Don't Let Me Stop You: Cancer Statistics

So how goes the war against cancer? Many people believe, erroneously, that there is a "cancer epidemic" today. More people are dying from cancer than in the past, but why is this?

Deaths from other causes have been reduced, but the overall mortality rate will always be 100%. Progress against cancer has been slower.
Diagnostic techniques have greatly improved, so deaths from "unknown" or non-specific causes (e.g. "old age") that were actually cancer are now correctly identified.
The average age of the US population has been increasing, and cancer is more common in older people.

These distortions in the mortality statistics of cancer cause, among other things, an irrational fear of "chemicals." In fact, literally everything is made of chemicals, and the distinction between "natural" and "man made" chemicals is virtually meaningless.

Similar distortions affect the statistics suggesting huge advances against cancer. Although there is no cancer "epidemic" and some forms of cancer are quite curable today, for other types progress remains elusive. One of the technical talks at the AACR described problems with interpreting cancer statistics. The speaker began with a Will Rogers quote, "When the Okies left Oklahoma and moved to California, they raised the average intelligence levels in both states." He noted that a population of cancer patients generally includes some whose disease is actually in a more advanced stage than is realized. If improved diagnostic techniques allow these patients to be correctly "staged," the survival statistics improve for both groups, even though no real treatment progress has been made. The early stage group will obviously have a longer average survival time when those with more advanced disease are removed. Less obviously, the shifted patients are also healthier than the average patient with advanced disease, so that group's statistics improve as well.

The ongoing, successful quest for ever earlier diagnosis has the same distorting effect on survival statistics. Finding the cancer 5 years earlier brings in patients that are likely to live about 5 years longer than the previous patient population even if treatment is not improved at all. Often these patients are easier to treat, too, improving the statistics even more.