Survival statistics

This page contains all the information you need to know about the survival portal. From the portal, you can access survival data by various different dimensions. Charts and tables will be provided for this data. You can also search all survival data and export tabular results.

Use the selection panel in the top section of each tab to select how you would like to search for estimates of relative survival for Irish cancer patients (malignant cancers only based on ICD-O-3 rules) by years survived. Figures are shown for the most common cancers, based on the first invasive/malignant cancer diagnosed in c300,000 adult patients (age 15-99) during 1994-2013 (excluding non-melanoma skin cancer which is rarely fatal). [For childhood cancer survival, see this report].

About our survival data

Net survival represents the probability of a patient surviving a given time in the  hypothetical situation in which the disease of interest is the only possible cause of death.  It is essentially an estimate of survival from the disease of interest having controlled for other possible causes of death (by comparison of observed survival with the expected survival of persons of the same age and sex in the general population). For cancer patients, it measures the effect of the excess mortality associated with a cancer diagnosis, providing an indirect alternative to estimation of cause-specific survival. Most commonly, five-year net survival estimates are presented. For example, if average five-year net survival for patients with a particular cancer type is 80%, on average 20 out of 100 patients die within five years who would not otherwise have died, based on our knowledge of 'background' mortality rates among populations of the same age and sex.

Survival estimates for Irish patients were derived by matching of registered cancer cases against death certificates for all deaths in Ireland up the end of 2014 and comparison against life tables showing expected probabilities of survival for the general population. Calculations use the ‘strs’ command with ‘Pohar’ option in Stata and estimates have been constrained so that net survival does not exceed 100% in any follow-up interval. Please interpret survival figures with caution - confidence intervals (a measure of statistical uncertainty) are wide on some estimates, and figures shown may be averages across patients of different age, disease stage or other prognostic factors. Figures updated January 2018.

Survival by year

Survival is presented by years of diagnosis (traditional ‘cohort’ estimates for 1994-1999, 1999-2003, 2004-2008. 2009-2013) and, in the case of the most recent years, by period of follow-up (‘hybrid’ estimates for 2010-2014).

Survival by sex

Survival statistics for males and females are provided for all years, but age-standardized figures are only available for 2010-2014 at this time.

Survival by region

Survival is presented by area of residence at the time of diagnosis, for the period 2010-2014.

Stage-specific survival

Survival is presented by AJCC/UICC 5^th edition TNM stage, for the period 2010-2014.

Age-specific survival

Survival is presented for the standard age-groups (usually 15-44, 45-54, 55-64, 65-74 and 75-99) used by EUROCARE and other projects for the purpose of age-standardization for international comparison of survival.

Search all survival

Results will be provided for all combinations of variables currently available in the underlying data (but some requested options may not be available).

Further notes

Reference for net survival:

Pohar Perme M, Henderson R, Stare J. 2009. An approach to estimation in relative survival regression. Biostatistics 10: 136-146.

‘Hybrid’ estimate for 2010-2014: 

This is the most up-to-date survival estimate and is based on follow-up of all cases alive at any point during this period, supplemented by follow-up 2009-2014 for cases diagnosed in 2009 (to provide more reliable estimation of the first year of survival). In practice, the hybrid estimate of 5-year survival for 2010-2014 is similar to the most recent (incomplete) cohort, i.e. cases diagnosed 2009-2013, reflecting substantial overlap of data. The real advantage of the hybrid approach is in allowing more up-to-date estimation of longer-term survival, presented up to 10 years here.

Age-standardisation:

Where noted, survival estimates are standardised to the International Cancer Survival Standard (ICSS) populations proposed for each cancer by:

Corazziari I., Quinn M. & Capocaccia R. 2004. Standard cancer patient population for age standardising survival ratios. Eur J Cancer 40: 2307-2316.

Exclusions:

Patients aged <15 or >99 at diagnosis; death-certificate-only (DCO) and autopsy-only cases; second or subsequent malignancies in the same patient (or the less serious of two or more synchronously-diagnosed malignancies); in situ carcinomas, benign tumours and tumours of uncertain behaviour. (However, in situ and uncertain-behaviour tumours are included in alternative survival estimates for bladder cancer, to allow for variation in coding practice geographically or over time.)