Related Experiment Videos
Interobserver and intraobserver variability in lung scan reading in suspected pulmonary embolism
Clinical Nuclear Medicine
|November 1, 1980
Summary
Lung scan interpretation shows low agreement between and within readers, especially for minor defects. This variability suggests lung scan readings may not be clinically useful for detecting smaller pulmonary embolism abnormalities.
Area of Science:
- Radiology
- Diagnostic Imaging
Background:
- Pulmonary embolism diagnosis relies on imaging.
- Lung scans are used to detect pulmonary embolism but their interpretation can be subjective.
Purpose of the Study:
- To assess interobserver and intraobserver variability in lung scan interpretation.
- To determine the clinical utility of lung scan readings.
Main Methods:
- Two experienced radiologists independently reviewed 96 lung scans from patients with suspected pulmonary embolism.
- Each radiologist read the scans twice, blinded to previous readings and clinical information.
- Agreement was measured using Kappa statistic.
Main Results:
- Low intraobserver agreement was observed, particularly for scans with minor defects (Kappa < .500).
- Interobserver agreement was even lower than intraobserver agreement.
- The distribution of abnormalities was similar in suspect and reference cases.
Conclusions:
- The reproducibility of lung scan readings is likely too low for clinical use in detecting segmental or smaller defects.
- Further studies are needed to assess reproducibility for lobar defects due to their low incidence in this cohort.