Whose risk? Whose responsibility? The Epidemiology of intervening in health risk behaviours for Aboriginal and Torres Strait Islander people

Abstract

This thesis presents a collection of applied epidemiological studies and is primarily focused on health risk behaviours and preventive health programs. The studies included: - a cross-sectional analysis of data from two national Aboriginal and Torres Strait Islanders surveys. The study examined estimates of smoking prevalence by Indigenous Regions, 2012/13 and 2014/15; - a pre- and post- study of a local Aboriginal-developed tobacco control campaign in Central Australia. The study examined change in smokers and ex-smokers knowledge, attitude and behaviours to tobacco use, 2016-17; - a retrospective cohort study of Aboriginal Community Controlled Health Services which examined if outcomes in smoking were associated with targeted Indigenous tobacco control funding, 2014-2017; - a clinical audit to identify and provide recommendations to improve the pathways available to help identified smokers quit, in six Aboriginal Community Controlled Health Services in Queensland 2017; - an evaluation of the Royal Flying Doctors Service Monitoring System for an obesity-prevention program, the Healthy Living program, delivered in remote communities in South Australia 2007-2016; and, - an outbreak investigation of gastroenteritis following Melbourne Cup luncheons in the Canberra, November 2016. In all but one study, I examined data availability, collection and analysis to support program evaluations. Although, Australian national survey data for the Aboriginal and Torres Strait Islander population are the most comprehensive source on health risk behaviours, the data’s utility for evaluating regional program level activities and campaigns is challenged by small sizes and data restrictions. In my study of regional smoking prevalence, I noted that wide confidence intervals meant that changes over time at a regional level could not be detected. Further, cross sectional analysis is restricted to comparing changes in sub-populations, jurisdictions or regions over time rather than an analysis of changes in individuals over time. As such, these analyses do not allow the attribution of exposures, including exploring what program specific activities and campaigns lead to changes.

In response to the national data limitations, managers and implementers of individual programs are opting to develop their own studies, monitoring and evaluation systems. In my pre- and post- study of smokers in Central Australia, I observed improvements in participants’ knowledge of smoking health risks and increased numbers of ex-smokers. I have recommended that next phase of Australian Government’s Tackling Indigenous Smoking evaluation needs to better examine which program specific exposures in tobacco control led to: the de-normalisation of tobacco use in communities; increased quit attempts; reduced uptake of smoking; and, the overall declines in smoking prevalence. I also examined the RFDS Monitoring System, a program monitoring system established to track changes in participants due to community-based health promotion activities. Local surveys/sentinel surveillance and program monitoring can help to improve program reporting and evaluations, is responsive to local data requests and builds health service capabilities. However, the system for ongoing data collection at an individual program level can also be compromised because of lack of comparability across sites/activities, limited budget, inadequate definitions of sampling frames, lack of standard operating procedures for data collection and the onus on nontechnical staff to analyse data. Australia needs an enhanced data collection system for preventive health that supports both local and regional data needs, while remaining comparable at a national level. Show more