Cancer incidence projections (long-term)
This page provides information about the cancer incidence projections for 2024 to 2033. The data are not accompanied by a data visualisation and are not included within this information page but can be found within the Data section of the report.
Within this section, the 2020 to 2023 incidence projections are referred to as short-term projections and the 2024 to 2033 as the long-term projections. The two sets of projections are derived by different methods. Because of this, on occasion the 2023 to 2024 incidence rate series may disconnect to some degree.
Both methods derive projections using historical trends in actual data. To that end, the reliability of projections reduces where new trends are yet to emerge, or haven’t emerged to the extent to fully influence projections.
Population projections from the Centre for Population are used for incidence rates from 2023 to 2033; the estimated resident population from the Australian Bureau of Statistics is used for earlier years. Cancer incidence projections are influenced by the estimated change in cancer rates and the change in the population size and composition. Population change often accounts for a considerable proportion of the increase in cases diagnosed and projected, particularly with the Australian population ageing, and cancer incidence being higher in older age groups. The reliability of projections is dependent on the accuracy of cancer rates projected and populations projected.
Given the uncertainties surrounding projections, they are provided as estimates rather than predictions. Projection time series are primarily provided as estimates of counts and rates for each given year, rather than estimates from which detailed time-series analysis can be undertaken.
The actual cancer incidence rates over time often have a degree of volatility. Accordingly, single years may on occasion vary from the overall trend. Projections do not have such volatility and instead project individual years as part of an overall trend.
Covid-19 may have some impact on the timing of diagnosis for some cancers. A possible impact may be that years where Covid-19 was at its height may result in fewer cases diagnosed for some cancers and a corresponding increase in later years where delayed diagnosis occurs. The CdiA projections do not attempt to adjust for Covid-19.
Short-term projections are derived from linear regressions. More information about the method can be found in the Methods section of CdiA. The short-term projections were not considered as suitable for longer-term as the single linear trend rarely continues for extended periods of time. The longer-term projections are derived from the Nordpred program, a software package developed by Harald Fekjær and Bjørn Møller at the Cancer Registry of Norway for predicting trends in cancer incidence using a modified age-period-cohort model based on existing incidence trends.
Prostate cancer projections are not derived using the short-term or long-term projection methods. Instead, these projections use the most recent actual incidence rates and apply these to the population projections. More information about prostate cancer incidence projections are available in Cancer data commentary number 9 (please note that the data in this commentary have been superseded by the new projections but the paper helps explain difficulties in projecting prostate cancer incidence).