Estimating the total prevalence and incidence of end-stage kidney disease among Aboriginal and non-Aboriginal populations in the Northern Territory of Australia, using multiple data sources

Lin Li, Steven Guthridge, Shu Qin Li, Yuejen Zhao, Paul Lawton, Alan Cass

Most estimates for End Stage Kidney Disease (ESKD) prevalence and incidence are based on renal replacement therapy (RRT) registers. However, not all people with ESKD will commence RRT and estimates based only on RRT registry data will underestimate the true burden of ESKD in the community. This study estimates the total number of Northern Territory (NT) residents with ESKD including: those receiving RRT, those diagnosed but not receiving RRT and an estimate of "undiagnosed" cases. Four data sources were used to identify NT residents with a diagnosis of ESKD: public hospital admissions, Australia and New Zealand Dialysis and Transplant Registry registrations, death registrations and, for the Aboriginal population only, electronic primary care records. Three data sources contained information recorded between 1 July 2008 and 31 December 2013, death registration data extended to 31 December 2014 to capture 2013 prevalent cases. A capture-recapture method was used to estimate both diagnosed and undiagnosed cases by making use of probability patterns of overlapping multiple data sources. In 2013, the estimated ESKD prevalence in the NT Aboriginal population was 11.01 (95% confidence interval (CI) 10.24-11.78) per 1000, and 0.90 (95% CI 0.76-1.05) per 1000 in the NT non-Aboriginal population. The age-adjusted rates were 17.97 (95% CI 17.82-18.11) and 1.07 (95% CI 1.05-1.09) per 1000 in the NT Aboriginal and non-Aboriginal populations respectively. The proportion of individuals receiving RRT was 71.4% of Aboriginal and 75.5% of non-Aboriginal prevalent ESKD cases. The age-adjusted ESKD incidence was also greater for the Aboriginal (5.26 (95% CI 4.44-6.08) per 1000 population) than non-Aboriginal population (0.36 (95% CI 0.25-0.47) per 1000). This study provides comprehensive estimates of the burden of ESKD including those cases that are not identified in relevant health data sources. The results are important for informing strategies to reduce the total burden of ESKD and to manage the potential unmet demand, particularly from comparatively young Aboriginal patients who may be suitable for RRT but do not currently access the services for social, geographic or cultural reasons.