Aboriginal and/or Torres Strait Islander people are under-reported in the Australia and New Zealand Dialysis and Transplant Registry (ANZDATA), according to a study comparing registry data with an algorithm for the Enhanced Reporting of Aboriginality.

The findings reinforce the potential for systematic bias in health data collection and the importance of ongoing efforts towards quality improvement.

Outside the research context, they are also a reminder for clinicians to ascertain and record the Indigenous status of patients using the national standard question, “Are you of Aboriginal or Torres Strait Islander origin?”

The retrospective cohort study, published in The MJA [link here], comprised 11,772 people in NSW who commenced kidney replacement therapy – either dialysis or transplantation – between 1 July 2006 and 31 December 2020.

The Enhanced Reporting of Aboriginality (ERA) approach links multiple datasets including public and private hospital admissions, ambulance records, the NSW Registry of Births, Deaths and Marriages, cancer screening participation data and community mental health records.

ERA identified 693 dialysis or transplant recipients (5.9%) as Aboriginal and/or Torres Strait Islander.

Meanwhile ANZDATA uses a single report from the renal unit when patients are diagnosed with end stage kidney failure requiring kidney replacement therapy. It identified 484 people (4.1%) as Aboriginal and/or Torres Strait Islander, including 20 people not recognised using ERA.

The two groups had different patient and clinical characteristics and outcomes.

For example, the study found a significantly lower proportion of people aged under 18 years (1.9% v 4.8%) and over 65 years (21.7% v 27.1%; p=0.028) were captured in ANZDATA compared with ERA only.

“A higher proportion of females were identified in ANZDATA compared with those in ERA alone, although the difference was not statistically significant (47.5% v 40.2%; P = 0.07),” the study said.

As well, ANZDATA captured significantly more of the most disadvantaged people than ERA only (37.6% v 25.3% (p=0.009) and more people living outside major cities (62.8% v 38.4%; p<0.001).

Regarding causes of kidney failure, there were significantly higher rates of glomerulonephritis or IgA nephropathy and polycystic kidney (13.4% v 21.0%) and lower rates of diabetes (60.7% v 36.7%) for the ERA only group.

There were no significant differences in rates of home dialysis, late referral, smoking or mental illness for the group captured in ANZDATA compared with those in the ERA group.

The study found a sensitivity of 67.0% and specificity at a high 99.8% regarding ANZDATA’s accuracy in identifying Indigenous status.

“Standardising the way in which the national standard question is asked across renal units, and inclusion of a preamble, may help to improve sensitivity and consistency within the ANZDATA data collection,” the study said.

Regarding patient outcomes, the study also found people not recognised by ANZDATA but recognised by ERA alone as Aboriginal and/or Torres Strait Islander had better outcomes.

“For the ERA only group, there were significantly higher rates of waitlisting (19.7% v 11.4% within 2 years; P = 0.004), the rate of transplantation within 2 years was more than three times higher (10.9% v 3.1%; P < 0.001), and the rate of death within 5 years was lower (27.9% v 38.0%; P = 0.008),” it said.

Solutions

The investigators, including Professor Stephen McDonald and Professor Angela Webster, said in the study that the under-recognition of Aboriginal and/or Torres Strait Islander people and the observed differences between the groups represented systematic bias.

“Further, routine linkage to external data sources such as those used in our study to enhance reporting would vastly improve ascertainment and reduce bias in research and reporting on kidney health for Aboriginal and/or Torres Strait Islander people,” they said.

Professor Stephen McDonald

Professor McDonald, Executive Director of ANZDATA, told the limbic that the data on Indigenous status is collected from renal units who in turn have collected it from their patients.

“So I guess the first question is, is it people not identifying to renal units, and if that’s the case then why might that be? Or is it simply that they [renal units] are not asking – not going down the usual process?”

One response to the study’s findings may be reminding dialysis and transplant units that there is a standard procedure across the health system for asking people their Indigenous status.

He said the issue of not asking or not recording Indigenous status would not just be isolated to New South Wales.

“This is going to be an issue everywhere. It is interesting that … the discrepancy was greater in metropolitan areas than remote areas and there was also a socioeconomic gradient to that. Does that mean that some renal units are making more assumptions? If somebody doesn’t look particularly indigenous, they may not be bothering to ask them if they’re in Sydney, whereas if they’re remote, it may be something that’s more front of mind.”

“Or it may be that people are choosing not to self identify…and that’s more challenging…and gets back to the presence of institutional bias or institutional racism in our health system. How do we move to a place where we’re providing care that is culturally safe?”

He said a simple example was that hospitals are fundamentally designed for the convenience of those who work in them.

“You come and see me in my place, in my room, on my terms…at the time of my choosing. And that’s a very simple example but that thinking cascades down through the health system.”

“Having said that, there are a bunch of programs on the way to call out those problems and then try and pick through the solutions. It all goes very slowly, and it’s difficult driving change in a resource-constrained environment.”

Professor McDonald said the National Indigenous Kidney Transplantation Taskforce, established in 2019, has run a series of pilot programs and demonstrated that change as possible. Some recommendations were working their way through strategy and budget processes at jurisdictional levels.

He said delivering care in communities, implementing patient navigators and involving Indigenous health practitioners in the process of care were other practical strategies to improve progress towards transplant waitlisting.

Professor Angela Webster

Professor Webster, from the Centre for Renal and Transplant Research at Westmead Hospital, told the limbic that data integration – the ability to easily link across multiple datasets – was critical to delivering equitable care.

“If you want to close the gap and you want to champion Aboriginal health, you need good data to be able to understand what’s going on. We did this as a research exercise, but these datasets are not all joined up in routine care.”

She also agreed that the issue of people not identifying as Indigenous was complex.

“People are choosing not to self identify and maybe that’s a legacy from colonialism. We know there’s racism within the health service, and we know that there are systematic biases at play, unconscious or conscious. We need to work out why people are choosing to hide their heritage; perhaps because they may believe they’ll get better care if people don’t know their origin.”

“This isn’t just us not asking a question or not recording it. There’s something else at play here and it’s likely to be a bit of both. It’s not delivering culturally safe care, not asking the question, not allowing people the space to choose to self identify,” she said.

“I’ve changed my practice since doing this paper. I make sure I’ve asked everybody that comes to the door [about Indigenous status], whereas before maybe I wouldn’t have prioritised that question in the short 20 minutes I’ve got with them. Now I actually do, because I realise it’s so important.”