Figuring Out the Subtleties of Using Statistics in Serious Illness Discussions

It seems to me that one of  the hardest things about serious illness discussions for the patient-professional team is figuring out how to think and talk about statistics.

This weekend I have been struggling to understand an article in the New York Times about this, headlined, I Had a 9 Percent Chance. Plus Hope.  The paragraph below  is, for me, the hardest to understand:

My surgeon refused to predict my future and told me to ignore my survival odds, which were a low 9 percent. He pointed out a cognitive error that people make when looking at statistics. If you survive, he reminded me, you’ll survive 100 percent. No one survives 9 percent. No one is 9 percent dead.

The numbers that matter are 0 and 100, dead or alive. To make an analogy to a commonly cited statistic, 50 percent of marriages end in divorce. There’s a part of the brain that therefore mistakenly believes all marriages experience 50 percent worth of divorce, making that number seem relevant to all couples, when it isn’t. For the happy couples, it is irrelevant. That meant that if I survived, it wouldn’t matter, looking back, whether I’d beaten chances of 9 or 19 or 90 percent.

I really did not understand.  What is the “cognitive error” here?

My family tried to explain by telling me that there is a difference between the use of statistics in making decisions about treatments, money, etc., and their use in living once the decisions are made.  While it makes all the sense in the world to know and use statistics to decide whether to use one or another treatment, or indeed whether a treatment attempt is best chosen or rejected, much of living that then happens does so without a focus on the survival numbers, it was suggested.  Or rather the family thought that that was what the writer was saying.

At first I found the distinction both persuasive and helpful.  And I certainly agree that numbers play very different roles in different situations and for different people.

But the more I thought about it, the less this makes overall sense.  As someone with an four to five year life expectancy because of bone marrow cancer, every experience in life, as well as every decision I make, is colored by that knowledge.  Its not just that it influences a decision abut whether oral surgery is worth the time disruption, effort and pain.  It influences my feelings about family holidays, seeing friends, and even politics, to put those in the context of a timeline and life expectation.  How can it not.

To put it another way, following the construct in the article, if your survival expectation is 100%, then each event is lived with that awareness and has a particular nuance.  If it is 0%, then it has a very different nuance.  Most importantly, if it is 50%, or any number between 0 and 100 then each event has bits of both sensibilities with an awareness of every possibility.

For me, at least, the only way I can live with the reality of constant risks of sudden change points, particularly when tests are done, and with knowledge that there will be inevitable bad news sometime, is to not to always find good or bad news with each test.  That only puts you on an emotional roller coaster.  I try to see it as a long process with gently smoothed ups and downs, gliding in the end, as with all roller coasters, down to a hopefully gentle stop.  For me, being aware of the stats, and of how they move for me, helps smooth the process out.  When there is potentially bad news, well, we knew that would come in the end.  When there is good news, well, that’s only for a time.

Most generally, I do conclude that different patients want, need, and can use very different information.  The doctor has to be able to figure out what will truly help the patient, and how to give it.  That may or may not be identical to the patients initial articulation.  On the contrary, the patient’s statement may be driven by a very different fear than that of actual knowledge about what is likely to come.  While the professional must respect and engage with the patient’s articulation of what and how they want to know, they have to recognize that that choice might change after honest joint exploration.