Study protocol
A multilevel study of the determinants of area-level inequalities in colorectal cancer survival
-
* Corresponding author: Peter D Baade peterbaade@cancerqld.org.au
1 Viertel Centre for Research in Cancer Control, Cancer Council Queensland, PO Box 201, Spring Hill QLD 4004, Australia
2 School of Public Health, Queensland University of Technology, Herston Road, Kelvin Grove 4059, Australia
3 School of Population Health, University of Queensland, Herston Road, Herston 4061, Australia
BMC Cancer 2010, 10:24 doi:10.1186/1471-2407-10-24
Published: 28 January 2010Abstract
Background
In Australia, associations between geographic remoteness, socioeconomic disadvantage, and colorectal cancer (CRC) survival show that survival rates are lowest among residents of geographically remote regions and those living in disadvantaged areas. At present we know very little about the reasons for these inequalities, hence our capacity to intervene to reduce the inequalities is limited.
Methods/Design
This study, the first of its type in Australia, examines the association between CRC survival and key area- and individual-level factors. Specifically, we will use a multilevel framework to investigate the possible determinants of area- and individual-level inequalities in CRC survival and quantify the relative contribution of geographic remoteness, socioeconomic and demographic factors, disease stage, and access to diagnostic and treatment services, to these inequalities. The multilevel analysis will be based on survival data relating to people diagnosed with CRC in Queensland between 1996 and 2005 (n = 22,723) from the Queensland Cancer Registry (QCR), area-level data from other data custodians such as the Australian Bureau of Statistics, and individual-level data from the QCR (including extracting stage from pathology records) and Queensland Hospitals. For a subset of this period (2003 and 2004) we will utilise more detailed, individual-level data (n = 1,966) covering a greater range of risk factors from a concurrent research study. Geo-coding and spatial technology will be used to calculate road travel distances from patients' residence to treatment centres. The analyses will be conducted using a multilevel Cox proportional hazards model with Level 1 comprising individual-level factors (e.g. occupation) and level 2 area-level indicators of remoteness and area socioeconomic disadvantage.
Discussion
This study focuses on the health inequalities for rural and disadvantaged populations that have often been documented but poorly understood, hence limiting our capacity to intervene. This study utilises and develops emerging statistical and spatial technologies that can then be applied to other cancers and health outcomes. The findings of this study will have direct implications for the targeting and resourcing of cancer control programs designed to reduce the burden of colorectal cancer, and for the provision of diagnostic and treatment services.