Key Point

ARDS is often underrecognized by clinicians, and the several clinical definitions for the condition vary widely in their specificity, which can further compound the problem of diagnosis.

TORONTO—Two clinical definitions are commonly employed to diagnose ARDS in study patients: the lung injury score and the American-European Consensus Conference definition. However, outside of research settings, these criteria are not necessarily used by clinicians. In fact, the American-European definition has performed poorly in limited reliability testing, and there is only moderate agreement between the two definitions.

In 2004, an international team of researchers compared the American-European definition of ARDS with autopsy findings of diffuse alveolar damage in a series of 382 patients and found that the sensitivity and specificity of the definition were only moderate.¹ More recently, they examined a subset of the original study population and compared the lung injury score and Delphi definitions of ARDS with the American-European definition. Clinicians’ accuracy in diagnosing ARDS was also determined.²

PUTTING ARDS DEFINITIONS TO THE TEST

All definitions were applied on a daily basis without knowledge of a patient’s disease evolution. Patients were included in the study if they had undergone intubation, mechanical ventilation, and autopsy. Patients’ clinical records were reviewed by two intensivists who were blinded to the pathology findings. The reviewers recorded source data used to determine the presence of ARDS according to the three definitions. Demographic information, ARDS risk factors, Simplified Acute Physiology Score II (SAPS II), and reason for ICU admission were all recorded, as were the time and cause of death. In addition, PaO2, FiO2, PEEP, peak inspiratory pressure, tidal volume, and pulmonary artery occlusion pressure were all recorded daily.

To determine whether the treating physician had recorded a clinical diagnosis of ARDS, patient charts were examined for any mention of ARDS or acute lung injury. The reviewers determined presence of ARDS by using computer algorithms according to specific definition criteria. Patients were classified as having ARDS according to a particular definition if they met its criteria on any day.

A total of 145 patients met the inclusion criteria. Of these, 138 patients were evaluable. Forty-two had evidence of diffuse alveolar damage at autopsy. However, only 20 patients had a diagnosis or suspicion of ARDS or acute lung injury recorded on their charts by a clinician. Of the 22 patients who were not recognized as having ARDS, 18 could be classified as having ARDS according to one or more of the definitions (15 by American-European definition, 15 by the lung injury score, and 13 by the Delphi definition). Patients identified as having ARDS met the American-European definition on 66.7% of the days they spent in the ICU; the lung score criteria, on 57.1% of days in the ICU; and the Delphi definition criteria, on 50% of days in the ICU.

Agreement between the Delphi definition and the lung injury score was good, but both showed significant disagreement with the American-European definition. Among patients who fulfilled the criteria for one or more clinical definitions of ARDS but did not have ARDS on autopsy, the most common clinical finding at autopsy was pneumonia.

ARDS ISN’T ALWAYS OBVIOUS

One point to keep in mind is that ARDS is rarely the primary diagnosis. "ARDS does not occur in isolation; it arises from either direct (eg, pneumonia or aspiration) or indirect (eg, sepsis or trauma) injury to the lung," explained Niall D. Ferguson, MD, Assistant Professor of Medicine at the University of Toronto and Director of Clinical Research (Critical Care) at University Health Network at Toronto Western Hospital. "I believe that clinicians may be focused on the primary event rather than its complications," he continued. "In addition, I think that many clinicians think of ARDS when there is four-quadrant airspace disease on a chest film and severe hypoxemia—but it’s clear that this represents only one extreme, and many patients may present with less striking findings."

Dr. Ferguson pointed out that asking the respiratory therapist to report the PaO2/FiO2 ratio during routine bedside rounds could help physicians to detect patients with ARDS. "When a patient is on 40% oxygen, ARDS may not spring immediately to mind," he said. "But hearing that their P/F ratio is low can act as a trigger for clinicians to take a closer look at the chest films and examine the clinical situation for ARDS risk factors."

While acknowledging that there aren’t as yet any effective pharmacologic therapies for ARDS, Dr. Ferguson noted that patients with ARDS would benefit from low tidal volume ventilation, which prevents and/or limits ventilator-induced lung injury.

—Gale Jurasek

Reference
1. Esteban A, Fernández-Segoviano P, Frutos-Vivar F, et al. Comparison of clinical criteria for the acute respiratory distress syndrome with autopsy findings. Ann Intern Med. 2004;141:440-445.
2. Ferguson ND, Frutos-Vivar F, Esteban A, et al. Acute respiratory distress syndrome: underrecognition by clinicians and diagnostic accuracy of three clinical definitions. Crit Care Med. 2005;33:2228-2234.

Return to table of contents