Purpose: Motivated to generate a series of stylized examples in cancer prevention that incorporate interventions and outcome measures that reflect disparities, we use a generic model for cancer screening to comparatively assess strategies that increase the number of people undergoing screening and those that increase adherence to screening among those already participating in the program.

Methods: We developed a state transition model to project lifetime costs and life expectancy for an annual cancer screening program for a population with a lifetime cancer incidence of 4.7%. We split the population into high-risk (HR) and low-risk (LR) individuals, where the HR group has twice the lifetime cancer than the LR group. We modeled screening disparities in terms of the proportion of the population getting screened (50% LR; 35% HR) and in terms of adherence among those already getting screened (50% LR; 35% HR). We evaluated both targeted strategies (only HR individuals change screening behavior) and untargeted strategies (equal relative effects on screening behavior for both groups).

Results: Compared with the status quo screening scenario, the benefit for strategies aimed at increasing the number of people undergoing screening was almost 5 times greater than strategies aimed at increasing adherence among people already being screened. The targeted approach to increase the number of screeners had a benefit of 0.35 discounted months, whereas the untargeted approach had a benefit of 0.26. In addition, the targeted strategy was associated with 3% fewer lifetime screening tests. Both strategies had a lower disparity measure (defined as the absolute difference in life expectancy between risk groups) than status quo screening, although this measure was lowest for the targeted strategies. From a cost-effectiveness perspective, we found that the targeted strategy remained efficient even if the cost per additional person screened under this scenario was as much as 20 times greater than the cost per additional person screened under the untargeted scenario.

Conclusions: Simple models such as this one can provide insight into the comparative costs and benefits associated with strategies intended to provide average health gains for a population versus strategies intended to reduce disparities.