Biomedical Engineering Reference
In-Depth Information
old and new medicine (the chance of dying from the drug) were formulated as either
a frequency or a probability (e.g., 0.0006). Results showed that those in the high-
risk level/frequency format condition were willing to pay a much higher price, not
only than those in the low-risk/frequency format condition but also than those in the
high-risk level/probability format condition. Interestingly, there was no differential
effect on willingness to pay between different risk levels for the probability format.
The latter result can be attributed to misinterpretation of risk stated as a probability.
After all, people of various educational backgrounds often have diffi culty in evalu-
ating numerical information in general (Schwartz et al.
1997
; Lipkus et al.
2001
;
Sheridan and Pignone
2002
), although natural frequencies are less likely to be
misinterpreted (Gigerenzer and Hoffrage
1995
; Hoffrage et al.
2000
; Cosmides and
Tooby
1996
). The results can also be attributed to the fact that, as described by the
affect literature, frequency information induced stronger emotional responses and
triggered vivid images of death.
Considering that research on decision-making has shown that small probabilities
tend to be underestimated (Kahneman and Tversky
1979
), though they may be over-
estimated if they are affect-laden (Rottenstreich and Hsee
2001
), people's (mis)
interpretation of verbal risk information is of little surprise. Verbal risk presentation
is generally conveyed in the form of “very common, common, uncommon, rare, or
very rare” (Berry
2006
). In the European Union, the European Commission has
imposed guidelines to direct such claims, such that, for instance, “common” must
refer to adverse side effects that occur in between 1 and 10 % of people who take
the medicine (Berry
2006
). Although in the United States there are no offi cial FDA
guidelines for the use of frequency descriptors, research shows that their use is
comparable to that in the EU and that their numerical meaning is well understood
among drug marketers (Cox et al.
2010
).
While the frequency and severity of side effects may be well understood among
pharmaceutical marketers, research suggests that this fl uency is absent among the
general consumer population. In both the United States and the European Union,
lay people tend to overestimate risk (Cox et al.
2010
; Knapp et al.
2004
; Berry et al.
2002
,
2003
) and conventionally interpret adverse reactions categorized by “com-
mon” as occurring in 45-50 % of cases; physicians, interestingly, also misinterpret
such language as around 25 % (Berry
2006
). In contrast to comparable numerical
descriptors, verbal descriptors have also been shown to correspond to more negative
perceptions of the medicine (Knapp et al.
2004
), greater perceptions of risk to
health, lower satisfaction with the information, and lower intensions to comply
(Berry et al.
2003
).
Overestimation of the likelihood of side effects, a signifi cant reason for consumers
opting to refrain from initiating the use of a particular medicine, is more prevalent
when risk information is presented verbally rather than numerically (Knapp et al.
2004
). For example, in a study conducted by Berry et al. (
2002
), the estimated mean
probability of experiencing an adverse reaction was over 3 times greater for those
given a verbal descriptor of side effects than those given a numerical value (64 % vs.
20 %, respectively) and verbal descriptors also corresponded to perceptions of increased
severity.
Search WWH ::
Custom Search