Hospital-wide Incidence, Clinical Characteristics and Outcomes of ARDS

A Multicenter Prospective Cohort Study of Hospital-wide Incidence, Clinical Characteristics and Outcomes of ARDS - The NITWA ARDS Study

The Acute Respiratory Distress Syndrome (ARDS) impacts one of every four patients requiring mechanical ventilation for respiratory support and carries a mortality rate of 40%. To diagnose ARDS, doctors currently use the Berlin definition, that requires chest radiographs and analysis of oxygenation in the blood (arterial blood gas). These tests are not available in areas of the world with constrained resources and may be unnecessarily invasive. A modification of the Berlin definition, using ultrasound and pulse oximetry (a small device that measures oxygen level non-invasively by clipping to the body, typically a finger), has been recently developed and tested in Kigali, Rwanda.

This study will try to confirm the validity of the Kigali modification initially in Boston and Toronto and subsequently in other hospitals worldwide. If confirmed, this new definition could allow for faster recognition and potentially improved treatment of patients with ARDS and facilitate studies worldwide.

The purposes of this study are:

1. To describe clinical characteristics and outcomes of patients diagnosed with ARDS according to the Berlin and Kigali definitions;
2. To determine how well chest radiograph and ultrasound of the chest are able to define ARDS, in comparison to chest computer tomography (CT).

We hypothesize that the hospital-wide incidence of ARDS, as defined by the Kigali modification, is similar in high resource settings (e.g., Boston and Toronto) as compared to the resource-constrained setting of Kigali, Rwanda. We also hypothesize that pulmonary ultrasound is a more sensitive and similarly specific imaging modality for bilateral opacities than chest radiograph, when compared to the reference standard of chest tomography.

We will test these hypotheses in a multicenter prospective cohort study with the following specific aims:

Aim 1: A) To estimate the hospital-wide incidence of ARDS defined according to both the Berlin definition and the Kigali modification, and B) To describe clinical characteristics and outcomes for these patients.

Aim 2: For the subset of patients who have chest CT, to determine the sensitivity and specificity for bilateral opacities of both chest radiographs and chest ultrasound done within 12 hours as compared to the reference standard CT scans.

As a part of the research study, we will perform a pilot study with the specific aim of assessing feasibility of a multicenter study. Criteria that will be used to assess feasibility include:

1. Number of hospitalised adult patients who fulfill Kigali or Berlin ARDS criteria over the first 7 days post-hospital admission;
2. Number of hospitalised adult patients who develop hypoxemia as detected on daily screening, during the first 7 days post-hospital admission (% hypoxemic patients/new admissions);
3. Proportion of recruited patients/eligible patients (see below for eligibility criteria);
4. Work-load per patient (lung ultrasound scanning time; average data collection time on the first day of hypoxemia);
5. Proportion of patients with CXR, CT scan and LUS available from the same +/-1 day.

All adults (≥ 18 years old) admitted to the hospital during either of two one-week study periods (winter and summer) will be screened daily for hypoxemia (defined as oxygen saturation < 90%) or use of any supplemental oxygen for a total of 7 days. For the initial feasibility phase, both in-person and electronic administrative records screening will be performed. Depending on the site and the results of the pilot phase, in the multicenter study the screening will be accomplished using electronic administrative records or in-person screening.

For any eligible patient who screens positive during the study period we will collect data as detailed in the table below:

Day 1 post-hypoxemia detection

• Demographic characteristics (year of birth, sex, height, weight)
• Admission data (type of admission - elective/emergency; transfer vs direct admission vs ED admission -; date of admission; if transfer from other hospital; ward - medicine, surgery, ICU)
• Main diagnosis/clinical presentation
• Co-morbidities
• ARDS risk factors at admission
• New or worsening symptoms within 7 days
• Institution of mechanical ventilation (invasive or non-invasive)
• Oxygenation data
• Lung Ultrasound data
• CXR and CT scans occurring up to 24 hours before onset of hypoxemia

Day 2-6 post-hypoxemia detection

• Oxygenation data
• Lung Ultrasound data
• Chest imaging

Day 7 post-hypoxemia detection

• Etiology of hypoxemia (as determined by MRP)
• New ARDS risk factors identified
• Need for ICU admission first 7 days
• Institution of mechanical ventilation first 7 days (invasive and non-invasive)
• Oxygenation data
• Lung Ultrasound data
• Structured focused lung ultrasound
• Chest imaging

Outcome data collection

• Vital status at hospital separation, censored at 90 days
• Date of hospital discharge (or death)
• ICU admission and duration of ICU stay

For any eligible patient who does not screen positive during any day of the study period (days 1-7 post hospital admission), we will collect the following data:

• Vital status at hospital separation, censored at 90 days
• Date of hospital discharge (or death)
• ICU admission and duration of ICU stay

For patients undergoing CT chest during the 7 days of data collection, we will attempt the performance of an extra lung ultrasound examination immediately before or after the CT scan

Inclusion Criteria:

• Hospital admission during the study period
• Age ≥ 18 years old
• New onset of hypoxemia (SpO2 < 90% or use of any supplemental oxygen) or, for patients on home O2 treatment, higher than baseline O2 flow needed during the first 7 days of hospitalization.

Exclusion Criteria:

• Patient in the Emergency Department but not formally admitted to the hospital
• Patient admitted in PACU