The sad news that Justice Ginsburg has pancreatic cancer has resurfaced the issue of cancer survival data. Epidemiologists have been trying to convince oncologists for decades that survival data are meaningless. And they more or less have, though recidivism is common among the latter. US News has an article that shows that we’ve still got a problem understanding survival data.
“But as with most cancers, early detection improves survival,” states the article. This is a tautology. How could it be otherwise? If you have a cancer that has no treatment – none – and it takes 10 years to kill you, your survival is one year if the disease is detected nine years from its inception. If the diagnosis is made one year after contracting the disease your survival increases 900% but you die at exactly the same time as you would if the diagnosis were made nine years later.
The right way to evaluate the effectiveness of early diagnosis and treatment is to look at age adjusted mortality for the cancer in question. If early diagnosis and treatment truly increase survival age adjusted mortality will decrease. But here’s where things get a little more complicated. If age adjusted mortality falls there may be other causes for the fall unrelated to early diagnosis. Prevention will decrease mortality rates, eg smoking cessation. Here lung cancer mortality falls not because early diagnosis makes a difference – it probably doesn’t, but because people can’t die from a disease they don’t get. Later treatment of some cancers may be just as effective as early treatment obviating the value of early diagnosis.
As I mentioned, the early diagnosis of lung cancer has yet to be shown to be of any therapeutic value. We are still arguing about the value of screening mammography for breast cancer though it’s standard of care. Breast cancer mortality rates have fallen, but whether this decrease is due to early diagnosis or more effective treatment or both is unclear. Prostate cancer is a mess. I’ve already commented on it. Colon cancer screening is the best example of the value of early diagnosis in reducing mortality. Removing colon cancers that are at an early stage or better removing polyps before they become malignant has markedly reduced death from this common disease.
So remember that survival data alone can’t be interpreted. Also let’s hope that Justice Ginsburg’s treatment is effective and that she makes a rapid recovery.
