|
For
many of the participants in the study by Ritov
and Baron (1995), the knowledge that they would know
the fate of each child reduced the number of deaths they would
tolerate from the vaccine. Knowing which children they chose
to vaccinate would have lived but for the vaccination induced
regret in many people. Among those participants for whom knowing
the fate of the children influenced their decision, the vaccine
was not given unless it caused no more than an average of 4.1
deaths. On the other hand, when the deaths occurred to anonymous
children, 6.5 deaths were tolerated.
The
role of regret has been demonstrated in actual patients contemplating
real consequences. For example, Smith
(1996) analyzed the net benefit of adding a 52-week course
of adjuvant chemotherapy to surgical resection of colonic carcinoma.
Five-year survival could be increased from 55% to 71% by adding
the chemotherapy. However Smith found that in a patient sample
of 8, both overall and for every individual patient, the addition
of the chemotherapy resulted in a loss of QALYs
(quality-adjusted life years). Nevertheless, every patient chose
to have the chemotherapy.
Choosing
an option characterized by a loss of QALYs would seem to define
not acting in one’s own best interests. Why would patients do
so?
Some
theorists (Loomes
and Sugden, 1982; Bell,
1982) have suggested that the value of any alternative
depends on the value of the other alternatives available when
the decision is made. If the outcome obtained is inferior to
an alternative outcome that might have been obtained had another
action been taken, then there is regret for having chosen the
action that led to the inferior outcome. Smith
(1996) hypothesized that all of the patients he studied selected
the chemotherapy, because they would have regretted not selecting
that option. Chemotherapy held out the hope of a better outcome,
which all 8 patients would have regretted forsaking.
There
are other factors that can trigger regret. Consider a non-medical
scenario
taken from Kahneman
and Tversky (1982):
Kahneman
and Tversky (1982) found
that those who got the “unusual route” version were likely to
change that aspect of the story when completing the “if only”
thought. Those who got the “earlier departure” version were less
likely to change the route and were more likely to change the
time of Mr. Jones’ departure. The point is that the unusual factor
is more likely to be “mutated” or changed in order to undo the
bad outcome. This is why patients so often take the “glide path”
rather than opt for some less orthodox course of action. More
regret accompanies a poor outcome that follows an atypical course
of action (“Why couldn’t I just do what the doctor suggested?”)
as opposed to typical (“I did what everyone else does in that
situation, so my decision cannot be faulted.”) Such reasoning
probably also contributes to the omission bias, because harmful
commissions are often more blameworthy than harmful omissions
(Ritov and
Baron, 1992).
|