Reliability of retrospective chlorophenol exposure estimates over five decades

Area of Science:

• Occupational Health
• Environmental Epidemiology
• Industrial Hygiene

Background:

• Assessing historical occupational exposures is crucial for epidemiological studies.
• Sawmill workers exposed to chlorophenates present a unique cohort for exposure assessment.
• Reliability of retrospective exposure estimation methods requires thorough evaluation.

Purpose of the Study:

• To estimate historical chlorophenate exposures in sawmill workers from the 1940s to 1980s.
• To assess the reliability of these historical exposure estimates using intraclass correlation coefficients.
• To determine the impact of time on the accuracy and stability of exposure assessments.

Main Methods:

• A cohort of sawmill workers was studied, with historical exposures estimated by experienced raters.
• Mill histories were divided into stable exposure eras.
• Individual (ICCind) and group (ICCgroup) intraclass correlation coefficients were calculated to measure rater agreement and estimate stability.

Main Results:

• Individual rater agreement (ICCind) was low, though it improved over time (mean 0.24 to 0.35).
• Group estimates showed high stability (ICCgroup means increasing from 0.78 to 0.88), indicating sufficient raters.
• Reliability of exposure estimates decreased for more distant historical periods.

Conclusions:

• While individual rater agreement was limited, the use of multiple raters ensured stable group estimates of historical exposure.
• The validity of exposure estimates is threatened more by random misclassification in earlier time periods.
• These findings highlight the challenges and necessary methodologies for reconstructing past occupational exposures in epidemiological research.