Elsevier

Journal of Critical Care

Volume 28, Issue 4, August 2013, Pages 352-357
Journal of Critical Care

Infection
Accuracy of the chest radiograph to identify bilateral pulmonary infiltrates consistent with the diagnosis of acute respiratory distress syndrome using computed tomography as reference standard

https://doi.org/10.1016/j.jcrc.2012.12.002Get rights and content

Abstract

Purpose

The purpose of the study is to evaluate the diagnostic accuracy of the anteroposterior chest radiograph to detect pulmonary abnormalities consistent with acute respiratory distress syndrome (ARDS).

Materials and methods

Ninety patients who met criteria for ARDS regardless of the radiographic one and had near simultaneous chest radiograph and computed tomography (CT) performed were identified. These radiologic studies were reviewed blindly and independently by 2 radiologists for the presence or absence of bilateral pulmonary abnormalities consistent with ARDS using defined radiologic criteria. Disagreements were resolved by consensus. Using the chest CT interpretation as reference standard, the chest radiograph diagnostic parameters were calculated.

Results

Sensitivity (Se) was 0.73; specificity, 0.70; positive and negative predictive values were 0.88 and 0.47, respectively. Female sex was associated with higher Se and lower specificity. When patients were divided according to disease distribution by CT, the Se was significantly lower for focal as compared with diffuse.

Conclusions

The accuracy of the portable chest radiograph to detect pulmonary abnormalities consistent with ARDS is significantly limited. These findings suggest that the use of the chest radiograph results mainly in underrecognition of the syndrome, particularly when disease is not diffusely distributed, but also in overdiagnosis.

Introduction

The presence of bilateral infiltrates on an anteroposterior chest radiograph is a necessary criterion for the diagnosis of acute respiratory distress syndrome (ARDS) by recommendation of the American-European Consensus Conference (AECC) in ARDS [1]. This definition has been the most commonly used in clinical practice and in epidemiologic [2] and other research studies with important clinical implications [3], [4], [5], [6]. However, this radiographic criterion is not precise and can result in significant interobserver variability [7], [8], [9] or equivocal interpretation [10]. The recent Berlin definition of ARDS, likely to be widely used in the future, recommends a similar radiographic criterion of bilateral opacities [11]. Those limitations to the radiographic criterion may have important implications in the correct selection of patients for research studies and in the appropriate diagnosis and consequent application of therapies in clinical practice.

This study aims to evaluate the accuracy of the anteroposterior chest radiograph to detect pulmonary abnormalities consistent with ARDS in patients who may have the syndrome, using chest computed tomography (CT) as reference standard. We conducted a comparison between chest radiograph and chest CT in patients meeting criteria for ARDS regardless of the radiographic one.

Section snippets

Subjects

Subjects were retrospectively selected from patients admitted to a 30-bed adult medical-surgical-trauma intensive care unit (ICU) of a university hospital between March 2009 and October 2011. The selection was performed using initially the hospital radiology computerized database and further refined by review of the subjects' medical records by an intensivist. Informed consent was waived by the local institutional review board given the nature of the study and that data were collected without

Results

Ninety subjects were identified meeting inclusion but no exclusion criteria. Ninety-seven percent were receiving invasive mechanical ventilation. The subjects' characteristics on the day the analyzed images were performed are summarized in Table 1.

There was disagreement between the 2 radiologists in the interpretation of the chest radiograph in 29 subjects. In 25 subjects, the disagreement included possible bilateral findings, whereas in 4 subjects, the disagreement was between unilateral vs

Discussion

By comparing portable chest radiographs with chest CT (as reference standard) in patients meeting criteria for ARDS regardless of the radiographic one, this study shows the limited accuracy of the radiographic criterion of the definition of ARDS. For the detection of “bilateral pulmonary infiltrates or opacities” consistent with the diagnosis of ARDS, we found the chest radiograph to have only fair interobserver agreement [14] between 2 radiologists and, after resolution of disagreements, an

References (26)

  • M.O. Meade et al.

    Interobserver variation in interpreting chest radiographs for the diagnosis of acute respiratory distress syndrome

    Am J Respir Crit Care Med

    (2000)
  • M.O. Meade et al.

    Agreement between alternative classifications of acute respiratory distress syndrome

    Am J Respir Crit Care Med

    (2001)
  • The ARDS Definition Task Force

    Acute respiratory distress syndrome: the Berlin definition

    JAMA

    (2012)
  • Cited by (0)

    Institution where work was performed: University Medical Center of El Paso, El Paso, Texas.

    View full text