Impact of population density on mortality in patients hospitalized for heart failure – JROAD-DPC Registry Analysis –

Background: There is little evidence regarding social and geographic factors in patients with heart failure (HF). We sought to investigate the association between in-hospital mortality of patients with HF and population density of the patients’ area of residence in Japan. Methods: The present study is a retrospective cohort analysis of the nationwide claim-based database, the Japanese Registry Of All cardiac and vascular Diseases-Diagnostic Procedure Combination (JROAD-DPC). We selected data of 196,286 hospitalized patients with HF (median age of 81 years and 51.8% males). Results: In-hospital mortality was 14.2%, 11.8%, and 9.5% in the lower, middle, and upper tertiles of population density, respectively. Age sub-analysis showed the largest absolute rural-urban disparity in mortality was 4.6% in the oldest subgroup (≥80 years). Multivariate analysis using mixed logistic regression model revealed that higher population density was associated with lower in-hospital mortality: multivariable-adjusted odds ratio (OR): 0.785, 95% confidence interval (CI): (0.713–0.864), p < 0.001 and 0.687, 95% CI: (0.623–0.757), p < 0.001 in the middle and upper tertiles, in comparison with the lower tertile as a reference, after adjustment for age, sex, the New York Heart Association class, comorbidities, and hospital volume. The same tendency was observed in propensity score analysis using 62,291 (in the lower vs. middle tertile) and 57,228 (in the lower vs. upper tertile) matched pairs (OR: 0.797, 95% CI: 0.725-0.877, p < 0.001 and OR: 0.695, 95% CI: 0.634–0.762, p < 0.001 in the middle and upper tertile, respectively). Conclusions: Higher population density was associated with lower in-hospital mortality in HF. More research is needed to gain insight into causality.