More Than a Statistic Cont.

Several factors need to be considered when discussing statistical averages versus possible outcomes for an individual. These include: comparing like cancers, variability in treatment, limitations of the 5 year time frame, compliance and future research.

Comparing Apples to Apples
When comparing outcomes you must compare cancers to "like" cancers and staging is disease specific. For example a stage I lymphoma of an indolent type (grows slowly) has a far better prognosis than a stage I resected lung cancer.

Variability in Treatment
Population based survival statistics are NOT comparable to the individual because of the variability in treatment and comorbities. Example, a stage 4 breast cancer that is of an indolent type that is treated with anti-estrogen therapy has a far higher survival than a stage 4 inflammatory breast cancer that is not hormone sensitive and is generally more rapid in metastasis

5 Year Limits
The whole issue of predicting "survival" is more social than scientific, in most of the medical literature. Funding for cancer research is usually not adequate to follow patients for "life" Patients move, change phones, change names, and change doctors so frequently that research studies tend to stop after 5 years, hence 5 year data limits. Ironically the "best" data on survival comes from actuarial life insurance databases. Research studies focus on "surrogate endpoints" or measures that are quicker and easier to get than survival. Example, time from cancer resection to relapse based on serial CT scans done at pre determined intervals. This is often called "Relapse Free Survival" and is a more common research endpoint.