1.1. What is the Australian Cancer Atlas?
The Australian Cancer Atlas is an interactive, online atlas designed to show how the burden of cancer varies across small geographical areas for the whole country, using the latest available data.
The Atlas highlights whether the burden of cancer in certain geographical areas is likely to be below, similar to or above the Australian average. This is reported for 20 different cancer types, including all cancers combined, and uses the Statistical Area level 2 (SA2) geographical classification.
The purpose of the Atlas is to give all Australians a better understanding of how the burden of cancer varies across Australia.
1.2. Where do the data shown on the map come from?
The maps show estimates that are produced using statistical modelling of real data from Australia’s state and territory population-based cancer registries. Since individual geographical areas often have small populations and, therefore, small numbers of people diagnosed with cancer, this modelling is used to provide more stability in the mapped patterns, and protect the confidentiality of people living in each area. The modelling also helps us to quantify the level of uncertainty, or imprecision, in the mapped estimates. These modelled estimates are called ‘smoothed estimates’, because they smooth out the random fluctuations in the original data. Further details about this modelling can be found here.
1.3. Why are the data modelled?
There are two main reasons: data privacy and statistical stability. Data privacy relates to the responsibility of data custodians to protect the identity of the individuals who are included in their data. Statistical stability relates to the inherent random fluctuations of statistics that occur when there are small numbers of records; the smaller the numbers the more they fluctuate, making correct interpretation difficult.
These issues are particularly important when considering data for small geographical areas. Small areas have small populations and correspondingly low numbers of people with cancer. Modelled estimates help us to see the real differences in the underlying cancer rates between areas without compromising privacy and taking into account natural fluctuations.
1.4. How do I interpret the maps?
The colours tell you how an area compares to the Australian average. Areas shown in blue are likely to have an average risk lower than the Australian average, those in yellow are likely to be equal to the Australian average, and those in orange or red are likely to be higher.
There is a series of interpretative tours that provide greater insights into how to interpret these maps.
1.5. What does the blue and red represent?
The colours shown in the Australian Cancer Atlas represent how the modelled estimates for a specific geographical area compare to the Australian average.
An area shaded blue means that the average risk for people living in that area being diagnosed with cancer (or dying from their cancer within 5 years of diagnosis) is likely to be lower than the Australian average.
An area shaded orange or red means that the average risk for people living in that area being diagnosed with cancer (or dying from their cancer within 5 years of diagnosis) is likely to be higher than the Australian average.
1.6. How were the specific cancers chosen?
Since individual geographical areas often have small numbers of people diagnosed with cancer, statistical modelling is used to provide more stability in the mapped patterns and protect the confidentiality of people living in each area.
This modelling starts with the assumption there is no variation in cancer burden, and then sees whether there is sufficient evidence in the observed data to change that assumption.
This is why we need to focus on those common cancers; if we included rarer cancers all we would see is a map showing the burden for those cancers is the same in every area across Australia.
These common cancers were chosen to be consistent with the most commonly diagnosed cancers as reported by the Australian Institute of Health and Welfare. This includes all cancers combined, along with nineteen other common cancer types. Some of these cancers are sex-specific, such as prostate cancer among men and cervical and uterine cancer among women. While men are diagnosed with, and die from breast cancer, the small numbers diagnosed were not sufficient to report in the Atlas. For this reason, the burden of breast cancer is only reported for females in the Atlas.
1.7. What years of data are used?
For the more common cancers, data for cancers diagnosed between 2010 and 2014 are included. For less common cancers, data for those diagnosed during 2005-2014 are included. When considering the survival-based measure of excess deaths [refer to Q.1.9 for more information about excess deaths], this is based on the experience of people diagnosed with cancer since 2001 and alive at some point during 2006-2014.
1.8. How often will the data be updated?
There are no fixed schedules for updates. However, when new data are available from the Australian Cancer Database, it is intended to update the modelled estimates shown in the Australian Cancer Atlas. Typically, these data are released about every twelve months.
In the future, it is planned to include other measures in the Australian Cancer Atlas. These will be from other data sources, so the timeframe for updates of those measures will depend on when new data becomes available.
Please [subscribe] if you would like to be notified of any updates or enhancements to the Australian Cancer Atlas.
1.9. What does the phrase ‘excess deaths’ mean?
Excess deaths is a way of expressing the survival of people diagnosed with cancer. Excess deaths among people with cancer are the number of deaths that occur within five years of diagnosis, above and beyond the number of deaths that would be expected in the general population.
1.10. What ages are included in the data?
Due to the different types of cancer that occur in children, we have restricted the Australian Cancer Atlas to those cancers diagnosed among Australians aged 15 years or older.
The analyses of excess deaths include people aged 15-89 years when diagnosed with cancer.
1.11. Don’t older people have more cancer? How does the map deal with that?
The likelihood of being diagnosed with cancer does increase with age. The statistical modelling used for the Australian Cancer Atlas takes into account age group at diagnosis. If one area has mainly younger people, while another area has mainly older people, this is adjusted for in the statistical modelling. This means that any differences between areas in the Australian Cancer Atlas cannot be explained by differences in age.
1.12. Why are both male and female rates shown?
There is a consistent pattern that, of the cancers common to men and women, men are more likely to be diagnosed with cancer and also die from their cancer than women. In particular, men are more likely to be diagnosed with bowel cancer, melanoma and lung cancer than women. While these are some of the most common cancers, they are also among the most preventable. Unfortunately, men are more likely to smoke, be overweight or obese, not use sun protection and drink excessive alcohol, at least one of which are known to increase the risk of being diagnosed with each of these cancers.
1.13. How do I navigate the Atlas?
1.14. Why is the Australian average used?
We used the Australian average to provide a context for the Statistical Area level 2 (SA2)-specific estimates. For example, if the Atlas only reported the diagnosis rate for bowel cancer in one SA2, then it wouldn’t provide any information about whether that was a high rate or a low rate when considering the national picture. For that reason, it was decided to specifically report on whether the diagnosis rate was higher than the Australian average, or lower.
1.15. How are the data on the map located and calculated?
The information presented in the Australian Cancer Atlas is current at the time of publication and will be updated when new data becomes available.
Each of the state and territory cancer registries assign Statistical Area level 2 (SA2) codes to each cancer diagnosis record based on the information they have about the patient’s usual residential address at diagnosis. While most of this process is now undertaken by geocoding, in which street address information is allocated a latitude and longitude co-ordinate and then mapped to SA2 boundaries, the actual process varies between the different cancer registries.
This process of allocating SA2 information to each record is not exact and relies on accurate address information being provided to the cancer registries. Challenges can include when a patient address is recorded as a post office box rather than a street address. Each registry has an ongoing process of checking and improving this process.
1.16. What key assumptions were made in modelling the atlas data?
In any statistical analysis, assumptions are made. There are various assumptions related to the original data and those related to the statistical models used to analyse that data.
Data-related assumptions include assuming that the cancer registries have collected details for all cancers diagnosed in each geographical area, that they have assigned the geographical area (SA2) correctly, and they have recorded the details of the cancer and the cancer patient correctly.
Some of the statistical assumptions include assuming that the average risk of cancer diagnosis, or the average excess death rate due to cancer, in any one area is likely to be more similar to the risk in its neighbouring areas.
Further details about the assumptions are included throughout the methods.
1.17. Do all maps show the same amount of variation?
No. For some types of cancer there is a large amount of geographical variation across the country, for others it is much less pronounced. For each map, there is an ’evidence of variation’ indicator. This shows whether there is ‘strong’ (3 bars), ‘moderate’ (2 bars), ‘weak’ (1 bar) or ‘no’ (0 bars) evidence for geographical variation.
2.1. What is the cancer burden in my area and what does it mean to be higher or lower than the Australian average?
This version of the Australian Cancer Atlas reports ‘cancer burden’ using cancer diagnoses and excess deaths due to cancer within 5 years of diagnosis (based on cancer survival).
All the estimates displayed in the Australian Cancer Atlas are compared to the Australian average. So if a specific geographical area has a ‘high’ rate of being diagnosed with cancer, it means that the average risk for all people living in this area being diagnosed with cancer is likely to be higher than the Australian average.
2.2. What does this mean for me specifically?
All the estimates for cancer diagnosis presented in the Australian Cancer Atlas are about the average risk and average cancer burden for all people living within a specific geographical area; this does not reflect the risk for every individual living in that area.
Where you live plays only a small role in your cancer risk as there are many other factors that need to be considered when determining your specific risk of being diagnosed with cancer or dying from this cancer, including age, family history of cancer, smoking, alcohol consumption, diet, physical activity and sun exposure.
2.3.The diagnosis rate in my area is above the Australian average – does this mean I have a high risk of cancer and should move?
Living in an area where the risk of cancer is above the Australian average does not mean you need to move. Cancer is not contagious. Even on a global scale, there are very few examples when high rates of cancer have been caused by some external factor specific to a geographical area.
It is much more likely that the higher risk of cancer in a specific geographical area reflects the general health and lifestyles of people living in that area. This does not necessarily reflect on your own individual risk. To ensure your own cancer risk is reduced, eat healthily (fruit and vegetables), stay active, maintain a healthy body weight, reduce alcohol intake, quit smoking, stay SunSmart and take part in all eligible cancer screening programs.
2.4. What causes some areas to have a higher or lower cancer diagnosis rates than the Australian average?
It is thought that most of the variation between areas is due to the characteristics and behaviours of the people living within each area. This may relate to known risk factors for cancer, such as smoking, excessive alcohol consumption, poor diet, inadequate physical activity and sun protection. It may also relate to the use of cancer diagnostic procedures, such as having access to skin checks for melanoma, or mammograms for breast cancer.
Tours for specific types of cancer give more details on reasons for these observed patterns.
2.5. What is a cancer cluster and how can one be identified?
A common misconception when identifying that an area, community, or suburb, has a higher rate of cancer than other areas is that there is a ‘cancer cluster’. As you can see from the Atlas, cancer rates vary a great deal from place to place. A true ’cancer cluster’ is rare and is said to occur when the number of people in a geographical area (or workplace, or some other group) who are diagnosed with a particular type of cancer is significantly higher than we would expect, given the actual background cancer rates.
Even though the average risk of being diagnosed with cancer (or dying from the cancer within five years of diagnosis) might be higher for people living in certain geographical areas, this is likely to be the result of the characteristics and behaviours of those people rather than exposure to something in these areas.
It is known that a third of cancers could be prevented by improving behaviours such as smoking, excessive alcohol consumption, poor diet, inadequate physical activity and sun protection. If a greater proportion of people within one community have these poor health behaviours, then that might be one explanation for the higher than average risk in that area.
If a true cancer cluster is still suspected after considering these types of factors, there are established procedures within each State and Territory Health Department to investigate these concerns.
There are extremely few cancer clusters around the world that have actually been confirmed after thorough investigations.
2.6. Where can I get more information about my risk of cancer?
Talk to your general practitioner about your personal risk of cancer.
2.7. Is cancer contagious?
Cancer is not contagious. Some cancer genes are inherited which can result in family members having the same or related cancers, however, this is due to the inherited genetic disorder, not through ‘catching’ cancer.
2.8. Who can I talk to about trying to reduce cancer diagnosis rates in my area?
At least one third of cancer cases can be prevented by avoiding known risk factors such as smoking, excessive alcohol consumption, poor diet, inadequate physical activity, and sun exposure. Any improvements in those risk factors are likely to result in long term reductions in cancer rates.
Talk to your general practitioner about programs that are already in place.
2.9. Who can I talk to about trying to reduce excess death rates due to cancer in my area?
The most important factor impacting on how long people live for after a diagnosis of cancer is how much the cancer has spread in the body when diagnosed. Early detection is vital. Talk to your general practitioner straight away if you notice any unusual changes on or in your body – no matter your age or the time since your last check-up.
In addition, treatments for specific types of cancer are continually being improved and made more effective. Your general practitioner can provide more information about treatments available for people diagnosed with cancer.
2.10. The excess death rate due to cancer in my area is above the Australian average (corresponding to lower survival). Does this mean that if I am diagnosed with cancer I should move?
On average, there is a general pattern that cancer survival outcomes are better for people living closer to cancer treatment centres than those living further away.
People living in more regional and remote areas are often faced with decisions about juggling family and work responsibilities with the treatment needs for their cancer.
Cancer Councils across the country provide individual assistance to those people who need to travel to access cancer treatment services. These include accommodation support lodges and transport services. There may also be government-funded financial assistance available in your state or territory to support the cost of travelling and accommodation.
2.11. What should I do if I think I might have cancer?
See your general practitioner immediately. If you have noticed any changes to your body, have signs and symptoms of cancer, or are concerned about your health, it’s important that you don’t delay visiting your GP.
Early detection plays a key role in successfully treating a cancer diagnosis so it’s always best to get anything of concern checked straight away.
2.12. Where can I get more detailed information about cancer?
General practitioners in your local area
Cancer Council’s information and support line 13 11 20
Cancer Council Australia (https://www.cancer.org.au/ )
Cancer Australia (https://canceraustralia.gov.au/ )
3.1. When were the data captured?
The data extract used for this release of the Australian Cancer Atlas was provided by the Australian Institute of Health and Welfare in March 2018. This extract covered the time period of 2001 to 2014.
For the more common cancers, data for cancers diagnosed between 2010 and 2014 are included. For less common cancers, data for those diagnosed during 2005-2014 are included in the Atlas. When considering the survival-based measure of excess deaths [refer to Q.1.9 for more information about excess deaths], this is based on the experience of people diagnosed with cancer since 2001 and alive at some point during 2006-2014.
3.2. What modelling method was used and why?
For the Australian Cancer Atlas, we are using Bayesian statistical models. The basic idea of the Bayesian method is that it uses a range of plausible values to describe the unknown quantities of interest. For the Atlas, these quantities are the numbers of cancer diagnoses rates and excess deaths rates within five years (related to survival) in each small area. Rather than calculating one specific estimate for each area, the Bayesian statistical models generate a probability distribution using the observed counts, population data and other (prior) information.
The assumption is that the factors that influence the cancer burden in one geographical area are more similar to those in areas closer to them, than to areas further away. In this way, the Bayesian analysis enables us to supplement the data observed in each area, which leads to the more stable and robust ‘spatially smoothed estimates’. The probability distributions that are obtained for the cancer diagnosis and excess death rates in each small area reflect the uncertainty of these estimates and enables more appropriate comparisons to be made between areas or with the Australian average.
3.3. What is the unit of analysis?
All the estimates shown are based on the Statistical Area level 2 (SA2) geographic classification used by the Australian Bureau of Statistics and reported in comparison to the Australian average.
3.4. How many people live within each SA2?
Population estimates for 2,148 SA2s (Statistical Area level 2) across Australia are included in the Australian Cancer Atlas. The average number of people living within each SA2 in 2014 was 10,941 (median of 9,210) and ranged from 4 to 54,773.
3.5. What is meant by ‘uncertainty’?
As is always the case with statistical estimates, including those in the Australian Cancer Atlas, every estimate has a degree of uncertainty, or imprecision, around it. The larger the uncertainty associated with an estimate, the less convincing it is that this is close to the true value.
Even though the cancer statistics are based on the total population, rather than a sample, there is still ‘uncertainty’, because these counts vary from year to year, or from area to area. The smaller the number, the larger this inherent uncertainty is.
In the Australian Cancer Atlas, the level of uncertainty is shown in three ways, the V-plot, the wave plot and the level of transparency.
3.6. What does the ’high probability of being above/below the Australian average’ mean?
The modelled estimates in the Australian Cancer Atlas are reported in relation to the Australian average. The Australian average is represented by the value of one. That means, if one area has a SIR (Standard Incidence Ratio) of 1.3, then that means that the average rate of that cancer being diagnosed in that area is 1.3 times higher than the Australian average (or 30% higher).
Similarly, if an area has an SIR estimate of 0.8, then that means that the average rate of that cancer being diagnosed in that area is 0.8 times the Australia average, or 20% lower.
However, to understand the cancer burden in a particular area you need more than just one statistic. There are many factors that contribute to how many people are diagnosed with cancer, and how long they survive – some factors are known, some are not. However, they all impact on when a cancer develops in an individual, whether that cancer is detected and diagnosed, and whether that cancer progresses and leads to death. For these reasons, the observed cancer statistics will vary from year to year, or area to area, leading to some ‘fuzziness’ around the true value. Generally, this uncertainty is higher when the population, or the numbers of cancer cases are low.
3.7. What is spatial smoothing?
While standard methods for reporting disease burden typically only adjust for age and sex in each area, spatial smoothing also incorporates the geographical structure of the data. It does this by ‘borrowing’ data from the neighbouring geographical areas.
This provides greater stability for the estimates and greatly reduces any risk of individuals being identified.
These protective effects of spatial smoothing are most pronounced for areas where it is needed the most, that is, those with the smallest numbers of cases. Smoothed estimates are designed to reflect the real differences in the underlying rate or risk between areas. See spatial smoothing tour.
3.8. Why have multiple years been used to produce a single estimate?
When looking at over 2,148 small geographical areas across Australia, there are not enough numbers of cancers diagnosed in each area each year to generate plausible estimates. Aggregating multiple years of data provides greater stability to the counts and enables the statistical models to run effectively.
3.9. Why is there no time series?
Due to the complexities of the statistical models, we have focused on the most recent time period for this first version of the Australian Cancer Atlas. In a subsequent release, we do intend to look at how the geographical patterns have changed over time. This may need to be restricted to the more common cancers.
3.10. Are there important correlations that can be observed in the data?
It is likely that there are specific factors that are driving the higher, or lower, rates for specific areas across Australia. Many of these are known. For example, we know that smoking causes most cases of lung cancer. This means that historical smoking rates in an area would likely impact on the average risk of lung cancer in that area.
However, there are still many causes of cancer that are not understood. It is hoped that by describing the geographical patterns of cancer across Australia that the Australian Cancer Atlas will motivate efforts to better understand what the driving forces behind that variation are. Ultimately, the goal is to use this knowledge to reduce the level of geographical variation in the future.
3.11. Will the uncertainty measures corresponding to each estimate be available?
Yes, in the Australian Cancer Atlas, the level of uncertainty is shown in three ways, the V-plot, the wave plot and the level of transparency.
3.12. Can estimates for multiple SA2s (e.g. a state) be combined to provide an overall estimate?
Due to the statistical modelling process, it is not possible to combine estimates from multiple SA2s (Statistical Area level 2) into one value. However, it is possible to look at the distribution of estimates within a broader area such as by State or Territory, by remoteness areas, or by area disadvantage.
4.1. Why is there so much liver cancer in the Northern Territory?
The primary cause of liver cancer is liver damage by excess alcohol consumption and chronic infection with diseases such as hepatitis B and C. Compared with other states and territories, the Northern Territory has a higher proportion of the population who have higher rates of harmful alcohol consumption and higher prevalence of chronic hepatitis B infection, which may be contributing factors.
4.2. There is a high level of lung cancer in my area. Am I at risk of developing lung cancer?
We know that smoking causes around 70-80 per cent of lung cancers. If an area has a higher rate of lung cancer diagnoses than the Australian average, then it is very possible that the prevalence of smoking in that area was also higher in the past (up to 20-30 years earlier). If this is the case, there are two considerations here. First, exposure to second hand smoke is known to cause cancer, so minimising your exposure to second hand smoke is important. Second, smoking can be more socially acceptable in certain communities, which causes an increased risk of tobacco-related cancers.
4.3. What are the driving factors behind the diagnosis patterns for melanoma?
The highest risk of being diagnosed with melanoma is predominantly in south-east Queensland and northern New South Wales. These regions are renowned for their beach-oriented outdoor lifestyle and relatively high UV levels. In addition to skin checks being conducted by general practitioners, there are a large number of primary care skin cancer clinics in these areas. Whole body skin examinations have been shown to reduce the diagnosis rate of thick melanomas, but can increase the diagnosis rate of thin melanomas by detecting them earlier than they might otherwise have been detected.
Most of the areas in Northern and Western Australia have a high proportion of Indigenous Australians, who are known to have a very low risk of being diagnosed with melanoma, so this is one explanation for the lower risk of being diagnosed with melanoma in these regions.
5.1. I am a journalist; how can I use the Atlas to find tours in the data related to my field?
The interpretative tours found here provide explanations about the interpretation of key findings from the Atlas.
5.2. What policy changes should be developed as a result of the findings?
The intent of the Australian Cancer Atlas is to stimulate further discussions, motivate targeted research designed to better understand reasons why the geographical variation exists. Once those reasons are better known, it is hoped that suitable interventions, including policy changes, would be implemented to reduce the current inequalities.
5.3. Can I download the modelled estimates?
Yes, the modelled estimates for each Statistical Area level 2 (SA2) that are shown in the maps can be downloaded under ‘Options’ in the atlas.
5.4. Can I generate a paper copy of the results?
Yes, there is an option (under ‘Options’) to generate a pdf document that contains the image currently on your screen, along with an Australian map for that cancer, and other citation and background information for that screenshot.
5.5. How can I link my data to your data (download)?
The mapped data can be downloaded under ‘Options’ in the atlas. These data are provided by Statistical Area level 2 (SA2). If you have other data by SA2, then there are no restrictions for combining your data with the data you download.
There is no capacity to link data at the unit record level.
5.6. How do I cite the Atlas?
Australian Cancer Atlas (https://atlas.cancer.org.au). Cancer Council Queensland, Queensland University of Technology, Cooperative Research Centre for Spatial Information. Version 09-2018. Accessed [give date]
5.7. What are the Copyright considerations for this Atlas?
Copyright notice: This material within the Australian Cancer Atlas is copyright Spatial Information Systems Research Limited, 2018. Spatial Information Systems Research Limited (formerly trading as the Cooperative Research Centre for Spatial Information and trading as FrontierSI from 1 July 2018), has no objection to the reproduction of the material within the Australian Cancer Atlas provided the following criteria apply:
- Content is used in a non-commercial setting.
- The required citation (see below) must be prominently displayed.
- No fees are charged by you for your presentation of this content.
- Content is not altered.
- Content is not to be reproduced for use on a Web page or with any online service or application without including the required citation.
- Other than as set out above, the content from the Australian Cancer Atlas may not be reproduced, adapted, distributed, stored in a retrieval system or transmitted without prior written consent from Spatial Information Systems Research Limited, unless expressly permitted by law.
Required citation: Australian Cancer Atlas (https://atlas.cancer.org.au). Cancer Council Queensland, Queensland University of Technology, Cooperative Research Centre for Spatial Information. Version 09-2018. Accessed [give date]
5.8. What are the legal disclaimers for this Atlas?
Disclaimer: Spatial Information Systems Research Limited, Cancer Council Queensland and Queensland University of Technology (the “Project Parties”) do not warrant that the information in the Australian Cancer Atlas is correct or complete or that it is suitable for any particular purpose. Your use of the information is at your own risk. To the fullest extent permitted by law, the Project Parties do not accept any liability for any reliance placed on the information including information which is not correct, complete or up to date or suitable for any particular purpose. If any warranty or guarantee cannot by law be excluded, then, to the extent permitted by law, the Project Parties’ liability for such warranty or guarantee is limited, at Spatial Information Systems Research Limited’s option, to supplying the corrected information or materials again via an update to the Australian Cancer Atlas.