
|
0.7
|
0.8
|
0.9
|
0.9
|
0.9
|
1.0
|
PEEP
|
5
|
5
|
8
|
8
|
10
|
10
|
10
|
12
|
14
|
14
|
14
|
16
|
18
|
18-24
|
During start or deterioration of patients included in the standard of care group, this table is used to determine the PEEP and FiO2 settings according to the low PEEP/ high FiO2 table of the ALVEOLI study. FiO2 = Fraction of inspired oxygen, PEEP = positive end-expiratory pressure in cm H2O.(18)
|
Figure 2. Flow diagram for enrollment of a PEGASUS participant. Patients with “focal” ARDS randomized in the intervention group will receive follow-up LUS exams to evaluate if their condition deteriorates to a “non-focal” ARDS. * ventilation strategy in Table 1. ARDS = Acute Respiratory Distress Syndrome, LUS = Lung Ultrasound.
Criteria for discontinuing or modifying allocated interventions {11b}
Criteria for discontinuation or modification of the intervention targets are described in the supplement 2. The oxygenation, pressure, and ventilation targets can be loosened during interventions (e.g. bronchoscopy, placing the patient in the prone position) or modified when the patient is not ventilated in a lung-protective way (e.g. Pplat >30 cm H2O). Specifically, in patients with “focal” ARDS and personalized ventilation, the PEEP can be increased when the FiO2 is higher than 80% and the PaO2/FiO2 is below 100 mmHg in the prone position for more than 6 hours, the physician is allowed to set the PEEP above 9 cmH2O and recruitment maneuvers can be applied. In patients with “non-focal” ARDS, prone positioning can be applied when the PaO2/FiO2 is lower than 150 mmHg for 6 hours in supine position and the FiO2 is higher than 80%. When a “non-focal” ARDS patient is breathing spontaneously, PEEP can be decreased to 10 cmH2O, and recruitment maneuvers are now only performed at the physician's discretion. For patient-triggered breaths, the tidal volume margins are no longer required.
Strategies to improve adherence to interventions {11c}
At every study site, there will be a dedicated study team that is trained by a member of the steering committee. Training consists of a LUS e-learning with an exam, reading standard operating procedures (SOP's) and a study handbook, and attending an initiation visit. The Steering Committee provided a ventilation handbook for detailed guidance on how to ventilate patients according to the study protocol and other best practices in lung-protective mechanical ventilation to minimize heterogeneity in patient management within the study. The dedicated study team will give clinical lessons to all nurses and physicians before starting the trial. Furthermore, they will monitor the ventilation settings of included patients on a daily basis. A ventilator card with study guidelines will be placed at the bedside of the patient to enhance protocol adherence. If the ventilation parameters deviate from the protocol, a member of the study site is required to write a protocol deviation form. The protocol adherence will be evaluated in the pilot study and in every meeting with the Data and Safety Monitoring Board (DSMB).
Relevant concomitant care permitted or prohibited during the trial {11d}
In controlled modes of ventilation, the default inspiration–to–expiration ratio will be 1:2. Expiratory time will be prolonged in case expiratory flow limitation is detected. The respiratory rate will be adjusted to obtain an arterial blood pH >7.25 but preferably under 35 breaths per minute. The oxygenation target ranges for SpO2 and PaO2 are 88% to 95%, and 7.3 kPa to 10.7 kPa, respectively.(19)
The attending physician decides when to extubate a patient, based on general extubation criteria or with folllowing a successful spontaneous breathing trial (SBT) with a T–piece or ventilation with minimal support (pressure support level < 10 cm H2O).
Early tracheostomy provides no advantage over late tracheotomy.(20) Therefore, tracheostomy is only to be performed on strict indications and preferably not earlier than 10 days after intubation. If a patient is treated with ECMO, the ventilator is set according to the local protocol for ventilation under ECMO. This means that the patient does not have to be ventilated according to the ventilation strategy of the randomization arm anymore. Sedation will follow the local guidelines for sedation in each participating unit. In general, these guidelines favor the use of analgo–sedation over hypno–sedation and use of bolus over continuous infusion of sedating agents. A Richmond Agitation Sedation Scale (RASS) score of –2 to 0 is seen as adequate sedation.(21)(22)
The routine use of neuromuscular blockage is not recommended. If neuromuscular blockade is required, single injections are preferred over continuous infusions.
If patients are expected to need ventilation for longer than 48 hours and/or are expected to stay in the ICU for longer than 72 hours, preventive measurements against VAP must be instituted according to the local guidelines.
A fluid balance targeted at normovolemia and diuresis of ≥ 0.5 ml/kg/hour should be maintained with diuretics or by crystalloid infusions.
Thrombosis prophylaxis will be given according to local guidelines.
Provisions for post-trial care {30}
There is no need for provisions for post-trial care after the patient is discharged from the ICU. The organization of study insurance depends on local regulations.
Outcomes {12}
The primary endpoint is all-cause mortality at day 90 (diagnosis of ARDS considered as day 0). Secondary outcomes are mortality at day 28, VFD at day 28, ICU length of stay, ICU mortality, hospital length of stay, hospital mortality, and number of complications (VAP using the clinical pulmonary infection score (CPIS), pneumothorax and need for rescue therapy). After a pilot phase, feasibility of LUS, correct interpretation of LUS images and correct application of the intervention within the safe limits of mechanical ventilation is evaluated to inform a stop-go decision.
Participant timeline {13}
The time points for enrolment, interventions and follow-up can be found in figure 3.
Figure 3. Timeline for a study participant from enrollment until last day of follow-up. a If a patient has “focal” ARDS and is randomized to personalized ventilation, a LUS will be repeated every 72 hours. b Informed consent is obtained before randomization or deferred consent is obtained within 72 hours after randomization depending on the local regulations. c Until day 7 or until extubation. d Until day 7 or ICU discharge. e Until day 90 or ICU discharge. f Until day 90 or hospital discharge. SOFA; Sequential Organ Failure Assessment, ICU; intensive care unit.
Sample size {14}
A sample of 538 patients (269 per group) is needed to detect an absolute between-group difference in 90-day mortality of 10% in favor of the intervention group, assuming a 27% mortality in the standard care, with a power of 80% at a two-tailed significance level of 0.047.(8) In the sample size calculation, an interim analysis of the primary endpoint after the recruitment of 269 patients (using a p-value of 0.003) with alpha spending has been considered.
The first 80 included patients will be used for the pilot phase of the trial. At least 20 patients in each personalized group are necessary to assess clinical feasibility and protocol adherence. As the expected ratio between “focal” and “non-focal” and the ratio between the intervention and standard care is 1:1, we would need a sample size of 80 patients for this pilot study. We expect an interobserver agreement among experts of κ: 0.85.(16) To be able to detect a clinically relevant decrease of κ towards 0.7 between experts and bedside clinicians, a total of 77 patients is needed for a power of 80% at a one-sided α level of 0.05. The primary endpoint will not be evaluated in the analysis of the pilot phase.
Recruitment {15}
Given the above sample size of 538 patients and an expected inclusion rate of 1 patient per 2 months, the recruitment period is approximately 2 years after all 40 sites started enrolling patients.(23)
Assignment of interventions: allocation
Sequence generation {16a}
Clinical research platform Castor Electronic Data Capture (EDC) (https://www.castoredc.com/) will be used to perform the randomization and is Good Clinical Practice (GCP) and Food and Drug Administration (FDA) compliant. Patients will be assigned to the personalized ventilation arm or to the control arm with a 1:1 ratio and only stratified by center.
Concealment mechanism {16b}
Patients will be randomized in Castor EDC immediately after the LUS exam. Concealment will be ensured by the use of blocks with randomly permuted sizes.
Implementation {16c}
The local principal investigator, who is trained by the steering committee, and the dedicated study time on the study site will enroll and randomize patients and make sure they will receive the ventilation strategy according to the randomization.
Assignment of interventions: Blinding
Who will be blinded {17a}
The PEGASUS study is a non-blinded RCT as it is not feasible to blind the ventilation parameters to the treating team during ICU stay. The type of lung morphology of the patients in the standard of care group will be blinded to the treating team to prevent any bias in the standard care.
Procedure for unblinding if needed {17b}
n/a. The PEGASUS study is a non-blinded RCT as it is not feasible to blind the ventilation parameters to the treating team during an ICU stay.
Data collection and management
Plans for assessment and collection of outcomes {18a}
All study data can be extracted from the paper or electronic patient record. Every center will have an initiation visit with a member of the steering committee before the inclusion of the first patient. In this visit, the data entry will be discussed to ensure the quality of the data entry. Furthermore, a SOP is provided containing detailed instructions. An extraction from the electronic case report form (eCRF) can be found in the supplement 3.
Plans to promote participant retention and complete follow-up {18b}
All the data for the PEGASUS study can be extracted from the patient’s record. Therefore, we expect a very low percentage of patients with a lost to follow-up for the primary endpoint. To further minimize the lost to follow up we inform relatives during the deferred consent procedure about the follow up procedure, researchers use a subject identification log with automatic follow-up reminders and the contact information of the research team is available on every consent form.
Data management {19}
Data quality of the PEGASUS study will be guaranteed through calculations and range checks in our database. Furthermore, data cleaning will be performed after closing the database, which will be described in our statistical analysis plan. A data management plan can be requested from the steering committee. All data will be stored for the length of the study and for 15 years afterwards.
Confidentiality {27}
Data of patients will be stored under a patient identification code that is not based on personal data. The codebook will be stored digitally and will be encrypted with a double password. The paper version will be stored behind a lock. All handling of personal data will comply with the General Data Protection Regulation (GDPR). Data can be shared between participating hospitals. Only coded information will be shared using data sharing systems developed for sharing medical data.
Plans for collection, laboratory evaluation and storage of biological specimens for genetic or molecular analysis in this trial/future use {33}
n/a. The PEGASUS study does not collect or analyze any biological specimens for genetic or molecular analysis in the current trial and for future use in ancillary studies.
Statistical methods
Statistical methods for primary and secondary outcomes {20a}
The statistical analysis will be based on the intention–to–treat principle, with patients analyzed according to their assigned treatment arms, except for cases lost to follow–up, or patients who are withdrawn due to lack of deferred informed consent. In addition, we will conduct per–protocol analyses, which only consider patients who completed the treatment according to the originally allocated protocol. Before the end of recruitment, a detailed statistical plan will be published.
When appropriate, statistical uncertainty will be expressed by the 95% confidence levels. P–value under 0.047 will be considered statistically significant for the primary endpoint and a p-value under 0.05 for secondary endpoints. Normality of data distribution will be assessed by visual inspection of histograms. For the experimental and control arms, continuous normally distributed variables will be expressed by their mean and standard deviation (SD) or, when not normally distributed, as medians and their interquartile ranges (IQR). Categorical variables will be expressed as frequencies and percentages. If less than 5% of data are missing or unavailable, no imputation data will be applied. All statistical analyses will be described in full detail in a statistical analysis plan, which will be published before the database is locked and analysis starts. Analysis will be performed with R software.
The goal of the primary analysis is to quantify the effect of LUS guided personalized mechanical ventilation in comparison with standard care on the 90-day mortality (with day of ARDS diagnosis as 0). The odds ratio for 90-day mortality is calculated using logistic regression analysis with mortality as dependent variable and randomization group as independent variable. Adjusted analysis will be performed according to European Medicines Agency (EMA) and FDA guidelines with the strongly prognostic variables age, clinical frailty and PaO2/FiO2 at admission as covariables.(24) The stratification variable (center) will be included as a random effect.
Since ventilator–free days is a highly skewed variable with a peak in -1 due to 28–day mortality, the mean ratio will be estimated using a generalized additive model for location scale and shape (GAMLSS) considering a zero–inflated and transformed beta distribution and using the delta method to estimate the confidence interval. A competing risk proportional hazard models will be used to evaluate the difference in time to extubation (accounting for mortality as a competing risk). Differences between groups in continuous variables will be analyzed with Student’s t–test or the Mann–Whitney U. Categorical variables will be compared with the Chi–squared test or Fisher’s exact test, as appropriate. Mortality rates and length of ICU and hospital stay will be compared using Kaplan–Meier mortality curves.
To assess interobserver agreement, an expert panel will score all LUS exams to assess and evaluate the clinicians’ diagnostic accuracy of distinguishing “focal” ARDS from “non-focal” ARDS by using Fleiss’ κ. The expert panel will be blinded to the randomization group and clinical parameters of the patient while scoring the LUS exams. Ventilator settings, complications, and the use of adjunctive strategies such as prone position and recruitment maneuvers will be summarized per morphology and randomized group.
Interim analyses {21b}
The study can be ended prematurely by the steering committee based on recommendations of the DSMB, for example as a result of low recruitment. There is a formal stopping rule after the first interim analysis. If the threshold of P=0.003 is passed in favor of either of the treatment arms, the study is automatically ended. In case the study is ended prematurely, the sponsor will notify the accredited institutional review board (IRB) within 15 days, including the reasons for the premature termination. Within one year after the end of the study, the investigator/ sponsor will submit a final study report with the results of the study, including any publications/abstracts of the study, to the accredited IRB.
Methods for additional analyses (e.g. subgroup analyses) {20b}
A predefined subgroup analysis stratified per morphology subphenotype (“focal” and “non-focal”) will be performed between the randomization arms for all primary and secondary outcomes.
Methods in analysis to handle protocol non-adherence and any statistical methods to handle missing data {20c}
Protocol non-adherence will be monitored by the DSMB and published in a pilot study after recruiting the first 80 patients. When the protocol adherence is not sufficient, patient recruitment will be stopped. The percentage of missing data is expected to be low as almost all data needed for the study can be extracted from the patient record. If data is missing for the primary endpoint we will perform a complete cases analysis. For other all data points, the primary method to handle missing data will be multiple imputation.
Plans to give access to the full protocol, participant level-data and statistical code {31c}
On reasonable request to a member of the steering committee the dataset, protocol and statistical code will be shared.
Oversight and monitoring
Composition of the coordinating centre and trial steering committee {5d}
The steering committee of the PEGASUS study consists of six researchers, all of whom have their different field of specialty and carry their responsibility for this specific topic in the study (e.g. imaging, clinical trials, invasive ventilation, methodology and statistics). The steering committee will meet regularly with the national coordinators to update them about the study. There will be one national coordinator for every country and he/she will be responsible for training the local investigators. The local investigator oversees the daily work of conducting the study in their hospital.
Composition of the data monitoring committee, its role and reporting structure {21a}
A DSMB is installed to monitor safety and the overall conduct of the trial. The DSMB consists of four individuals with one as the chair. The DSMB will first meet after inclusion of the first 80 patients in the pilot phase, approximately 6 months after the first patient is enrolled. Subsequent to this meeting, the DSMB will meet virtually every 6 months. The DSMB will review the overall status of the study, the number of patients enrolled and adherence to the protocol (in total and per center). The DSMB will monitor safety of both ventilation strategies by monitoring the secondary endpoints of ventilation specific complications. The report and/or advice of the DSMB will be sent to the sponsor of the study and to the IRB of the Amsterdam UMC.
Adverse event reporting and harms {22}
The risks of the PEGASUS study are considered minimal as the ventilation methods in the intervention group are already being applied in the standard care of ARDS patients. For this reason, we are not expecting Serious Adverse Events (SAEs) related to the study. Therefore, we report secondary endpoints of this trial, which incorporate ventilation-specific complications, in a line listing two times per year to the IRB to monitor safety of both treatment strategies. The IRB will receive a line listing of the secondary endpoints incorporating ventilation-specific complications (incidence of pneumothorax and VAP) and ICU mortality. These endpoints will be specified per study arm in the line listing without disclosing the specific arms.
Frequency and plans for auditing trial conduct {23}
On-site monitoring will be performed by an independent monitor or a researcher at all participating sites. The roll of the monitor is to verify the completeness and correctness of the research dossier and the presence of complete informed consent forms. The monitors will follow an approved monitoring plan. A monitor visit must be performed at least when the first 5 patients are included and at the end of the trial.
Plans for communicating important protocol amendments to relevant parties (e.g. trial participants, ethical committees) {25}
Every substantial amendment must be reported and approved by the ethical board of the sponsor. Every participating centre and other relevant parties will be notified of the approval of an amendment by e-mail and trial registries will be updated.
Dissemination plans {31a}
The results of the study will find their way into (inter–) national scientific journals and guidelines. Every participating center is encouraged to submit substudies and to publish these after approval of the sponsor and publication of the main article. Participants can find the results of the study on https://clinicaltrials.gov/ under the name PEGASUS or with the identifier (NCT05492344).