This website places cookies on your device to help us improve our service to you. To find out more about the cookies, see our Privacy and Cookies notice
ISD Scotland better information, better decisions, better health

Frequently Asked Questions

Cancer

NHS National Services Scotland ISD Scotland & NHS National Services Scotland

Cancer Information FAQs

Where do the data come from?
What is your definition of incidence?
What is your definition of mortality?
What is your definition of survival?
What is your definition of prevalence?
What is your definition of lifetime risk?
What is your definition of deprivation?
What do you mean by 'person years at risk' in your rate calculations?
What is the latest year of complete cancer incidence data?
Why are data for 2016 not yet available?
When do you expect complete cancer incidence data for 2016 to be available on the website?
What is the latest available year of cancer mortality data?
Why do the data I am looking at differ from other published data for the same time period?
Where can I find out the codes used to classify each type of cancer?
Why are data for non-melanoma skin cancer sometimes excluded?
Where can I get cancer registration data covering other parts of the UK?
How do you assess the quality of the cancer registration data?
Why have you published the information that you have on your website?


Where do the data come from?

The majority of the analyses are based on the Scottish Cancer Registry (SCR) data. Managed by ISD, the SCR has been collecting data since 1958 and is the authoritative source of cancer data in Scotland. The mortality data come from the National Records of Scotland.

The population data, used to calculate rates of cancer, also come from the National Records of Scotland. Normally the Mid-year Population Estimates (MYPEs) (population estimates at 30 June of a given year) are used.

back to top


What is your definition of incidence?

Incidence is the total number of new cases (registrations) of the cancer diagnosed in Scotland for the given period.

back to top


What is your definition of mortality?

The number of patients who died and whose primary cause of death was the specified cancer during the given period. Deaths data is sourced from National Records of Scotland. Information on coding, and changes in coding over time can be found here. The mortality statistics published on our website are based on the date of registration of the death rather than the date on which the death occurred. This is in order to be consistent with the information published by National Records of Scotland. By law, a death should be registered within 8 days. More information can be found on the National Records of Scotland website.

Cancer mortality statistics specifically for under 75 year olds are published to inform the Scottish Government target of a 20% decrease in mortality in this age group (European Age Standardised Rate, both sexes combined) during the period of 1995 and 2010. This target was achieved and statistics continue to be published to enable monitoring over the longer term. For more information on this target, please see the Better Cancer Care document, published by the Scottish Government in February 2008.

back to top


What is your definition of survival?

Relative survival is calculated as the ratio of the observed survival of a group of cancer patients, divided by the expected survival for a group of the general population having a similar structure in terms of age and sex. This can be thought of as a measure of net survival expectation after contracting cancer, or the probability of survival from cancer in the absence of other causes of death.

Observed survival estimates vary between age groups, not only because their risk of dying of the cancer varies, but also because of their differing risk of dying from other causes. Relative survival estimates adjust for the latter differences and so can be used to compare survival between age groups or between populations with different age distributions. However relative survival is unlikely to completely represent the true 'net' survival. For example, it is possible that patients diagnosed with cancer are more likely to die from other causes than the general population and hence relative survival may still be an overestimate of 'net' survival.

back to top


What is your definition of prevalence?

Cancer prevalence is a measure of the burden of cancer in a population at a particular point in time.

One-year prevalence is the number of patients diagnosed in the previous year who are still alive; 20-year prevalence is the number of patients diagnosed in the previous 20 years who are still alive. For example at 31 December 2011, 10 year prevalence is the number of patients diagnosed between 2002 and 2011 who are alive at 31 December 2011; while 20-year prevalence is the number of patients diagnosed between 1992 and 2011 who are alive at 31 December 2011.

Prevalence is estimated by counting the number of patients diagnosed before 31 December 2011, excluding those who are known to have died or migrated by that date. It is a person-based rather than a tumour-based measure. Each patient is counted only once within each specific cancer, or grouping of cancers, using the most recent date of diagnosis up to 31 December 2011. Therefore the number of patients diagnosed with, for example, the category of head and neck cancer (ICD-10 C00-C14, C30-C32) is expected to be less than the sum of the individual prevalence estimates for cancer of the lip, oral cavity and pharynx (ICD10-C00-C14), cancer of the nasal cavity and middle ear (C30), cancer of the accessory sinuses (C31) and cancer of the larynx (C32).

There are several sources of error which are likely to increase as the elapsed time since the date of diagnosis increases. For example, patients diagnosed with cancer may leave Scotland without this fact becoming known to the cancer registry. Inclusion of these patients will inflate the estimates. On the other hand, the estimates of prevalence will exclude patients who lived outside Scotland at the time of diagnosis and have subsequently moved to Scotland.

Because of these sources of uncertainty, and because it is often more appropriate to consider only those patients diagnosed fairly recently, who are most likely to require health service care, estimates of prevalence are shown by time survived since diagnosis (up to 1 year, >1-5 years, >5-10 years and >10-20 years). To estimate how many people are alive with cancer whose diagnosis was within the last 10 years, for example, the prevalence estimates for up to 1, >1-5 and >5-10 years survived since diagnosis need to be summed. The fourth column of the table shows the percentage of patients within each of these time intervals; for example, only 13.1% of males living with lung cancer were diagnosed more than 10 years previously.

Prevalence is influenced by both incidence and survival rates. If the survival prognosis for a particular cancer is low, then the prevalence of patients with that cancer will also be low. As incidence and survival are both influenced by age, prevalence estimates are also shown by age (this is age at 31 December 2011, not age at cancer diagnosis). The fourth column of the table shows the percentage of patients within each age band; for example, 76.6% of males living with lung cancer are aged 65+.

Prevalence is calculated only up to the point where cancer registrations are considered complete; otherwise the prevalence estimate would be skewed downwards, particularly in the most recent diagnosis years.

back to top


What is your definition of lifetime risk?

A person's chance (or risk) of developing cancer up to a specific age may be expressed either as a percentage (Y%) or as 1 in X. For example, a man might experience a 5% or 1 in 20 chance of developing a particular cancer up to the age of 84 years (where X=100/Y).

A person's chance (or risk) of developing a particular cancer from a specific age is the average chance that an individual has of subsequently developing the cancer if he/she has already reached that age without getting cancer. For example, a man of 64 might have a 2% or 1 in 50 chance of developing a particular cancer during the remainder of his life.

We present information using a method of calculation known as the 'lifetime risk' method, which we believe gives a truer estimate of the risk of developing cancer than the 'cumulative rate' method. This is because our chosen method makes allowance for the fact that people may die of other causes, and further, once they have developed cancer they are no longer at risk. For instance when calculating the risk of developing cancer up to the age of 84, it takes into account the fact that a proportion of the population will die of other causes before the age of 84 and are therefore no longer at risk of developing cancer, and also that once a person has been diagnosed with cancer they are also no longer at risk of developing cancer for the first time.

The 'cumulative rate' method, on the other hand, is calculated by summing the age-specific incidence rates for the cancer in question up to the specified age. It does not take into account competing causes of death, and therefore over-estimates the risk, at birth, of developing cancer at older age groups. This is because it assumes that the same people are at risk at (say) age 84 as at birth, whereas in practice some of these people will have died of other causes, or will already have developed cancer.

The calculations are based on death rates and cancer incidence rates for the period 2007-2011. This means that, although the above discussion refers to 'a person's risk of developing cancer' the figures presented do not show the specific experience of any particular person or generation; this would require the calculation of age-specific rates for consecutive time periods throughout the person's life.

back to top


What is your definition of deprivation?

Historically, ISD has used mainly Carstairs 1991 to measure deprivation in Scotland. However, with the availability of the Scottish Index of Multiple Deprivation (SIMD), ISD reviewed how deprivation should be measured in national health statistics. Full details of the review and subsequent recommendations can be found at Deprivation.

back to top


What do you mean by 'person years at risk' in your rate calculations?

Rates are intended to be annual rates (number of cases per population), or average annual rates in the case of combined year periods. Person-years is based on the mid-year population estimate for a single year, or the sum of the mid-year population estimates in the case of combined year periods. Say, for example, that you have a defined geographical area that happens to have a consistent population of 10 people; your population at risk for that year, as well as your person years at risk for one year, would be 10. Over a 5 year period, that will give you 50 person years at risk. It may not be the same 10 people each year - some may leave and others enter - but that does not matter because we are looking at the incidence rate within that defined area.

To expand on the example of person years at risk and population at risk, say you want to calculate rates for person years at risk if you have 50 cases of a cancer diagnosed in 2years and a starting population of 200,000. Your incidence rate per 100,000 is (50/(200,000*2 years))*100,000 = 12.5 cases per 100,000 people in 2 years. Note that 400,000 would be the number of person years at risk.

If you are looking at incidence rate for 1 year then you would divide your rate by the number of years observed. So for the number of cases for 1 year then ((50/(200,000*2))*100,000)/2 = 6.25 per 100,000 person years at risk for 1 year.

The 'at risk' subset of the population is taken to mean everyone of the specified gender(s) and within the specified age range. For example, rates of cervical cancer are only calculated for females in the population, as they are the only ones at risk of cervical cancer.

back to top


What is the latest year of complete cancer incidence data?

Cancer registrations are currently considered to be complete for 2015.

back to top


What is the latest available year of cancer mortality data?

Mortality data are currently complete for 2015.

back to top


Why are data for 2016 not yet available?

To achieve a completed cancer registration, multiple data sources are processed (for example, computerised hospital discharge records, pathology records and death records). Registration cannot be completed until at least six months after diagnosis to allow treatment information to accrue. As a result of this, the target for UK cancer registries is to achieve complete registration for a calendar year within 12 months of the end of that year. The majority of cases for a given incidence year should be registered by this point, although it is recognised internationally that registrations may come to light many months or even years after diagnosis and the cancer registration database is continually being updated. Once registrations are essentially complete we allow up to 2 months to work on Quality Assurance (QA) of the data then wait for several more months before publishing data to allow for the inclusion of late registrations.

back to top


When do you expect complete cancer incidence data for 2016 to be available on the website?

It is planned that 2016 cancer registrations will be published on the website at the end of April 2018.

back to top


What is the latest available year of cancer mortality data?

Mortality data are currently complete for 2015.

back to top


Why do the data I am looking at differ from other published data for the same time period?

Cancer registrations may come to light many months or even years after diagnosis and the cancer registration database is continually being updated. The data presented here may therefore, particularly for recent years, differ slightly from other published data relating to the same time period. Hence the statistics we publish are updated on a regular basis.

back to top


Where can I find out the codes used to classify each type of cancer?

Details are provided here Download Excel file [20KB] for all sites or can be found in the pages for each individual cancer site. Background information on the classification and coding used over time in the Scottish Cancer Registry, with some notes on the impact of these changes, can be found on the Scottish Cancer Registry page

back to top


Why are data for non-melanoma skin cancer sometimes excluded?

Registrations for non-melanoma skin cancer (ICD-10 C44) are likely to be less complete and less accurate than for other cancer sites. Such cancers are relatively common and usually non-fatal. There is a propensity for multiple tumours to occur in one individual and cancer registries adopt different practices in recording these. The tumours are most common in the elderly population and the completeness of registration in the very elderly is likely to be less than for younger patients. Furthermore, increasing numbers of these cancers are diagnosed and treated within GP surgeries and the registration scheme is not confident that all such cases are notified. Because cancer registries across the world have different practices for recording non-melanoma skin cancer (some do not record them at all), the category "all cancers combined" often omits these tumours in the interests of making international comparisons of cancer incidence more valid.

back to top


Where can I get cancer registration data covering other parts of the UK?

England: Office for National Statistics
Wales: Welsh Cancer Intelligence and Surveillance Unit
Northern Ireland: Northern Ireland Cancer Registry

back to top


How do you assess the quality of the cancer registration data?

Several studies have been undertaken to examine the quality of the data held in the Scottish Cancer Registry, and where possible to compare the quality to that of other registries. Information on the studies can be found in this note Download pdf file [25KB].

back to top


Why have you published the information that you have on your website?

The depth and breadth of information on the cancer area of the ISD website is a product of information requests received, clinical demand, and data to inform Scottish Government policy. The Scottish Cancer Taskforce and its subgroups prompt evaluation of new or existing publications. Cancer Screening data largely address standards and/or targets set by the Scottish Government in conjunction with the National Services Division, who run the Cancer Screening Programmes.

Because of the internationally recognised extent and quality of the data contained in the Scottish Cancer Registry, we attempt to balance provision of data and information with manageability of volume of information. Any comments and suggestions would be most welcome.

back to top


© ISD Scotland 2010
Information Services Division,
NHS National Services Scotland,
Gyle Square,
1 South Gyle Crescent,
Edinburgh EH12 9EB,
Tel: 0131 275 7050
NSS.isdSHIS@nhs.net  
Contact Us

About Information Services Division | Contact Us | Accessibility | Complaints
Copyright | Terms and Conditions | Privacy | Freedom of Information

ISD is a division of NHS NSS
Working at the very heart of the health service, NSS delivers services critical to frontline patient care and in support of the efficient and effective operation of NHS Scotland. Visit the NSS Website for more information