HRs by multivariable Cox regression for each of the potential risk factors for 1-year survival in incident patients on RRT. Age group, ethnicity, diabetes, primary renal diagnosis, first treatment modality, healthcare sector (public/private), province of residence and sex are all modelled individually and adjusted by a different suite of confounders. The confounders are selected based on the prior knowledge that they have an association with mortality and are associated with the variable of interest but not an effect of the variable of interest nor a factor in the causal pathway of mortality. Age group was adjusted for healthcare sector and ethnicity; ethnicity was adjusted for age and healthcare sector; diabetes was adjusted for age, ethnicity and healthcare sector; primary renal diagnosis was adjusted for age, ethnicity and healthcare sector; first modality was adjusted for age, ethnicity, diabetes, primary diagnosis and healthcare sector; healthcare sector was adjusted for age and ethnicity; province was adjusted for ethnicity and healthcare sector and sex was adjusted for age.