Executive Summary

A diverse array of tobacco-related datasets has been developed in Australia. The purpose and content of these datasets varies substantially. They cover a broad range of issues including smoking prevalence rates by age and gender, health, wellbeing and lifestyle, morbidity and mortality rates, expenditure and economic activity, monitoring and surveillance, young people and schoolchildren, and population groups such as prisoners and Aboriginal and Torres Strait Islander peoples.

No single tobacco-related dataset addresses all tobacco-related issues and can meet the needs of researchers, practitioners and policy-makers.

The Agency has developed this resource to provide an overview of Australia’s key datasets on the basis of their tobacco-related content, methods and key characteristics. A scoping audit of Australia’s tobacco-related datasets was undertaken using the following selection criteria:

• Direct measures of tobacco-related issues
• Data representativeness (national or specific sub-groups of the Australian population, e.g. young people, Aboriginal and Torres Strait Islander peoples)
• Use of a valid or appropriate methodology
• Longitudinal studies
• Ongoing rather than one-off data collection
• Data accessibility (including cost).

A total of 50 datasets met the inclusion criteria. They were grouped into one of the following four categories:

• national datasets
• jurisdictional datasets
• longitudinal datasets
• special population datasets (e.g. school children, Aboriginal and Torres Strait Islander peoples, and prisoners).Details of each of the 50 identified tobacco-related datasets are provided including information about the data custodian, data availability, purpose of the data collection, methodology, year of most recent collection, and strengths and limitations. Identification of the strengths and limitations of current datasets may inform future tobacco-related data collection and appropriate utilisation.

The datasets have been identified to assist researchers, practitioners and policy-makers to continue to develop the evidence base underpinning preventive health and tobacco control measures in Australia.

Introduction

Tobacco smoking is a leading risk factor for chronic disease including many types of cancer, respiratory disease and heart disease (Ministerial Council on Drug Strategy, 2011). In 2004-2005, the cost of tobacco use to Australian society, including costs to the health and hospital systems, productivity losses and loss of life was estimated at $31.5 billion (Collins and Lapsley, 2008).

The Australian Institute of Health and Welfare (AIHW), in its report Australia’s Health 2010, identified smoking as a leading cause of death and disease and noted that it was responsible for more drug-related hospitalisations and deaths than alcohol and illicit drugs combined (Australian Institute of Health and Welfare, 2010).

The National Tobacco Strategy 2012-2018 and the National Drug Strategy 2010-2015 both emphasise the importance of reducing tobacco-related health, social and economic harms among individuals, families and communities.

All efforts to address tobacco-related problems and monitor the success and effectiveness of preventive and intervention initiatives rely heavily on access to appropriate and high quality data. Such data may be purpose designed or derived from generic datasets primarily established for other purposes. Knowledge about the availability of all potential datasets is therefore crucial. Access to available data also maximises the investment made into the acquisition of such data in the first instance, thereby value-adding to such investments.

This document represents the first comprehensive identification of Australian key tobacco-related datasets. It is noted that a detailed online resource ‘Tobacco in Australia: Facts and Issues’ has been developed by the Cancer Council Victoria (Scollo & Winstanley, 2012) that provides information about a limited number of datasets. However, no comprehensive coverage of tobacco-related datasets has been developed previously.

Identification of available data will also assist in reducing unnecessary duplication of effort and collection of data that is already available through other sources. It will help to maximise the utilisation of available data so that it can be fully interrogated. It will also help to harmonise the way in which different data sets collect data, such that different data sets may achieve a higher degree of comparability into the future. Finally, it will help shine a light on gaps, limitations and deficits and provide a basis for refining future data collection initiatives.

Smoking Prevalence

The importance of establishing a comprehensive information base in regard to tobacco-related datasets, to support prevention and intervention efforts, is underscored by the prevalence of tobacco use in Australia. Data from the 2010 National Drug Strategy Household Survey (NDSHS) found that:

• The proportion of people aged 14 years or older smoking daily (15.1%) had declined, continuing a downward trend that began in 1995. The decline in daily smoking was greatest among those aged in their early-20s to mid-40s, while the proportion of those aged over 45 years who smoked daily remained relatively stable or slightly increased between 2007 and 2010.
• In the 12–17 years age group, girls were more likely to smoke daily than boys (3.2% to 1.8%). This was the only age group where females were more likely than males to smoke daily.
• Support for policies aimed at reducing harm caused by tobacco remained high in 2010. In particular, there were increasing levels of support for a rise in tax on tobacco products to pay for health education and to contribute to treatment costs (Australian Institute of Health and Welfare, 2011).



Tobacco-Related Data

Efforts to prevent and reduce tobacco-related harm, including tobacco control initiatives, are informed by tobacco use patterns including uptake and cessation. This understanding is in turn informed by a range of data covering health and socioeconomic status, demographic, employment, education and sales information. Data on tobacco use is critical to efforts in tobacco control together with socio-demographic and other data, and this data may be found in a single data set (e.g. some national and state surveillance systems) or by linking different complementary data sets.

The World Health Organization (WHO), through the WHO Framework Convention on Tobacco Control (WHO FCTC), requires its Member States to develop, maintain and report data on health outcomes related to tobacco use and exposure (World Health Organization, 2012). The WHO further recommends that tobacco surveillance programs should be integrated into national, regional and global health surveillance programs to ensure that data are comparable and can be analysed at regional and international levels (Scollo & Winstanley, 2012).

In the Australian context, the National Preventative Health Taskforce highlighted the importance of a comprehensive national surveillance system for tobacco for the purposes of collecting and managing relevant datasets, monitoring progress against specified targets and reporting trend information over time (National Preventative Health Taskforce, 2009). The Taskforce also noted that in order for a surveillance system to be effective, it needed to provide data that was representative of the whole population as well as populations of special interest such as, Indigenous people, children and adolescents, and disadvantaged groups (National Preventative Health Taskforce, 2009).

Data Access

Tobacco data are collected for various purposes. They are diverse and scattered across a range of health and social fields, and held by a range of data custodians. Knowledge of available tobacco-related datasets is not necessarily immediately apparent, even to those who may have worked in this field for some time. Gaining a comprehensive understanding of the patterns of tobacco use and associated problems is not likely to be achieved through a single dataset alone regardless of sample or methodological rigour. Hence, knowledge of available tobacco-related datasets and their relative strengths and weaknesses is essential.

Access to most of the datasets detailed in this resource requires the permission of the data custodians. This may include the need to address ethical and privacy-related issues as per the National Health and Medical Research Council’s National Statement on Ethical Conduct in Human Research (National Health and Medical Research Council, 2007). In addition, acquiring some datasets may incur a financial cost.

Opportunities for Data Linkage

Further to the identification of specific datasets that contain relevant and quality tobacco-related data, the reader’s attention is also drawn to the scope that exists for data linkage. Data linkage allows two or more datasets to be examined together to explore possible relationships. For instance, data linkage offers potential to examine patterns of health service utilisation by (de-identified) individuals and/or groups for tobacco-related issues and interrogation of demographic and other variables of interest over time. Without data linkage, we are restricted in our understanding of the patterns of tobacco use and its associated health issues as determined by single sources of data. Retrospective examination of existing data through data linkage processes is a cost effective way of conducting cohort studies and potentially offers large scale samples to enable accurate assessment of the risk of relatively infrequent events. For example, researchers in Western Australia (WA) can access health data from the WA Health and Wellbeing Surveillance System (HWSS), Death and Hospitalisation data sets using data linkage. Data linkage capacity is also increasing in other states.

Australia is comparatively new to systematically linking tobacco-related data and opportunities exist for a better understanding of the aetiology of health and wellbeing behaviours and the potential for further data exploration and utilisation by researchers and policy makers.

Data Gaps

Australia has high quality tobacco-related data that encompasses most key areas of relevance. There are, however, some notable gaps. Some gaps include the lack of data on people experiencing homelessness, people living with a mental illness, and on culturally and linguistically diverse communities.

Australia’s National Drug Strategy 2010-2015 notes that tobacco use is common among homeless people (Ministerial Council on Drug Strategy, 2011). However, apart from occasional research studies there is no regular data collection on smoking prevalence among homeless people. The National Tobacco Strategy 2012-2018 identifies that smoking rates for homeless people are much higher than for the general population and that many of the factors that underlie social disadvantage such as homelessness are also predictive of smoking uptake (Intergovernmental Committee on Drugs, 2012). Lack of basic data about smoking prevalence rates among homeless people represents a major gap in Australia’s tobacco-related data collection and action is required to redress this anomaly.

There are limited data specifically collected about prevalence rates in culturally and linguistically diverse populations. Scollo and Winstanely (2012) note that, in the past, one-off studies have been conducted examining the smoking rates of the Arab-speaking population and the Vietnamese community in Sydney. More recently, the updated results from the Australian Health Survey, released in June 2013, contain data on smoking status by selected population characteristics, including country of birth (Australian Bureau of Statistics, 2013). Given Australia’s cultural diversity there is an imperative to continue to better understand the smoking rates amongst members of Eastern and Southern European, African, Middle Eastern and Asian population sub-groups. This has important implications for the development of appropriate education, prevention and support initiatives.

Summary

Australia holds a substantial array of high quality tobacco-related data. This report provides a detailed compilation and examination of 50 key datasets. The available data can be used to examine the patterns of tobacco use, including uptake and cessation, to determine reliable estimates of the prevalence and impact of tobacco use in Australia, and to inform future directions for tobacco control.

Identification of key tobacco-related datasets provides researchers, practitioners and policy-makers with a useful resource and will contribute to the development of an enhanced evidence base to inform tobacco-related policy and initiatives, inform the development of appropriate and evidence-based tobacco control measures and policy responses aimed at preventing uptake and reducing the rate of smoking in Australia. Equally important, the identification of gaps and limitations in current datasets will also help improve tobacco-related data collection and data utilisation in the future.