Where do prognostic estimates come from?
Doctors use information from groups of people who had the same or a similar cancer. Studies and cancer registries can show what happened across those groups over a defined period.
The estimate for an individual may then be informed by factors such as:
- cancer type, stage and grade;
- biological or molecular features;
- where the cancer is and whether it has spread;
- treatments available and their purpose;
- age, general health and performance status;
- how the cancer responds to treatment.
The relevant factors are not identical for every cancer.
Why do statistics not predict an individual outcome?
A survival percentage describes a group, not the future of each person within it. People in the group may differ in health, cancer biology and treatment. The data may also come from people diagnosed several years earlier, before a newer treatment became available.
Statistics can describe a range and help inform decisions, but they cannot identify where one person will fall within that range. Even a very experienced clinician cannot reliably convert a population average into an exact date for an individual.
Why can a prognosis change?
The first estimate is based on the information available at diagnosis. New information can alter it. This may include a pathology result, an operation finding, a molecular test, a change in health or evidence about how the cancer is responding.
A response to treatment may make the outlook more favourable than first expected. Cancer growth despite treatment may point in the other direction. This does not mean the original explanation was dishonest; it means prognosis is updated as the clinical picture develops.
Prognosis and treatment prediction are related but different
Prognosis concerns the likely course of the illness. Prediction concerns the chance that a particular treatment will help. A cancer can have an overall difficult prognosis while still having a treatment that offers worthwhile benefit. Another cancer can have a generally favourable outlook but gain little from an additional treatment.
The distinction matters when discussing benefits, side effects and alternatives.
How much do I have to know?
People differ in how much prognostic information they want. Some want figures and ranges; others prefer a broad explanation. It is reasonable to tell the team what would be helpful and whether you want a family member present.
If a number is given, ask what group it came from, what period it covers and how closely that group resembles your situation. Online survival tools can be misleading when the cancer type, stage or treatment context is not an exact match.
Questions to ask
- Which factors are most important for my prognosis?
- Are you describing the chance of cure, recurrence, treatment response or length of life?
- Is this estimate based on a broad group or people similar to me?
- What uncertainty remains?
- What future result might change the estimate?
Prognostic information can support planning and treatment decisions, but uncertainty is unavoidable. A useful conversation makes that uncertainty clear instead of presenting a group statistic as a promise about one person.
