Number of diagnoses
The total number of diagnoses for each cancer type by sex (males, females, persons) aged 15 years and over during 2010-2014 were divided by five to give an average number per year. These five-year averages are shown for each cancer type to enable comparisons across the cancer types [even though the modelled estimates shown in some maps represent ten years of data (2005-2014)].
Diagnosis rate (age-standardised)
Age-specific rates in 5-year age bands, starting at 15 years (i.e. 15-19, 20-24,…85+) were calculated for each cancer type and sex (males, females, persons) for Australia during the years 2010 to 2014. These age-specific rates equal the number of cases within a cancer type, sex and age group combination divided by the corresponding population.
Age-standardised rates are then calculated by multiplying the age-specific rates by standard population weights, which for the Atlas was the Australian Standard Population 2001, and then adding together.
Excess deaths (survival)
Number of excess deaths
The number of excess deaths within 5 years provide an estimate of the deaths caused by a specific cancer that occur within 5 years of people being diagnosed with that cancer.
The excess refers to the difference between the number of deaths among cancer patients and the number of deaths expected if each cancer patient had the same mortality as their counterparts in age, sex and year.
Excess deaths is calculated using relative survival methods. Relative survival is the most commonly presented measure of cancer survival when using data from population-based cancer registries [Dickman et al. 2004]. One key advantage is that no knowledge of the specific cause of death is needed.
Patients who were still alive at 31st December 2014, or who were alive more than five years after diagnosis, were considered ‘censored’. Even though they might have died after this date, they are considered to be alive for the purposes of these calculations.
Survival calculations only included people aged between 15 and 89 years at diagnosis. Patients whose cancer diagnosis was based on death certificate or autopsy only have also been excluded, as well as those with a survival time of zero days or a date of diagnosis after date of death.
Relative survival estimates can be calculated using either the period or cohort methods [Brenner 2002], and we used the period approach. Rather than following a cohort diagnosed within a certain time period, this examines anyone alive within 5-years of diagnosis within a certain time period.
The life table method was used to calculate observed survival. This approach involves dividing the total period being studied into a series of discrete time intervals. Survival probabilities were calculated for each of these intervals, and then multiplied together to produce the observed survival estimate. Expected survival (based on total Australian mortality data published by the Australian Bureau of Statistics [ABS Cat. No. 3302.0.] was calculated for each age (15,16,17,…,100+ years) and individual calendar year (2010, 2011,…,2014) based on the Ederer II method [Ederer et al. 1961].
Any follow-up time intervals extending past 5 years were excluded, and the number of observed deaths minus the number of expected deaths was summed and divided by five (given the 5-year ‘at-risk’ time period of 2010-2014) to provide an average per year.
Excess death rate
The 5-year excess death rate is calculated as 100 minus percent relative survival (see Interpretation). Here relative survival was calculated as above, but also internally age-standardised. This involved calculating the proportion of cases diagnosed among broad age groups (15-54, 55-64, 65-74, 75-89 years) and weighting the relative survival estimates for each of these age groups by this proportion.
Level of evidence for spatial variation
This summary provides details of the level of evidence for spatial variation in the estimates across Australia. The evidence is split into four categories, denoting strong evidence for variation (3 bars), moderate (2 bars), weak (1 bar) and no evidence for spatial variation (no bars). Weak or none indicates a lack of variation in the cancer patterns overall. However, there may still be some areas that are likely to be above/below the Australian average. Further details about the method used to calculate the summary measure can be found here
Australian Bureau of Statistics. Deaths, Year of registration, Age at death, Age-specific death rates, Sex, States, Territories and Australia
ABS Cat. No. 3302.0., retrieved on the 18th May 2018. ABS: Canberra, 2018
Brenner H. Long-term survival rates of cancer patients achieved by the end of the 20th century: a period analysis. Lancet. 2002; 360(9340):1131-1135.
Dickman PW, Sloggett A, Hills M, Hakulinen T. Regression models for relative survival. Stat Med. 2004; 23(1):51-64.
Ederer F, Axtell LM, Cutler SJ. The relative survival rate: a statistical methodology. NCI Monogr. 1961; 6:101-121.