Wednesday, June 9, 2010


As it does regularly, the issue of survival prediction arose today in our support group meeting.

When we’re informed that we have a life-threatening disease, it’s natural for us to ask how life-threatening it is. Is the Reaper at the front door now, or what? “How much time do I have, Doc?” is probably the most common way of asking the question.

The answer, of course, is that no one knows. No one can know. We’re aware that's the case, but still desire some guideline, a frame. So the doc gives us statistics, maybe saying something like, “Well, eighty percent of people with your diagnosis survive two years.”

There’s something essential to know about disease statistics: by definition, they describe the characteristics of a group, never an individual. How does a group picture apply to you? Do you have an eighty percent chance of surviving two years? No. In two years you’ll either be alive or you won’t—one hundred percent or zero.

How should you interpret these numbers, then? More often then not, patients remember only the time factor—in this case two years—and carve that into their memory, forgetting the percentage: “Doc gave me two years.”

That’s one problem these statistics can generate. Another one is that once you hear “two years” you can’t forget it. Think of a judge admonishing the jury to “ignore the previous testimony.” Impossible, so that duration pronouncement, sitting in the back of your mind like a dour raven, can become voodoo. I’ve seen too many people succumb right on time. Hasn’t much of world literature taught us how self-fulfilling prophecies can be?

There’s another problem, too. We traditionally describe survival rates with a mathematical model called a “bell curve.” You’ve seen plenty of these. It looks like this:

The curve says a small number of patients die soon, a similar number survive a long time, and the majority fall in between. When you hear statistics involving your life-threatening disease, all you know for sure is that you’re somewhere, anywhere, in that curve. (In fact, even when you’re as fit as a fiddle, you, being mortal, are somewhere in that curve.)

Bell curves are based on the results of large-scale studies that take just a few categories into account: type of cancer, age of patient, and gender, for example. The studies seldom consider less easily measurable characteristics, like the patient’s degree of interest and involvement, quality of relationships, exercise, diet, stress management, spiritual skills, and so on.

A few do, though. Several years ago a study of women with stage four breast cancer compared their survival with the number of “confidants” they enjoyed. A confidant was defined here as a close personal friend, someone with whom the patient felt comfortable discussing anything. Women with no confidants evinced a particular average survival time. Those with one confidant lived longer, on the average, and those with two even longer. The correlation held for up to six confidants.

I know oncologists who, even when pressed, offer no numbers. They prefer that their patients spend their energy moving themselves rightward under the curve  rather than fretting about their location in it.

No comments:

Post a Comment