Member preview

Expected-Value Calculations Don’t Apply to Cancer

The problem with cancer is not so much the physical struggle, but the mental struggle to make rational decisions about treatments. We can withstand nearly any adversity so long as we believe it’s rational. And to believe something is rational, we have to believe in its value. But how do you measure the value of a life-or-death decision? Expected-value calculations are undefined when the costs and benefits are unlimited. Death is an infinite loss. Prolonged life is an infinite benefit.

We have some capacity to deal with expected-value in life-or-death situations. For example, when we swerve out of the way to avoid a car accident, we take on some risk. But the probability of success is close to one, which practically removes it as a factor for consideration. There is no expected-value, just value. One action to save one life. That’s easy.

In the case of cancer, treatment may have a 100% chance of six months of suffering for a 20% chance of prolonging your life another ten years. 20% multiplied by ten years is two years, but you can’t tell a patient, “this choice is worth two years.” While there is some research on maximizing QALYs or quality-adjusted life years, they require a battery of thought experiments. Patients have to consider various trade-offs, gambles, and preference-ranks to get to the bottom of what they “truly want.” But even if such tests could approximate the rational choice 95% of the time, the patient would have to undergo significant duress to comply. It takes an insane amount of guts to refuse treatment that has a 20% of success, even if it’s supposedly worth two QALYs. A 20% chance is high enough to activate hope, but not low enough to activate resignation.

Our ancestors rarely encountered these kinds of probabilistic scenarios. Our default approach is always to grasp for survival. If there is famine, strive for success. If you’re sick, try to fight it. If you fail, well, it was a good two-week battle, and then you’re done. Not a bad way to go out. There were no decades-long quests to cling onto life. In a way, modernity is worse, because we spend our golden years maximizing the number of last-ditch efforts to extend an increasingly poor quality-of-life.

One ray of hope is in cultivating public heuristics. According to this 2015 article in American Psychologist, humans rely on fuzzy traces for memory, which are a synthesis of verbatim memory and gists. Cancer options could, therefore, be framed with gist-like shorthand, such as “the graceful exit” or the “honorable choice.” Such work could help nudge patients to the rational choice, and help close the gap between how ordinary people die and how doctors themselves die.