Who doesn't love to think about the nonrational components of medical decisions and choices? If you know me, you know I'm fascinated by the decisions we make, both as doctors and as patients. Why do doctors and patients make choices that seem to fly in the face of “evidence based medicine”?
I came across an interesting opinion article in JAMA this month, about prospect theory and understanding choices and medical decisions. Right up my alley! Did I like it so much because the authors cited some of the same original work I had recently cited in my Anesthesiology review? Maybe. (This is called “Ingroup Bias” – I am more likely to feel favorably about those who share my views, and so although I don't know these authors, we are in the same “group”) But mostly, I loved the following sentence:
“Disagreements about medical decisions are often attributed to
differences in in how individuals balance risks and benefits.”
In other words, we have different preferences. Our decisions are not strictly rational, even as scientists.
Framing disease prognosis in terms of potential for survival may be received differently than framing in terms of decline in quality of life, or chance of death. This is true even when the facts are identical – the “frame” of the outcome as “good” or “bad” influences decisions. This is also tied in tightly to loss aversion. Humans tend to be risk-averse when there is a choice with certainty, and more inclined to gamble for a bigger win when a loss is possible.
Anchoring (though the authors did not specifically use this term) is important because it describes a reference point from which an evaluation is made. For example, the authors points out that many healthy people view severe disability and death as nearly equivalent (“I’d never want to live like that”) because the state of severe disability or death is far remote from their current state of health. A person in very poor health or disability, however, is likely to see a significant gap between their state of living and not living at all.
Moreover, the authors argue that the concept that value of outcome is generally rated more highly for proportional differences than absolute differences. As an illustration, the difference between one death and zero deaths (100% reduction) seems more significant than the difference between 1000 and 999 deaths (0.1% reduction), despite the absolute number of deaths being the same. Perhaps this explains why there is so much recent focus on “never” events, which are certainly important adverse outcomes that are perceived to be preventable, and are also fairly rare. Why not focus on adverse outcomes which are much more common? Does the commonness reduce the value of the reduction? If it is not the commonness, but instead the “catastrophicness” of the outcome, then why is wrong sided surgery – a shockingly catastrophic and surely preventable outcome – on the 2009 AHRQ list next to medication error, pressure ulcer, and patient falls. These are all bad outcomes, certainly, but I do not know if they are indeed always preventable, and they often do not carry the same catastrophic sequelae.
“Patient dissatisfaction, diagnostic error [my favorite topic], overtreatment or undertreatment, and even litigation can result when communication about key decisions goes awry. After clinical events unfold [outcome and hindsight bias], patients and families are too often left feeling that they would have made a difference choice has they known better [regret].” Although the brackets are mine, and refer to concepts I’ve discussed elsewhere, I cannot agree more that communication and aligned expectations are paramount to both good medical decisions and “customer satisfaction” in the age of patient centered care.
Have you had a recent decision that was influenced by one of these nonrational factors? How did you choose?