P LAWTON1, J CUNNINGHAM1, Y ZHAO2, M JOSE3,A CASS1
1Menzies School Of Health Research, Darwin, Australia, 2Health Gains Planning Branch, NT Department of Health, Darwin, Australia, 3Faculty of Medicine, University of Tasmania, Hobart, Australia
In his research, he addresses questions about kidney disease care disparities and outcomes among Indigenous Australians, using larger already existing datasets, including some data linkage. How can we do better for disadvantaged populations, and why aren’t we?
Background: Many clinicians are concerned about Indigenous transplant outcomes, comparing them to non-Indigenous outcomes. Without randomized controlled trials, observational data is needed to measure survival after kidney transplantation compared to dialysis-only treatment, a more patient-centred comparison.
Aim: To explore any survival benefit of kidney transplantation over dialysis-only treatment for both Indigenous and non-Indigenous patients.
Methods: Using ANZDATA, all Australians commencing renal replacement therapy from 1st April 1995 were followed until 31st December 2015. Baseline characteristics were used to create propensity scores for transplantation, which were used for Indigenous and non-Indigenous parallel 1:1 matched case-control studies (transplanted vs. dialysis-only). All-cause survival was compared using life-table methods, and stratified Cox proportional hazards models for three time periods post-transplant (given non-proportional hazards), adjusted for demographic and clinical differences and a transplanted-remoteness interaction term. Quantitative sensitivity analysis measured the effect size unmeasured confounders would need to explain any survival differences seen.
Results: Unadjusted five year survival was better for transplanted patients than their well-matched dialysis-only pair in non-Indigenous (p<0.0001) and Indigenous studies (p=0.0005). Adjusted Cox models showed early (0-0.25 years post-transplant) survival equivalence within both Indigenous and non-Indigenous studies, with improvements in subsequent transplanted survival clearest for all non-Indigenous patients [major cities (MC) Hazard Ratio (HR) 0.16 (0.13-0.19)] except from very remote areas, and for Indigenous patients in MC [HR 0.12 (0.03-0.59)] and inner regional (IR) areas [HR 0.0 (no deaths)]. Unmeasured confounders would need very large effect sizes for both Indigenous (HR >3.27) and non-Indigenous patients (HR>6.75) to account for survival differences.
Conclusions: Indigenous transplanted patients have similar or better survival to similar dialysis-only patients, with long-term benefit in MC/IR. Modern epidemiological methods can minimize or quantify biases.
Dr Paul Lawton is a kidney specialist who has been working as a clinician across the Northern Territory since 1999, including four years as Director of Northern Territory Renal Services.