Lung Cancer Screening Centralization Is Associated With Improved Screening Uptake: The Veterans Healthcare Administration's Experience 2015-2021.

Abstract

Lung cancer is the leading cause of cancer mortality, yet lung cancer screening (LCS) remains underutilized. Centralizing LCS into dedicated teams with tracking systems (vs. decentralized, individual provider-led screening) may improve LCS uptake, but effectiveness across diverse settings and populations is uncertain. Centralization also varies: hybrid programs share responsibilities with primary care clinicians, whereas fully-centralized programs manage nearly all screening and patient navigation. Whether one model is more effective in enrolling patients remains unknown. Is level of LCS program centralization associated with increased LCS uptake? We performed a retrospective, longitudinal nationwide cohort study using a difference-in-differences Poisson model using Veterans Health Administration facilities. We included Veterans aged 55-80 years old who entered screening from October 1, 2015 through September 30, 2021. The primary exposure was Facility LCS program type (decentralized vs hybrid vs fully-centralized). The main outcome was the incident rate ratio (IRR) of facility-level quarterly LCS uptake, normalized to their estimated LCS-eligible population. We identified 151,194 unique Veterans who underwent screening for our analysis. Overall, LCS uptake was lower for Hispanic Veterans (IRR=0.92, 95% CI [0.90-0.93]), women (IRR=0.80 [0.74-0.86]), those ≥65 years (IRR=0.81 [0.78-0.84]), and rural Veterans (IRR=0.96 [0.95-0.97]). Compared to decentralized programs, increasing centralization was associated with increasing LCS uptake (hybrid programs IRR 1.47 [1.25-1.72], fully-centralized programs IRR=1.75 [1.49-2.07]). Hybrid programs were associated with particularly improved LCS uptake for rural Veterans and more modestly for Veterans from more disadvantaged regions, while fully-centralized programs were associated with improved LCS uptake for Veterans from more disadvantaged regions and Veterans with higher out-of-pocket costs, though they recruited fewer rural Veterans. Both hybrid and fully-centralized LCS programs were associated with improved LCS uptake, with subgroup differences suggesting model-specific advantages. However, persistent disparities highlight the need for tailored approaches to expand screening.