Equitable cancer control: better data needed for indigenous people

• Published in: The lancet oncology 2015-11, Vol.16 (15), p.1442-1444
• Main Authors: Sarfati, Diana, Robson, Bridget
• Format: Article
• Language: English
• Subjects: Bias; Censuses; Colorectal cancer; Ethnicity; Female; Health risk assessment; Hematology, Oncology and Palliative Medicine; Humans; Indians, North American; Male; Native peoples; Neoplasms - epidemiology; Oceanic Ancestry Group; Studies
• ISSN: 1470-2045; 1474-5488
• DOI: 10.1016/S1470-2045(15)00295-8

Subsequent direct linkage of census data to cancer registry data showed that during that period there was a 30% underestimation of cancer incidence.2 More recently, there has been a substantial focus in New Zealand on improving ethnicity data in all health records, in terms of both completeness and consistency; this effort is reflected in a reduction in the underestimation of cancer incidence over time.2,3 Researchers have also reported substantial underestimation of cancer cases in indigenous populations in the USA4,5 where, for example, incidence rates of cancer in an indigenous population based on augmented indigeneity status in cancer data resulted in incidence rates that were more than twice those estimated with Surveillance, Epidemiology, and End Results data.6 As a result, the incidence data are probably wrong and most likely underestimated in all jurisdictions where this numerator-denominator bias is a problem. Despite the data shortcomings, Moore and colleagues highlight the striking findings relating to the higher incidence of potentially highly preventable cancers, especially those associated with smoking (eg, lung, and head and neck cancers) and chronic infection (including stomach, liver, and cervical cancers).