Interobserver agreement using computed radiography in the adult intensive care unit

Area of Science:

• Radiology
• Medical Imaging
• Intensive Care Medicine

Background:

• Conventional radiography is a cornerstone of diagnosis in intensive care units (ICUs).
• Assessing radiologic findings and interval changes is crucial for patient management in ICUs.
• Interobserver agreement can be variable with conventional imaging techniques.

Purpose of the Study:

• To compare computed radiography (CR) with conventional radiography (CR) in an adult ICU setting.
• To evaluate reader agreement in describing radiologic findings and interval changes between CR and conventional radiographs.
• To assess the overall quality and diagnostic utility of CR versus conventional radiographs.

Main Methods:

• Simultaneous acquisition of conventional and computed radiographs using a sandwich cassette in 28 adult ICU patients.
• Four readers evaluated chest films, assessing radiologic descriptors and interval changes using standardized confidence scales.
• Evaluation included overall exposure and film quality assessment.

Main Results:

• Computed radiographs were rated as having better overall quality compared to conventional radiographs.
• Detection accuracy for tubes and lines was comparable between computed and conventional radiographs.
• Higher interobserver agreement was observed with computed radiographs for radiologic descriptors, but not for interval changes.

Conclusions:

• Computed radiography demonstrates diagnostic usefulness comparable to conventional films in the ICU.
• The enhanced interobserver agreement for radiologic descriptors with computed radiography represents a significant advantage.
• Computed radiography is a viable alternative for chest imaging in the intensive care setting.