Formulae for calculating subgroup disease burden from overall and reported or assumed relative burden estimates

The risk of disease varies across populations based on factors like age, sex, race, ethnicity, socioeconomic status, and underlying medical conditions. Subgroup or subpopulation data are critical in planning, executing and evaluating public health interventions. However, most studies report the values for the overall (total) population with little or no information on the subgroups. As a result, finding subgroup specific data can be challenging.

In this report, a set of formulae that can be used to calculate subgroup or subpopulation data using the overall estimates and the reported or assumed relative estimates were derived.

A simple numerical example was used to illustrate the methodology. Next, symbolic formula for calculating the burden (e.g., incidence, prevalence, or average cost) for 3 (and extended to n number of) subgroups or subpopulations were derived. To account for uncertainty in the data, two statistical methods were used to estimate confidence intervals for the point estimates.

The derived formulae indicated that each subgroup or subpopulation's burden (incidence, prevalence, or average cost) can be calculated as the overall burden adjusted by the ratio of that subgroup or subpopulation's relative burden to the sum of the proportion-weighted relative burden (incidence, prevalence, or average cost) of all the subgroups or subpopulations within the population.

These formulae can help to avoid or minimize potential quantitative and qualitative errors in subgroup or subpopulation disease burden estimates used for health research, interventions and/or policy analyses or deliberations.