Availability of national data on cancer in Aboriginal and Torres Strait Islander people has improved.1 Complemented by regional data, they show Aboriginal and Torres Strait Islander people to have a higher age-adjusted incidence (all cancers combined), an excess of more lethal cancer types, more advanced cancer stage at diagnosis, more comorbidity, less comprehensive treatment, and lower survival.1-3 Indigenous cancer disadvantage is clear, albeit poorly quantified. Despite gains, data coverage and quality for Indigenous identification are inadequate, commonly leading to under-estimation of the Indigenous cancer burden and complicating assessment of differences and changes across Australia.4
Data collection has largely relied on recording Indigenous status at point of care in hospitals, general practices, and other health services, plus death records. Major challenges exist in gaining a level of consistent engagement across multiple service outlets to obtain data of adequate quality. Consistency in asking people as to whether they are of Aboriginal or Torres Strait Islander origin, or both, has not been achieved.5, 6 Guidelines for collecting these data have been developed and strong political support given through Council of Australian Governments (COAG) and other sources, and training sessions provided for data collectors, which would have helped, but data inadequacies are appreciable.5, 6 Barriers can include a lack of understanding of the importance of collecting these data, and uncertainty how to ask patients in the context of busy service settings that focus on individual patient care.5, 6 Changing IT systems for data collection can be additional barriers.5
Collecting Indigenous status at the point of care has the advantage of availability to contribute to individual patient care, but it is not a reliable means of collecting nationally consistent data for population-health monitoring and policy making. An international survey of 83 cancer registries in 25 countries showed inadequate collection of Indigenous status, and common barriers.7
More recently, data linkage has improved Indigenous identification in population statistics. Linkage has been particularly advanced in NSW with linkage of cancer registry, inpatient, emergency department, Medicare, screening and other health-related data for population health and health service monitoring and feedback. That this can enhance Indigenous status recording of NSW people with cancer was demonstrated where linkage increased the recorded age-standardised Indigenous cancer incidence by 42% for males and 27% for females.8 In Queensland, Indigenous status was successfully obtained by linking cervical screening data to hospital records.9 In Western Australia, gains in Indigenous identification in health records from data linkage have been documented in multiple settings.10 In Victoria, the identification of Indigenous status for communicable disease surveillance has benefited from data linkage.11 In South Australia, the scope of Indigenous and non-Indigenous cancer data has been increased through data linkage of cancer registry, hospital, Medicare claims, and screening data.3
For accurate determination of cancer rates in Indigenous people, Indigenous status needs to be consistently recorded both for those with cancer (the numerator) and the population (the denominator). This would be theoretically possible through linking cancer data to data from the Australian census. An efficient privacy-protecting means of doing this is needed.
References
1. Australian Institute of Health & Welfare. Cancer in Aboriginal & Torres Strait Islander people of Australia. Cat. no: CAN 109. Canberra: AIHW, 2018.
2. Zhang X, Condon JR, Rumbold AR, Cunningham J, Roder DM. Estimating cancer incidence in Indigenous Australians. Aust NZ J Public Health 2011; 35(5): 477-485.
3. Banham D, Roder D, Eckert M, Howard NJ, Canuto K, Brown A, on behalf of CanDAD Community Reference Group and other CanDAD investigators. Cancer treatment and the risk of cancer death among Aboriginal and non-Aboriginal South Australians: analysis of a matched cohort study. BMC Health Serv Res 2019; 19(1): 771.
4. Ford BK, Kong M, Ward JS, Hocking JS, Fairley CK, Donovan B, Lorch R, Sparks S, Law M, Kaldor J, Guy R. Incomplete recording of Indigenous identification status under-estimates the prevalence of Indigenous population attending Australian general practices: a cross sectional study. BMC Health Services Res 2019; 19(1): 567.
5. Schütze H, Pulver LJ, Harris M. What factors contribute to the continued low rates of Indigenous status identification in urban general practice? – A mixed-methods multiple site case study. BMC Health Serv Res 2017; 17(1):95.
6. Morgan S, Thomson A, Tapley A, O’Mara P, Henderson K, van Driel M, Scott J, Spike N, McArthur L, Magin P. Identification of Aboriginal and Torres Strait Islander status by general practice registrars: Confidence and associations. Australian Family Physician 2016; 45(9): 677-682.
7. Diaz A, Soerjomataram I, Moore S, Whop LJ, Bray F, Hoberg H, Garvey G. Collection and reporting of Indigenous status information in cancer registries around the world. JCO Global Oncology 2020; 6: 133-142.
8. Tervonen HE, Purdie S, Creighton N. Using data linkage to enhance the reporting of cancer outcomes of Aboriginal and Torres Strait Islander people in NSW, Australia. BMC Med Res Methodol 2019; 19(1): 245.
9. Whop LJ, Diaz A, Baade P, Garvey G, Cunningham G, Brotherton JML, Canfell K, Valery PC, O’Connell DL, Taylor C, Moore SP, Condon JR. Using probabilistic record linkage methods to identify Australian Indigenous women on the Queensland Pap Smear Register: The National Indigenous Screening Project. BMJ Open 2016; 6(2): e009540.
10. Thompson SC, Woods JA, Katzenellenbogen JM. The quality of Indigenous identification in administrative health data in Australia: insights from studies using data linkage. BMC Med Inform Decis Mak 2012; 12(1):133.
11. Rowe SL, Cowie BC. Using data linkage to improve the completeness of Aboriginal and Torres Strait Islander status in communicable disease notifications in Victoria. Aust NZ J Public Health 2016; 40(2): 148-153.