Socioeconomic Disadvantage, All-Cause and Cause-Specific Mortality in Patients Treated With Maintenance Dialysis: A Mediation Analysis of Geographical Inequity and Multimorbidity.

Rationale & Objective: Social gradient in health (a "social gradient in health" refers to the observed pattern in which individuals with lower socioeconomic status typically experience poorer health outcomes than those with higher socioeconomic status. This indicates that health disparities exist across different social levels, with the most disadvantaged groups experiencing the worst health outcomes) is significant and established in patients with kidney failure, but the pathways of this relationship are unknown. We aimed to assess the mediating effects of multimorbidity and geographical remoteness in the socioeconomic status (SES)-death associations. Study Design: A cohort study. Setting & Participants: All patients with kidney failure aged 18 years and above, who commenced dialysis in Australia from 2005 to 2019. Exposure: Area-level SES. Outcomes: All-cause and cause-specific death. Analytical Approach: The effect of SES on all-cause and cause-specific death was analyzed using the inverse probability stabilized weighting. Mediating effects of geographical remoteness, diabetes mellitus (DM) and cardiovascular disease (CVD) on the association between lower SES and all-cause and cause-specific death were explored. Results: A total of 35,239 patients receiving incident dialysis were included, with a median (p25, p75) follow-up period of 3.3 (1.7-5.9) years. Compared with patients from higher SES, the average hazard rate for all-cause death among those from lower SES was 17% higher (total effect [TE] = 0.17, 95% CI [0.12-0.23]). Proportions of the effects between SES and all-cause mortality mediated by geographical remoteness, CVD, and DM were 29.4%, 11.8%, 17.6%, respectively, whereas SES explained 41.2% of the TE directly. Compared with patients from high SES, patients from lower SES have on average a higher hazard rate of CVD (TE = 0.26, 95% CI, [0.15-0.38]) and infection-related deaths (TE = 0.12, 95% CI, [0-0.25]). The effects of SES on CVD and infection-related deaths were mediated by CVD and DM, but not geographical remoteness. Limitations: Potential residual confounding and other latent mediators. Conclusions: Geographical remoteness, diabetes, and CVD are potential mediators that lie in the pathways between SES and all-cause and cause-specific deaths. A multifaceted approach with sustained efforts from multiple sectors to address these factors may reduce the social disparities observed in patients treated with dialysis.