Quantification of Systemic Congestion by Ultrasound As a Predictor of Weaning Failure From Mechanical Ventilation

• Protocol overview and Rationale Weaning patients from invasive mechanical ventilation represents a significant daily challenge within the intensive care unit (ICU). The process is complex, and failure can lead to prolonged ventilation, increased morbidity, and worsened patient outcomes. The "Quantification of Systemic Congestion by Ultrasound As a Predictor of Weaning Failure from Mechanical Ventilation (SCOUT)" study is a prospective investigation designed to evaluate a novel, non-invasive method for identifying patients at high risk of weaning failure before a trial is even attempted.

The core clinical problem addressed by this study is the high incidence of weaning failure, which occurs in 10-30% of patients. While respiratory complications are a common cause, up to 30% of failures are attributed to cardiovascular origins. A key mechanism in these cases is systemic congestion, which can lead to weaning-induced pulmonary edema (WIPO). This condition occurs when the physiological stress of a spontaneous breathing trial (SBT) causes a rapid shift of fluid into the central circulation, overwhelming cardiac capacity and flooding the lungs, even in patients without a prior history of heart failure.

Current methods for assessing a patient's fluid status and risk of congestion have significant limitations. Clinicians often rely on metrics such as cumulative fluid balances, daily weight changes, physical examination findings, and biomarkers. However, these tools have proven to be insufficiently reliable or specific for accurately predicting which patients will develop WIPO during an SBT. While tools like echocardiography are useful for diagnosing pulmonary edema after it has already occurred, they have not demonstrated utility in identifying at-risk patients beforehand. Similarly, pulmonary ultrasound can detect lung edema (B-lines) but not identify high risk patients before it occurs.

This study will investigate the Venous Excess Ultrasound (VExUS) grading system as a predictive tool. VExUS is a novel, point-of-care ultrasound protocol that provides a comprehensive assessment of systemic venous congestion. It integrates ultrasound measurements of the inferior vena cava with Doppler flow patterns in the hepatic veins, portal vein, and intrarenal veins to generate a score that quantifies the degree of venous congestion.

Given that systemic congestion is a central mechanism in WIPO, a tool that can reliably measure it at the bedside holds immense potential. VExUS provides real-time, repeatable information, allowing for dynamic patient monitoring. The central justification for this study is that by using the VExUS score to assess a patient before initiating an SBT, clinicians may be able to identify those at high risk of cardiovascular-related weaning failure. This proactive identification would allow for targeted therapeutic measures to be implemented preemptively, thereby improving a patient's readiness for weaning, reducing the likelihood of failure, and minimizing the consequences associated with prolonged mechanical ventilation. This clinical rationale underpins the study's formal hypotheses and objectives.

• Study Hypotheses and Objectives

• Primary Hypothesis: The assessment of systemic congestion using the VExUS ultrasound score immediately prior to a SBT can identify patients who are at risk of weaning failure.
• Primary Objective: To evaluate the utility of the VExUS score for predicting the failure of weaning from mechanical ventilation.

Secondary Hypotheses and Objectives

• Secondary Hypotheses: The VExUS score demonstrates superior diagnostic performance in identifying patients at risk of SBT failure compared to other methods (pulmonary ultrasound, cardiac ultrasound, and biomarkers).

Patients with a higher VExUS score before the SBT experience a longer total weaning duration.

The assessment of systemic congestion via VExUS can predict the need for reintubation.

• Secondary Objectives: To compare the diagnostic performance of different methods (biomarkers, pulmonary ultrasound, cardiac ultrasound) for predicting SBT failure.

To evaluate whether a correlation exists between the VExUS-assessed degree of congestion and the number of days required for weaning.

• Study Design and Methodology The study is designed as a prospective, multicenter registry. The design is considered quasi-experimental because it evaluates the effect of an intervention, the comprehensive ultrasound assessment, on a cohort of patients.

• Study Population and Sample Size

Inclusion criteria

Patients aged > 18 years. Admitted to an Intensive Care Unit (ICU). Requiring invasive mechanical ventilation for more than 48 hours. Meeting the center's standard criteria to begin weaning and undergo an SBT.

Exclusion criteria

Inadequate ultrasound window for assessment. Presence of "do not reintubate" orders. Patients with a tracheostomy. Presence of atrial fibrillation. Diagnosed liver cirrhosis (portal hypertension). Absence of qualified personnel to perform the ultrasound at the time of the SBT.

Denial of informed consent.

• Sample Size: The study requires a total of 350 patients for the final analysis. This calculation is based on an assumed 30% weaning failure rate, an alpha risk of 0.05, and a statistical power greater than 0.8. The enrollment target will be higher to account for an estimated 20% loss to follow-up.

The following sequence of procedures and data collection will be applied to each enrolled patient.

At Study Inclusion The following baseline variables will be collected:

• Demographic data (age, sex, patient history).
• Reason for ICU admission and for initiation of mechanical ventilation.
• Severity scores (SAPS III and APACHE II).
• Duration of mechanical ventilation prior to the first SBT.

Prior to SBT initiation A comprehensive set of measurements will be performed immediately before the trial begins:

• Vital Signs and Ventilator Parameters: Heart rate, blood pressure, respiratory rate, support level, tidal volume, FiO2, and PEEP.
• Analytical Parameters: Arterial blood gases (pH, O2, CO2), and the most recent laboratory results for renal function, liver function, electrolytes, and ProBNP.
• Ultrasound Assessments: A complete set of pulmonary, cardiac, and VExUS ultrasounds will be performed and stored for later analysis by blinded experts.

The SBT will be conducted according to each center's protocol, lasting between 30 and 120 minutes via T-tube or pressure support. Immediately before the patient is reconnected to the ventilator (in case of failure) or extubated (in case of success), a second set of measurements will be taken, including vital signs, an arterial blood gas sample, Pro-BNP level, and repeat pulmonary and cardiac ultrasounds.

Definition of Weaning Failure Weaning failure, the primary study outcome, is defined as either a failure of the spontaneous breathing trial or a failure of extubation.Failure of the Spontaneous Breathing Trial: The SBT is interrupted due to clinical intolerance, requiring reconnection to the ventilator. Extubation Failure: The patient requires rescue with invasive or non-invasive mechanical ventilation within 48 hours of extubation, or the patient dies within 48 hours of extubation.

At ICU Discharge The final variables to be collected include:

• Fluid balance from admission to the day of the SBT.
• Total fluid balance
• Total duration of mechanical ventilation, including the number of days from the first SBT to final extubation, and ICU/hospital length of stay.
• Patient status (alive/dead) at ICU discharge, 30 days post-discharge, and 90 days post-discharge.
• Data Management and Quality Assurance The study will utilize a custom database designed in the RedCap application at the coordinating center, Hospital de la Santa Creu i Sant Pau. This system is fully compliant with all applicable data protection regulations. Upon completion of the study, the database will be irreversibly anonymized to protect participant confidentiality.

The General Coordinating team holds responsibility for monitoring data quality across all participating centers. This process includes a random review of 5% of participants from each center to control for data accuracy and completeness.

The Coordinating Team will issue automated alerts to investigators to recover any missing information. To ensure the integrity of the final dataset, any participant with missing primary variables or more than 10% of missing secondary variables will be excluded from the final analysis.

Each site's principal investigator is responsible for ensuring that the data entered into the electronic Case Report Form (CRF) is accurate and consistent with source documents, such as patient medical records. Investigators must agree to provide direct access to these source documents for monitoring, auditing, and inspection by ethics committees or regulatory authorities upon request.

• Statistical Analysis Plan

• Data Presentation Qualitative Variables: Data will be summarized as absolute numbers and corresponding percentages.

Quantitative Variables: Data will be presented as mean and 95% interval confidence, or as median and interquartile range for non-normally distributed variables. The Kolmogorov-Smirnov test will be used to assess data distribution.

• Comparative Analysis Group Comparisons: Differences between groups will be analyzed using the Student's t-test, χ2 test, or Mann-Whitney U test, as appropriate for the data type and distribution.

Paired Comparisons (pre- vs. post-SBT): Measurements taken before and after the SBT will be compared using a paired Student's t-test or a Wilcoxon signed-rank test, depending on data distribution.

Predictive Performance Analysis To address the primary objective, Receiver Operating Characteristic (ROC) curve analysis will be employed to evaluate the ability of VExUS and other methods to predict weaning failure. The optimal threshold value for each predictive index will be determined using the Youden index, which maximizes the sum of sensitivity and specificity. The non-parametric DeLong method will be used to formally compare the areas under the ROC curves for different predictors.

Software All statistical analyses will be conducted using the SPSS v25 and Stata software packages.

• Acknowledged Protocol Limitations

• Operator-Dependence of Ultrasound: Ultrasound assessments are inherently operator-dependent, meaning that results can vary based on the skill and technique of the person performing the scan. The protocol mitigates this by requiring all assessments to be performed by trained personnel, but inter-observer variability remains a potential limitation.
• Heterogeneity of Pre-SBT Care: While the protocol standardizes the SBT procedure itself, the clinical management of patients on mechanical ventilation leading up to the SBT may differ between participating centers. This variability in pre-trial care introduces a potential confounding factor that could influence patient outcomes.
• Inherent VExUS Limitations: The VExUS score itself has known limitations, as certain clinical conditions can influence its measurements and alter the results. Although the protocol addresses most of these through its detailed exclusion criteria, they represent a limitation of the tool itself that must be considered when interpreting the findings.

Inclusion criteria

• Patients aged > 18 years.
• Admitted to an Intensive Care Unit (ICU).
• Requiring invasive mechanical ventilation for more than 48 hours.
• Meeting the center's standard criteria to begin weaning and undergo an SBT.

Exclusion criteria

• Inadequate ultrasound window for assessment.
• Presence of "do not reintubate" orders.
• Patients with a tracheostomy.
• Presence of atrial fibrillation.
• Diagnosed liver cirrhosis (portal hypertension).
• Absence of qualified personnel to perform the ultrasound at the time of the SBT.
• Denial of informed consent.