Cancer Survival Statistics: What the Numbers Do Not Tell You
Cancer survival statistics describe what happened to groups of people diagnosed in the past. They cannot say what will happen to any one person. The American Cancer Society puts it plainly: survival rates “cannot tell you exactly what will happen with any one person.” That does not make the numbers useless, since they help researchers, health services and policy makers, but it does mean they need careful reading. This page explains the common terms and the limits of the figures. It is general information, not medical advice.
Where the numbers come from
In the United States, much of the data comes from the National Cancer Institute’s Surveillance, Epidemiology, and End Results programme, usually called SEER. It collects information on cancer cases from population-based registries that cover about 48% of the US population, according to the NCI. Other countries have their own registries, and their methods can differ.
Registries record who was diagnosed, with what type and stage of cancer, and what happened to them over time. Statisticians then turn those records into rates.
The terms you will meet
Five-year survival rate. The percentage of people in a study group who are still alive five years after diagnosis or the start of treatment. Five years is a convention, not a magic point, and some cancers can return later.
Relative survival. A comparison between people with the cancer and people in the general population of the same age and sex. The American Cancer Society explains it with an example: if a five-year relative survival rate is 80%, people with that cancer are on average about 80% as likely as people without it to live at least five years after diagnosis. This measure takes into account that people also die of other causes.
Stage groupings. The SEER database sorts cancers by how far they have spread rather than by the numbered stages used in clinics: localised (confined to where it began), regional (reached nearby structures or lymph nodes) and distant (spread to distant parts of the body).
What the numbers do not tell you
They describe the past
The American Cancer Society notes that survival rates are based on people diagnosed and treated at least five years earlier. Treatments change, so people being diagnosed now may have a different outlook from the one the figures suggest. The NCI makes a similar point by stating which years each figure covers: incidence, mortality and survivor counts all refer to specific past periods.
They describe groups, not people
A rate is an average across many individuals who differ in age, general health, the exact subtype of the disease, where it is located, how it responds to treatment and many other factors. The NCI states that statistical trends are usually not directly applicable to individual patients. Two people with the same diagnosis can have very different experiences.
They depend on how groups are defined
A figure for “all sarcomas” mixes many different diseases. A figure for one subtype and one stage is more specific but rests on fewer cases. Very rare cancers, such as alveolar soft part sarcoma, generate small numbers, so statistics for them may be less stable and may be reported only in broad groupings. When numbers are small, a few cases can shift a percentage noticeably.
They are affected by how cancer is found
The NCI notes that rising incidence does not always mean worse outcomes. Screening can find cancers that would never have caused problems, a situation called overdiagnosis, so death rates are considered a better indicator of progress than the number of new cases.
They can be compressed in the news
Headlines often reduce a detailed report to one percentage, losing the definitions that give it meaning. The summary of news coverage on this site is a reminder that the same story can be told in many ways, so it is worth reading past the headline to the source.
A worked example, with made-up numbers
Imagine a registry follows 100 people with a hypothetical condition, and 60 are alive five years after diagnosis. The five-year survival rate is 60%. Now suppose that in the general population of the same age and sex, 95 of every 100 people would be alive after five years. The relative survival is 60 divided by 95, which is about 63%. The numbers are invented to show the arithmetic and say nothing about any real cancer.
Questions to take to a medical team
If you are reading statistics in connection with a diagnosis, these are useful questions for a doctor:
- Which group do these figures describe, and how does my situation differ?
- How old is the data, and have treatments changed since?
- What do the numbers look like for my specific type and stage?
- What factors in my own case might make my outlook better or worse?
Reliable places to read further
- The NCI’s pages on understanding cancer statistics explain incidence, mortality and survivor counts.
- The American Cancer Society publishes survival-rate explanations for individual cancer types, including how to read the SEER stage groups; its soft tissue sarcoma page is an example of the format.
Frequently asked questions
What is a good five-year survival rate?
There is no single threshold. Rates differ widely between cancer types and stages, and a number only makes sense when you know which group it describes and when the data were collected.
Can survival statistics predict my outcome?
No. They describe groups of people diagnosed in the past, so they cannot predict an individual’s course. A doctor can discuss what they mean in a specific situation.
Why do survival rates for the same cancer differ between websites?
Sources may use different time periods, countries, age groups, stage definitions or data sets. Check the dates and definitions behind each figure.
Does “five-year survival” mean a person has only five years?
No. It is a measurement point used for comparison. Many people live well beyond five years after diagnosis.
Numbers can inform, but they are not a forecast. Read the definitions, check the dates, and bring questions to a qualified medical professional.