Assessment of reliability of lung function screening programs or longitudinal studies

Area of Science:

• Occupational Health
• Pulmonary Medicine
• Biostatistics

Background:

• Lung function measurements are crucial for monitoring occupational respiratory health, particularly in mining environments.
• Ensuring the reliability of these measurements is vital for accurate health assessments and decision-making in screening programs.
• The American Thoracic Society (ATS) provides guidelines for standardized lung function testing.

Purpose of the Study:

• To assess the reliability of lung function measurements conducted under ATS guidelines within a South African gold mine screening program.
• To evaluate the utility of the reliability coefficient G for ongoing monitoring of lung function measurement reliability in mass screening.

Main Methods:

• Calculated the reliability coefficient G, representing random error relative to total measurement variation.
• Analyzed G as a correlation between consecutive lung function tests in 3,378 miners over 43 months.
• Determined optimal sample size and testing frequency for monitoring reliability fluctuations.

Main Results:

• Significant temporal variability in reliability was observed.
• For FEV(1), coefficient G initially varied, stabilized at 0.93 for 23 months, then declined.
• An optimal sample size of approximately 900 randomly examined miners annually is sufficient for monitoring reliability.

Conclusions:

• The reliability coefficient G is a valuable tool for improving data quality in lung function screening programs.
• Monitoring reliability with coefficient G enhances confidence in decisions regarding temporal changes in individual lung function.
• The time interval between tests up to 15 months did not affect the reliability coefficient G's value.