YSPH Biostatistics Seminar: “We’re granular, we’re nationwide: Scaling geographic access with big data and high-performance computing.”

NOTE: BIS 526 students are required to attend in person. Others are invited to attend in person, but may also attend via zoom.

Title: "We’re granular, we’re nationwide: Scaling geographic access with big data and high-performance computing.”

Abstract

Geographic access to healthcare resources has a dramatic impact on healthcare utilization and outcomes. Quantifying access is important for understanding how it differentially affects geographically delineated subpopulations. While analysis of specific regions is locally meaningful, the most robust insights come from large-scale (i.e. nationwide) analyses – an approach that demands considerable computational resources and technical acumen. Harnessing open-source software and massively parallel high-performance compute, we have developed a methodology that makes highly granular (census block) yet large-scale (nationwide) travel time analysis feasible. We have applied this approach to projects analyzing access to a diverse assortment of healthcare resources. Furthermore, the robustness of the approach allows us to implement two novel analytic products with potentially broad application: empirically derived travel time thresholds and empirically-derived catchment areas.