Oral Presentation Clinical Oncology Society of Australia 2014 Annual Scientific Meeting

The National Indigenous Cervical Screening Project: can we finally identify why Indigenous women fare worse with cervical cancer? (The Queensland linkage results) (#157)

Lisa Whop 1 , Gail Garvey 1 , Catherine Taylor 2 , Abbey Diaz 1 , Patricia C Valery 1 , Peter Baade 3 , Joan Cunningham 1 , Julia ML Brotherton 4 , Karen Canfell 5 , Kamalini Lokuge 6 , Dianne L O’Connell 7 , Suzanne P Moore 1 , John R Condon 1
  1. Menzies School of Health Research, Brisbane, QLD, Australia
  2. Queensland Health, Brisbane, QLD, Australia
  3. Cancer Council Queensland, Brisbane, QLD , Aus
  4. Victorian Cytology Service, Melbourne, VIC, Australia
  5. University of New South Wales, Sydney, NSW, Australia
  6. Australian National University, Canberra, ACT, Australia
  7. Cancer Council New South Wales, Sydney, NSW

Background: Indigenous Australian women have much higher cervical cancer incidence and lower survival than non-Indigenous women.(1) The National Cervical Screening Program (NCSP) has reduced cervical cancer incidence in Australia, but it cannot report screening participation or outcomes for Indigenous women because Pap Smear Registers (PSRs) do not routinely record Indigenous status.

Aim: Tobuse record-linkage to identify Indigenous women in PSRs, to measure their screening participation and outcomes, including time trends and regional variation.

Methods: In Queensland, the study cohort was defined as women aged 20-69 years with at least one Pap test recorded on the PSR in 1999-2011, linked to women with cervical cancer recorded on the Cancer Registry (QCR) in 1997-2011. This cohort was probabilistically linked to two extracts of public hospital data (1995-2011): extract 1, all episodes of women ever-identified as Indigenous; and extract 2, all episodes with a diagnosis/procedure code indicating cervical cancer diagnosis/treatment.

Results: Study approvals for Queensland took 17.5 months and the record-linkage process 3.5 months with approximately 30,000 clerical reviews. The study cohort (PSR+QCR); consists of 1.4 million unique women with 5 million clinical records. Linkage to hospital extract 1 identified 28,872 (2.1%) Indigenous women in the study cohort; 3.3% of the Queensland female population aged 20-69 are Indigenous. Linkage to extract 2 matched to one or more hospital inpatient episodes for 1385 (71%) of the 1955 women identified with cervical cancer in the QCR, those which did not link were notified by private hospitals.

Conclusion: Record-linkage has produced data that are being used to assess NCSP performance for Indigenous women for the first time. However it was a lengthy, resource intensive and costly process that is probably not feasible for regular reporting. Alternatives, such as having Indigenous identification on pathology request forms, have been discussed at length but never implemented.

  1. Australian Institute of Health and Welfare & Cancer Australia 2013. Cancer in Aboriginal and Torres Strait Islander peoples of Australia: an overview. Cancer series no.78. Cat. no. CAN 75. Canberra: AIHW.