What Is a Survival Rate and How Is It Calculated?

A survival rate is the share of people with a disease who are still alive a set time after diagnosis, most often five years. It answers a plain question about a group: out of everyone diagnosed, how many were still here when the clock ran out? The numbers come from the records of past patients, so they describe how a group fared, not what will happen to one person.1PubMed. Cancer patient survival–patterns, comparisons, trends–a population-based Cancer Registry study in Finland

What a Survival Rate Measures

The simplest version, called overall survival, is the percentage of people still alive a specific time after diagnosis.2EBSCOhost. Understanding disease survival rates It is worked out as the number alive at the end of the period divided by the number at the start, times 100.2EBSCOhost. Understanding disease survival rates If 1,000 people are diagnosed and 650 are alive five years later, the 5-year survival is 65%. Every death counts, whatever the cause, not just deaths from the disease.

The most quoted figure is 5-year survival. Five years is long enough to show how many cancers usually behave, and it is widely used to track progress against cancer.3PubMed Central. Five-Year Survival is Not a Useful Measure for Cancer Control in the Population: an Analysis Based on UK Data3PubMed Central. Five-Year Survival is Not a Useful Measure for Cancer Control in the Population: an Analysis Based on UK Data It is not a cure rate, though. Some people alive at five years relapse later or eventually die of the disease.

A 65% figure does not mean everyone lives exactly five years. People alive at the five-year mark may live for decades; the figure simply stops counting there.

A figure is only as specific as the group it came from, so a number covering every stage of a cancer combined can look very different from the number for one stage. When a percentage appears in a news story, find out what disease, what stage, what years and what endpoint were counted. Without those, the number is hard to interpret.

How Survival Rates Are Calculated When Some People Drop Out

In real records, not everyone’s outcome is known. Some people move away, withdraw from a study, or are still alive when the study ends.4University of Oxford. Survival Analysis For these people researchers only know that they survived at least until their last contact. Their true survival is equal to or longer than the time they were watched.5The Basics of Survival Analysis. The Basics of Survival Analysis Researchers call this censoring, and you may see the term in reports: the person is set aside from that point on rather than counted as a death.5The Basics of Survival Analysis. The Basics of Survival Analysis

Handling these people any other way would mislead. Counting dropouts as survivors makes the rate look too good, and counting them as deaths makes it look too bad. Dropping them entirely wastes the years they were known to be alive. Censoring uses exactly the information available, no more and no less.

The most common method splits follow-up into stretches and works out the share surviving each one.6PubMed Central. A PRACTICAL GUIDE TO UNDERSTANDING KAPLAN-MEIER CURVES Those shares are multiplied together to get overall survival up to that point.6PubMed Central. A PRACTICAL GUIDE TO UNDERSTANDING KAPLAN-MEIER CURVES Say all 6 people survive the first stretch, 5 of 6 survive the second, one more person drops out, and 4 of the remaining 4 survive the third… Survival is 6/6 × 5/6 × 4/5, about 67%.6PubMed Central. A PRACTICAL GUIDE TO UNDERSTANDING KAPLAN-MEIER CURVES The third factor is 4/5 rather than 4/6 because someone who drops out is removed from the count of people at risk for the next stretch, so they don’t skew the result.6PubMed Central. A PRACTICAL GUIDE TO UNDERSTANDING KAPLAN-MEIER CURVES

The same logic applies to someone followed for two years and then lost. That person is counted as at risk for every stretch they were watched and removed only after their last contact, so their two years of known survival still count. It also means the estimate rests on fewer people at later times, so the tail end of a survival figure is less certain than the beginning.

An older “life table” method groups time into fixed intervals such as years.7Charles University. What Is a Survival Rate and How Is It Calculated? It assumes people who drop out did so halfway through the interval, counting them as half a person at risk.7Charles University. What Is a Survival Rate and How Is It Calculated? It works best when dropouts are spread evenly over time.7Charles University. What Is a Survival Rate and How Is It Calculated?

Reports often show the result as a survival curve, a stepped line that falls over time. Each step down is a death, and small tick marks often show people who dropped out. To read it, find a year along the bottom and look up to the line: the height there is the survival at that time.

Relative Survival and Other Types You May Hear About

Many cancer statistics give relative survival instead of overall survival.8PubMed Central. Relative Survival Analysis Using the Centers for Disease Control and Prevention’s National Program of Cancer Registries Surveillance System Data, 2000-2007 That is because not every death among cancer patients is caused by the cancer.9Robert Koch Institute. Survival rates Relative survival compares patients with people of the same age and sex in the general population.9Robert Koch Institute. Survival rates A relative 5-year survival of 90% means patients were 90% as likely to be alive as similar people without the cancer.8PubMed Central. Relative Survival Analysis Using the Centers for Disease Control and Prevention’s National Program of Cancer Registries Surveillance System Data, 2000-2007

This removes much of the effect of ordinary mortality. It is especially helpful for diseases common in older people, where many patients would have died within five years anyway. It also doesn’t require knowing the exact cause of each death, which can be unreliable or hard to record. For low-risk cancers, relative survival can approach 100%. That means patients fare about as well as the general population, not that no one died.

Treatment trials often report progression-free survival, the time from starting treatment until the disease grows or worsens.10PubMed Central. Relationship between Progression-free Survival and Overall Survival in Randomized Clinical Trials of Targeted and Biologic Agents in Oncology Overall survival, how long patients live after starting treatment, is the accepted direct measure of benefit.10PubMed Central. Relationship between Progression-free Survival and Overall Survival in Randomized Clinical Trials of Targeted and Biologic Agents in Oncology The two don’t always move together: in cancer drug trials about 1 in 5 comparisons improved disease control without extending life.10PubMed Central. Relationship between Progression-free Survival and Overall Survival in Randomized Clinical Trials of Targeted and Biologic Agents in Oncology

So a headline saying a drug “extends survival” may be based on the time before the disease worsens, which is not the same as living longer. That time can still matter, since a longer stretch of control may mean better days, but it is a different benefit from more total life. When a treatment is described as improving survival, it’s worth asking which kind. If the study did not measure overall survival, the benefit might still be real but unproven.

Statisticians also use separate methods to estimate the chance of the disease itself causing death when other causes compete.11PubMed Central. Introduction to Survival Analysis in the Presence of Competing Risks These estimates follow people who survived both the disease and other causes in earlier periods.12British Journal of Cancer. A note on competing risks in survival data analysis

What a Survival Rate Can and Can’t Tell You About One Person

Survival rates come from grouped data and can’t be applied directly to predict one patient’s outcome, but they can be used as guidelines.1PubMed. Cancer patient survival–patterns, comparisons, trends–a population-based Cancer Registry study in Finland They average people of many ages, health levels, stages and treatments. Doctors weigh factors like stage, age and response to treatment when discussing a specific outlook.

Stage at diagnosis is often one of the strongest factors, which is why figures broken down by stage are more informative than a single combined number. How a tumor responds to treatment is unknown at diagnosis, so the outlook can be revised as the response becomes clear. Age and other health conditions affect which treatments are possible and how well they are tolerated, which in turn affects outlook.

Old data can also understate today’s outlook. A 5-year figure needs five years of follow-up, so it describes people diagnosed at least five years ago. Treatments introduced since then aren’t reflected, and for many cancers the numbers improve over time.

Framing changes how numbers feel, too. The same facts land differently depending on whether they’re stated as survival or as deaths.13PubMed. Framing effects in medical situations: distinctions of attribute, goal and risky choice frames People offered treatment options described by survival chances choose riskier options than when given death rates.14PubMed Central. Five Different Types of Framing Effects in Medical Situation: A Preliminary Exploration “70% survive” tends to feel hopeful and “30% die” tends to feel alarming, though they describe exactly the same facts. If only one version is given, subtract from 100 to get the other.

How Your Outlook Changes After Surviving Each Year

Conditional survival is the chance of surviving a further number of years, given that someone has already survived a certain number after diagnosis.15PubMed. Conditional Survival: A Useful Concept to Provide Information on How Prognosis Evolves over Time It updates the outlook as time passes, instead of relying on the figure from diagnosis day.16PubMed. Dynamic prognostication using conditional survival estimates

The reason is that for many cancers the riskiest period is the first few years, so those who get past it often face better odds. At diagnosis, the 5-year figure includes everyone who dies early. Once you have lived through the first year, those early deaths no longer apply to you. The shift tends to be largest for cancers that are most dangerous early on, and smaller for those with a steady risk over many years.

In practice, someone three years out can ask for figures for people who reached three years. Those figures are often more encouraging than the original, and they help with planning work, family and follow-up care. Not all cancers improve the same way, so ask about the specific type.

Why Screening Can Make Survival Rates Look Better Than They Are

Screening finds disease earlier, so the survival clock starts sooner.17Catalogue of Bias. Lead time bias Survival time then looks longer in screened people even if early detection doesn’t change when they die.17Catalogue of Bias. Lead time bias In one esophageal cancer screening program, this effect inflated 5-year survival by more than 10%.18PubMed Central. Lead-time bias in esophageal cancer screening in high-risk areas in China

Take a tumor found through symptoms at age 67, with death at 70. That is a 3-year survival. Now suppose the same tumor is found by screening at 63, with death still at 70. That is 7-year survival. The statistic rose by four years, yet nobody lived a day longer. Survival is measured from diagnosis, so moving the diagnosis date earlier adds time to the count, and the extra time was simply spent knowing about the disease. This is why a higher survival rate alone doesn’t prove a screening benefit.

Screening is also more likely to find slow-growing, less aggressive cancers, which exist longer before symptoms.19National Cancer Institute. What Cancer Screening Statistics Really Tell Us Screen-detected cases then look like they survive longer, an advantage not caused by earlier detection itself.20PubMed. Continuous tumour growth models, lead time estimation and length bias in breast cancer screening studies In some cases screening finds cancers that would never have caused harm, and these add to the survivors without anyone being helped.

This means 5-year survival is limited as a way to judge whether cancer control is working.3PubMed Central. Five-Year Survival is Not a Useful Measure for Cancer Control in the Population: an Analysis Based on UK Data Death rates in the whole population are a fairer test of whether screening saves lives.

Why Survival Rates Differ Between Countries

Five-year breast cancer survival ranges from 39.1% in WHO’s African Region to 88.5% in the Americas.21World Health Organization. First global breast cancer survival estimates show wide regional and income disparities It also rises with national income, from 41.9% in low-income countries to 87.3% in high-income ones.21World Health Organization. First global breast cancer survival estimates show wide regional and income disparities Survival reflects early detection, timely diagnosis and access to effective treatment, and it was lower where breast cancer was usually found at a later stage.21World Health Organization. First global breast cancer survival estimates show wide regional and income disparities Access to surgery, drugs and radiation also varies a great deal between places, and this affects survival independently of when the cancer was found.

A national figure reflects health systems, not just biology. It blends together people diagnosed early and late, treated well and poorly, so any one patient can sit far from the average. Someone diagnosed at an early stage and treated promptly generally has a better outlook than a low-survival country’s number would suggest. Only some countries have cancer registries, so global figures are partly estimated.