A Small Area Analysis of Mammography Capacity in the Southern United States: Recommendations for Strategic Resource Allocation and Policy-making
Source: MD Anderson Cancer Prevention Research Training Program (R25CA57730)
Period: January 2011-December 2013
Role: Postdoctoral Fellow
The goal of this project was to estimate mammography capacity at the small area level and characterize clusters of geographically contiguous small areas with inadequate mammography capacity.
Mammography screening detects breast cancer at an early stage when treatment is most effective. The American Cancer Society recommends yearly mammograms for women starting at age 40. Utilization varies across age, racial/ethnic groups, income levels, and geographic areas, and access to services is a contributing factor. Inadequate mammography capacity, defined as too few mammography machines to provide services for all eligible women in counties or states, has been linked to under-utilization of mammography screening and late-stage diagnosis of breast cancer. These studies demonstrate a critical need for additional mammography capacity in states and counties, but provide no guidance for placement of new facilities, addition of machines, or routing mobile mammography units. Examination of capacity in small geographic areas and identification of clusters of underserved small areas is required to support the efficient allocation of mammography screening services. We will fill this knowledge gap by estimating mammography capacity at the small area level and characterizing clusters of geographically contiguous small areas with inadequate mammography capacity. I aim to:
- Estimate mammography capacity at the small area level over time.
Small area level capacity at time t will be measured using a modified gravity model, which predicts the distribution of mammography facilities over space and produces a summary measure of capacity for each small area. The measure can be categorized to represent differing levels of access (e.g., high/low).
- Identify and characterize clusters of geographically contiguous small areas with inadequate mammography capacity. Geographic clusters of inadequate mammography capacity will be identified using the Getis-Ord Gi statistic. Descriptive statistics will be used to characterize the populations within these clusters.
Eberth JM, Eschbach K, Morris JS, Hossain Md M, Nguyen HT, Elting LS. Geographic disparities in mammography capacity: a longitudinal assessment of supply and demand. Health Services Research. 2014 Feb; 49(1):171-85. DOI: 10.1111/1475-6773.12081. PMCID: PMC3922472.
Eberth JM, Elting LS, Eschbach K, Morris JS, Mossain Md M, Hguyen HT. Access to mammography in the south: a spatial perspective. Annals of Epidemiology. 2012 September;22(9):662.
Eberth JM, Eschbach K, Morris J, Nguyen HT, Hossain Md M, Elting S. Decreasing mammography capacity in the southern United States, 2002-2008. ESRI International User Conference. 2013 Jul. [Oral]
Eberth JM, Elting LS, Eschbach K, Morris J, Hossain Md M, Hguyen HT. Access to mammography in the South: a spatial perspective. American College of Epidemiology Annual Meeting. Chicago, IL. 2012 Sep. [Poster]
Elting LS, Eberth JM, Chang Y, Nguyen H. Is mammography screening capacity in Texas sufficient to achieve cancer control goals? Cancer Prevention Research Institute of Texas Conference. Austin, TX. 2010 Nov. [Poster]