Health Information from the Australian Bureau of Statistics Presentation for the National Centre for Geographic and Resource Analysis in Primary Health Care Health Information from the Australian Bureau of Statistics Louise Gates Director Health Section

The Australian Health Survey is the largest most comprehensive health survey ever conducted in Australia. The survey is funded by the ABS work program, the Commonwealth Department of Health and Ageing and the National Heart Foundation of Australia

What are some of the questions? Are Australians healthy? What kind of health conditions do we have? How accessible are health services? What are we eating and drinking? How many of us smoke? Are we getting enough nutrients in our diet? How do our physical activity patterns play a role in our health? Are there differences between population groups? This slide lists some of the questions that policy makers and researchers are interested in answering now and into the future. In the last few years there has been an increasing focus on preventive health, both internationally and in Australia. In particular, the focus has been on chronic disease and lifestyle-related risk factors and on developing a strong evidence base for policy and program decision-making in these areas.

Structure of the Australian Health Survey General population sample size = 26,000 households NATIONAL HEALTH SURVEY (NHS) 16,000 Households 1 Adult + 1 child = 21,120 persons   Detailed conditions Medications and supplements Health related actions Days of reduced activity Social & emotional wellbeing (18+) Physical activity (15+) Private health insurance status (18+) Breastfeeding (0-3) Disability status Alcohol consumption (15+) Family stressors (15+) Personal income (15+) Financial stress   NATIONAL NUTRITION AND PHYSICAL ACTIVITY SURVEY (NNPAS) 10,000 Households 1 Adult + 1 child (2+) = 12,890 persons Food security Food avoidance Dietary recall Physical activity Selected conditions CORE CONTENT Household information Demographics Self-assessed health status (15+) Self-assessed body mass (15+) Smoking (15+) Physical measures (2+) Fruit & veg consumption (2+) Salt use (2+) Blood pressure (5+) Female life stage (10+)   NNPAS Telephone follow-up 2nd dietary recall 8-day pedometer (5+) This diagram shows the main topics of information collected in the general population. It shows where the topic is collected for a subsample of the population i.e. age shown in brackets () Of the 26,000 households selected in the survey, 16,000 will be selected for the NHS and 10,000 for the NNPAS 1 adult and 1 child (if any) are randomly selected from all the household residents. Only children 2 years and over are in scope for the NNPAS A personal interview is conducted with the assistance of a computer based instrument. Physical measurements are taken if the respondent agrees. A core set of questions are common across both surveys (shown in the blue circle in the centre). All respondents 5 years and over selected in either the NHS or NNPAS are invited to participate in the voluntary biomedical component. Only those respondents 12 years and over are asked to provide a blood sample. NATIONAL HEALTH MEASURES SURVEY (NHMS) All survey participants (aged 5 yrs +) invited to VOLUNTEER Key blood tests (12yrs +) and urine tests (5yrs +) of nutritional status and chronic disease markers

Sample size = 7,700 households Proposed Structure of the Australian Aboriginal and Torres Strait Islander Health Survey Sample size = 7,700 households 2 Adults + 2 children (non-remote) - 6372 persons 1 Adult + 1 child (remote) - 3878 persons NATIONAL ABORIGINAL AND TORRES STRAIT ISLANDER HEALTH SURVEY (NATSIHS)   Detailed conditions Health related actions Days of reduced activity Social & emotional wellbeing (18+) Physical activity (15+) Private health insurance status* (15+) Breastfeeding (0-3) Disability, injuries & short term conditions Alcohol & substance use (18+) Adult immunisation (15+) Women’s health (e.g. pap smears) (18+) Female contraception (18-49) Male contraception (18+) Family stressors (15+) Personal income (15+) Financial stress Cultural identification & discrimination (18+) 1 Adult + 1 child (2+) Non-remote - 1250 persons Remote - 1750 persons NATIONAL ABORIGINAL AND TORRES STRAIT ISLANDER NUTRITION AND PHYSICAL ACTIVITY SURVEY (NATSINPAS)   Dietary recall Physical activity Selected conditions CORE CONTENT Household information Demographics Self-assessed health status (15+) Self-assessed body mass (15+) Fruit & veg consumption (2+) Salt use (2+) Food Avoidance (2+) Food Security Smoking (15+) Physical Measures (2+) Blood pressure (5+) NATSINPAS Telephone follow-up* (Non-Remote) 2nd dietary recall 8-day pedometer (5+)   The three Australian Health Survey components that collect information specifically from the Aboriginal and Torres Strait Islander population are sometimes referred to as the Australian Aboriginal and Torres Strait Islander Health Survey. The majority of the topics are the same as those collected for the general population with some variations based on data priorities and with some reduction in content for remote areas (in order to ensure that high quality data is collected). *Excludes remote areas Population is all persons unless otherwise indicated. NATIONAL ABORIGINAL AND TORRES STRAIT ISLANDER HEALTH MEASURES SURVEY (NATSIHMS)  All adult survey participants invited to VOLUNTEER Key blood and urine tests of nutritional status and chronic disease markers (18+)

Release schedule – 15 October Australian Health Survey: First Results, 2011-12 (cat. no. 4364.0.55.001) focus on risk factors and long-term health conditions Australian Health Survey: Users’ Guide, 2011-13 (cat. no. 4363.0.55.001) Profiles of Health, Australia, 2011-13 (cat. no. 4338.0)

Release schedule – 2013 Content Date Description Health service usage February 2013 Focus on health services and actions taken, including medications Conditions and risk factors – revised release May 2013 Revised indicators from earlier releases plus additional material and depth of analysis, based on full AHS sample Physical activity release June 2013 Focus on pedometer data and physical activity indicators from NNPAS Biomedical measurements Focus on high level results from the biomedical measures Nutrition September 2013 Selected highlights from Nutrition component of NNPAS Basic CURF releases and data access tools Second half of 2013 Multiple CURFs will be released that combine the common elements of the AHS

Release schedule – 2014 Content Date Description National Aboriginal and Torres Strait Islander Health Survey (NATSIHS) September 2013 Focus on health risk factors, for COAG reporting National Aboriginal and Torres Strait Islander Nutrition and Physical Activity Survey (NATSINPAS) June 2014 Aboriginal and Torres Strait Islander biomedical measurements 2nd half of 2014

New data products Survey Table Builder ABS.Stat build customised tables through drag-and-drop interface available 2013 ABS.Stat aggregate data in flexible and dynamic way machine-to-machine data exchange http://stat.abs.gov.au/#

National Health Survey 2007-08: 20,800 people (data since 1977) Personal interview One adult (18+) and one child (where applicable) for each dwelling prevalence of long term health conditions (cardio-vascular, cancer, arthritis, asthma, diabetes, mental health etc.) health risk factors and lifestyle (smoking, alcohol, BMI, fruit and veg, physical activity) use of health services (GP, specialist, allied health) use of medications medications for specific health conditions NHS 20,800 people One adult (18+) and one child (where applicable) for each sample dwelling Provides benchmark measures on a wide range of health related issues. Also designed to monitor changes in health over time Risk factors: alcohol and smoking 15-17 yrs nutrition in children – fruit, veg and milk consumption for 5+ physical activity at home and work measured and self reported measurements of height, weight, hip and waist

Patient Experience Survey (PEx) 2010-11: 27,000 people (data since 2009) computer assisted telephone interview general practitioners medical specialists dental professionals imaging and pathology after hours care hospital/emergency departments care experience information provision Waiting times Barriers to access PEX Measures access and barriers to a range of health care services People 15+ years Information on children’s use of health services was also collected from households with children under 15 Conducted annually and collects data on barriers and access to the following information: GP, medical specialists, dental professionals, imaging and pathology tests, after hours care, hospital/emergency departments. Also collects data on patient experiences of care and information provision. Includes data from people that did not access health services as well as those who did which enables analysis of health service information in relation to particular population groups.

National Aboriginal and Torres Strait Islander Health Survey 2004-05: 10,400 people Personal interview 6 yearly – remote areas and discrete Indigenous communities prevalence of long term health conditions (cardio-vascular, cancer, arthritis, asthma, diabetes, mental health etc.) health risk factors and lifestyle (smoking, alcohol, BMI, fruit and veg, physical activity) use of health services (GP, specialist, allied health) use of medications medications for specific health conditions NHS 20,800 people One adult (18+) and one child (where applicable) for each sample dwelling Provides benchmark measures on a wide range of health related issues. Also designed to monitor changes in health over time Risk factors: alcohol and smoking 15-17 yrs nutrition in children – fruit, veg and milk consumption for 5+ physical activity at home and work measured and self reported measurements of height, weight, hip and waist

Survey of Disability, Ageing and Carers Collect information about people with a disability older people (i.e. those aged 60 years and over) people who provide assistance to older people and people with disabilities. Disability = any limitation, restriction or impairment which restricts daily activities that has lasted or is likely to last at least 6 months e.g. Range from loss of sight not corrected by glasses, arthritis that causes difficulty dressing and dementia that requires constant help and supervision

Survey of Mental Health and Wellbeing (SMHWB) 2007 - Measures the nature and severity of mental illness in Australia selected lifetime and 12 month mental disorders 3 major groups – anxiety disorders, affective disorders and substance use disorders levels of impairment health services used for mental health problems physical conditions social networks includes people who don’t access health services SMHWB Mental health is one of 8 national health priority areas identified by WHO in its global strategy to improve health outcomes for all people. Others Asthma Arthritis Cancer cardiovascular health diabetes Obesity injury prevention 8,800 people 16-85 yrs CIDI - Composite International Diagnostic Interview 3 disorder groups: Anxiety disorders (eg social phobia) Affective disorders (eg depression) Substance abuse disorders (eg harmful use of alcohol and drugs) Did not use proxies Response rate was lower than expected (60%) – users should exercise caution

Providing information at small area levels (Louise – there wasn’t much in the way of analytical papers on PEx, so I thought that we could talk about the future directions of the survey here, in relation to the plans for expanding it etc…)

Small area estimation Governments are increasingly interested in the characteristics of small areas. Small area estimation is one way of fulfilling these requests. Uses the known characteristics of a region to model predicted values of a particular variable. Best for variables that are closely correlated with demographic variables and have relatively high prevalence. Often, governments want detailed data on particular geographical regions that we are unable to provide directly using numbers collected from our survey sample. In this case, we can produce small area estimates for particular regions as a customised request. There are several methods for producing small area estimates, however, I won’t go into detail about them in this presentation. Essentially, small area estimation uses the known characteristics of a particular geographic region to model the predicted values for particular variables. Generally, known characteristics of small areas are their demographic characteristics, such the age-sex profile of the area. This information can be easily attained from Census counts. Given this, synthetic estimates work best for data which is correlated with demographic variables such as age and sex. This technique can be useful in the case of health and disability data as many health conditions are age-related. However, small area estimation typically only works well on variables that are relatively prevalent in the population, such as overweight/obesity. If there is a small prevalence of a particular condition (as there is for many, but as an example, diabetes), there may already be large RSE’s associated with the estimates at the national or state/territory level. As such, using small area estimation is not appropriate for these data, as it is in essence, adding error on top of error. So despite the fact that diabetes is an age-related condition, it is not appropriate to use small area estimation for this variable due to its low prevalence.

Small area estimation Often, governments want detailed data on particular geographical regions that we are unable to provide directly using numbers collected from our survey sample. In this case, we can produce small area estimates for particular regions as a customised request. There are several methods for producing small area estimates, however, I won’t go into detail about them in this presentation. Essentially, small area estimation uses the known characteristics of a particular geographic region to model the predicted values for particular variables. Generally, known characteristics of small areas are their demographic characteristics, such the age-sex profile of the area. This information can be easily attained from Census counts. Given this, synthetic estimates work best for data which is correlated with demographic variables such as age and sex. This technique can be useful in the case of health and disability data as many health conditions are age-related. However, small area estimation typically only works well on variables that are relatively prevalent in the population, such as overweight/obesity. If there is a small prevalence of a particular condition (as there is for many, but as an example, diabetes), there may already be large RSE’s associated with the estimates at the national or state/territory level. As such, using small area estimation is not appropriate for these data, as it is in essence, adding error on top of error. So despite the fact that diabetes is an age-related condition, it is not appropriate to use small area estimation for this variable due to its low prevalence.

Contact details for Health: Louise Gates Director louise.gates@abs.gov.au 02 6252 6415 Sally Rayner Assistant Director sally.rayner@abs.gov.au 02 6252 6014