A reformulation of the serially additive expected dose method for occupational cohort data.

The serially additive expected dose method for the analysis of occupational cohort data can be biased whenever worker exposure is associated with length of follow-up. An illustration shows that this bias can act to either enhance or reduce the degree of association between disease and exposure. The serially additive expected dose method can be modified, then embedded in a classical Cox proportional hazards model context to eliminate the analytic bias. A chisquare test of no exposure association is derived.