Related Experiment Videos
Radiologic detection of pleural thickening
H Frumkin1, G Pransky, I Cosmatos
1Clinical Epidemiology Unit, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104.
The American Review of Respiratory Disease
|December 1, 1990
Summary
Radiologic detection of pleural thickening using chest X-rays shows suboptimal accuracy and significant reader variability. Caution is advised when interpreting findings with International Labour Organization procedures due to these limitations.
Area of Science:
- Radiology
- Occupational Medicine
- Pulmonary Medicine
Background:
- Pleural thickening diagnosis is crucial for occupational lung diseases.
- Chest X-rays (CXRs) are commonly used but their accuracy for pleural thickening is debated.
- International Labour Organization (ILO) guidelines standardize CXR interpretation.
Purpose of the Study:
- To evaluate CXR specificity for pleural thickening by B readers.
- To assess the impact of positivity thresholds on test performance.
- To determine interreader variability and film quality effects on CXR accuracy.
Main Methods:
- Review of 421 autopsy-correlated CXRs by six independent B readers.
- Use of modified ILO forms for standardized assessment.
- Analysis of specificity, sensitivity, and interreader agreement (Kappa statistics).
Main Results:
- Average specificity was 0.85 with a strict positivity definition, decreasing with liberal definitions.
- Sensitivity estimates were unstable but appeared to increase with liberal definitions.
- Considerable interreader variability (Kappa 0.06-0.40) and decreased sensitivity with lower film quality were observed.
Conclusions:
- CXR demonstrates suboptimal test characteristics for diagnosing pleural thickening.
- Significant interreader variability necessitates caution in ILO-based interpretations.
- Further research may be needed to improve diagnostic accuracy and consistency.