Methodological issues involving the standardized mortality ratio and proportionate mortality ratio in occupational studies.

The algebraic relationship between the standardized mortality ratio (SMR) and the proportionate mortality ratio (PMR) is derived. This relationship clearly shows that the validity of PMR as a risk assessment index depends on the homogeneity of the age-specific overall SMRs and the value of the "average" overall SMR (usually estimated by the age-adjusted overall SMR). Based on actual data from various epidemiologic studies, the adequacy, or lack of it, of using the PMR or the converted SMR is examined. Variations of the PMR method using different bases can be useful. Incomplete death ascertainment, which is a common problem in PMR studies, can lead to biased estimates.