Lack of broadband access and preventable hospital stays in relation to cardiovascular mortality: a nationwide ecological study

Associations of digital connectivity and primary care access with county-level cardiovascular mortality, and their spatial patterning, remain unclear. We examined associations of lack of broadband access (LBA) and preventable hospital stays (PHS) with cardiovascular disease (CVD) mortality across US counties. We conducted an ecological analysis of 2021–2023 County Health Rankings & Roadmaps and CDC WONDER data. Age-standardized mortality was compared across exposure quartiles and demographic strata. Quasi-Poisson models estimated adjusted rate ratios (RRs). Correlation, bivariate Moran’s I, and Bayesian spatial models assessed county-level associations and spatial patterns. Fully adjusted RRs for the highest vs lowest quartiles were 1.11 (95% confidence interval [CI], 1.09–1.13) for LBA and 1.07 (95% CI, 1.06–1.09) for PHS. Positive gradients were consistent across age and sex strata and most urbanization groups but varied by race. The strongest associations were LBA with acute myocardial infarction mortality (RR, 1.37 [95% CI, 1.28–1.47]) and PHS with heart failure mortality (RR, 1.19 [95% CI, 1.13–1.26]). LBA and PHS were positively correlated (Spearman ρ = 0.410) and showed significant spatial concordance. High-LBA/high-neighboring-PHS clusters were concentrated in the South and Southeast. Associations with total CVD mortality persisted in continuous-exposure and Bayesian spatial models. In this ecological study, higher LBA and PHS were associated with higher county-level CVD mortality and exhibited geographic clustering. Causality cannot be inferred from this ecological design; longitudinal and individual-level studies are warranted.