Individual and Catchment Area Factors Associated With Breast and Cervical Cancer Screening Within the Military Health System.

Introduction: Breast and cervical cancer screening is critical to identifying cases at earlier stages in order to begin treatment earlier and improve survival. Screening rates have been shown to vary within the Military Health System (MHS). The goal is to estimate drivers of variation in screening rates within the MHS.

Materials and methods: We used 2007-2019 MHS Data Repository Data to examine individual-level and catchment area-level factors associated with 1- and 2-year breast and cervical cancer screening rates. Specifically, we estimated univariate and multivariate association between 1- and 2-year probability of breast and cervical cancer screening rates and age group, marital status, rank, service branch, beneficiary type (service member vs. dependent), race/ethnicity of service members, catchment area fraction of overall care through purchased care and average per capita spending. The project was approved by both the Penn State Institutional Review Board (IRB) and the Defense Health Agency's electronic IRB.

Results: Overall, we observed a 45.6% 1 year and a 65.7% 2-year mammography screening rate and a 30.5% 1 year and a 51.9% 2-year Pap testing rate. For breast cancer screening, we found higher screening rates for older (ages 50-64 years), married, service members, more senior ranked women or those married to more senior ranked members, and non-Hispanic Black women in both unadjusted and multivariate analyses. Conversely, we found higher rates of cervical cancer screening for younger, unmarried, more junior ranked women as well as for non-Hispanic Black women. We also found higher rates for both breast and cervical cancer screening in catchment areas with a greater fraction of care delivered through the private sector.

Conclusion: Our finding of higher screening in catchment areas with higher rates of purchased care warrants additional study to understand what factors may drive this result. The differential findings of the association between individual characteristics and breast and cervical cancer screening suggest important differences in these 2 types of screening with potentially different policies required to encourage and enhance breast vs. cervical cancer screening. Finally, our results showing higher screening among non-Hispanic Black women suggests important features of the MHS, such as universal, low-cost sharing coverage may help to reduce racial and ethnic disparities in breast and cervical cancer screening.