SEER: where cancer survival statistics come from

Maybe you have read a survival percentage for your cancer. There is a good chance it traces back to one program. SEER stands for Surveillance, Epidemiology, and End Results. It belongs to the National Cancer Institute. It describes itself as an authoritative source of information on cancer incidence and survival in the United States. Incidence means how many new cases are diagnosed.

SEER does not treat anyone. It counts. It follows. It publishes.

That sounds dry until you realize what it makes possible. Without a program like this, nobody could say whether a cancer is becoming more common. Nobody could say whether people diagnosed with it are living longer than they used to.

SEER collects and publishes data from population-based cancer registries. A population-based registry tries to capture every cancer diagnosed among people living in a defined area. It is not just the cases at one hospital.

Those registries cover approximately 45.9 percent of the U.S. population. Coverage is not even across groups. SEER reports this coverage:

• 39.6 percent of Whites
• 43.5 percent of African Americans
• 64.9 percent of Hispanics
• 59.3 percent of American Indians and Alaska Natives
• 68.2 percent of Asians
• 69.9 percent of Hawaiian and Pacific Islanders

The participating registries are a mix of whole states and regions. The list includes Connecticut, Iowa, Utah, New Mexico, Louisiana, New Jersey, Hawaii, Idaho, Illinois, Kentucky, Georgia, Texas, and New York. It also includes regional registries such as San Francisco-Oakland, San Jose-Monterey, Los Angeles, Greater California, Detroit, and Seattle-Puget Sound, plus the Alaska Native Tumor Registry.

Data collection began in 1973 with a limited number of registries and has continued to expand.

Registries routinely collect a set list of items:

• patient demographics, such as age and sex
• the primary tumor site, where the cancer started
• tumor morphology (the type of cell under the microscope) and stage at diagnosis
• the first course of treatment
• follow-up for vital status, meaning whether the person is still living

That last pair is the heart of it. Registries record how far a cancer had spread when it was found. Then they follow whether people are living. That lets SEER produce survival figures broken out by stage. SEER says it is the only complete U.S. source of population-based data that includes both stage at diagnosis and patient survival.

Some pieces come from elsewhere. The mortality data SEER reports come from the National Center for Health Statistics. The population counts used to calculate rates come periodically from the Census Bureau.

It can give you honest, free numbers. The data are updated annually. They are provided as a public service in print and electronic formats.

The friendliest entry point is Cancer Stat Facts. These are short statistical summaries for a number of common cancer types. They are built to answer frequently requested questions quickly. Depending on the cancer, a page may include incidence, mortality, survival, stage, prevalence, and lifetime risk. There are pages for dozens of sites. They range from breast, lung, and colorectal to smaller groups such as vulva, small intestine, and certain lymphoma and leukemia subtypes. There are also pages for all cancer sites combined, for childhood cancers, and for adolescents and young adults ages 15 to 39. These pages are updated annually with the SEER data release. They link out to NCI information on risk factors, treatment, and clinical trials.

It can also help you tell a real statistic from a headline. Say you see a claim about cancer trends. You can check whether the underlying source is a program like this one.

There is a related tool worth knowing about. State Cancer Profiles, a joint project of NCI and CDC, offers cancer statistics for specific states and counties.

It cannot tell you what will happen to you. SEER publishes what happened to groups of people. Those people were diagnosed in past years, in covered areas, and were followed over time. Your age, your other health conditions, your tumor's biology, and the treatments available to you now are not in that number.

It cannot represent everyone equally. Its registries cover under half the country, and coverage differs by group. So national estimates involve pooling and modelling. NCI works with the North American Association of Central Cancer Registries. Together they help state registries produce data that can be pooled to improve national estimates.

It cannot be up to the minute. Survival statistics require years of follow-up by design. So the most recent release still describes earlier diagnoses.

It cannot answer questions about your own care. There is no SEER helpline for treatment decisions. There is no way to look up your personal record there.

A published survival figure is a description of a crowd. You are one person walking into a specific clinic on a specific day.

A number may frighten you. That reaction is not a failure of understanding. It may simply be the wrong tool for the moment. It is entirely reasonable to set the statistics aside and ask your team a narrower question instead: what are we doing next, and what are we watching for.