Evaluating occupational hazards using information known only to employers: a comparative study.

Area of Science:

• Occupational health
• Epidemiology
• Biostatistics

Background:

• Occupational health studies often rely on employer records, raising concerns about potential data bias.
• The accuracy of mortality data in these studies can be affected by deaths unknown to the employer (UTE).

Purpose of the Study:

• To empirically assess the bias in occupational health studies using employer records.
• To compare mortality ratios derived from different data sources (known to employer vs. total deaths).

Main Methods:

• Comparative analysis of standardized mortality ratios (SMRs) and proportional mortality ratios (PMRs).
• Inclusion of five occupational cohorts with extensive follow-up.
• Comparison of PMRs based on deaths known to employer (KTE), unknown to employer (UTE), and total deaths.

Main Results:

• Bias was minimal for major causes of death (all cancer, respiratory cancer, heart disease, accidents) when UTE deaths were below 75%.
• A significant inverse relationship was found between the percentage of UTE deaths and the agreement between KTE-based PMRs and total death PMRs.
• Workers leaving the industry showed similar overall mortality to those remaining, supporting the findings.

Conclusions:

• Employer-based occupational health studies are generally reliable when UTE deaths are infrequent.
• The proportion of UTE deaths is a critical indicator of potential bias in mortality data.
• Estimating disease risks using KTE deaths can be a valuable approach when complete data is limited.