Transcript What*s in a name? The effect of a disease label on parent*s

Cancer screening decisions:
The role of feelings vs. beliefs
LAURA D. SCHERER
K.D. VALENTINE
NIRAJ PATEL
S. GLENN BAKER
ANGELA FAGERLIN
Affect and Decision Making
 Many ways in which affect influences
judgments/decisions



Affect as information (Schwarz & Clore, 1988)
Affect as a heuristic (Slovic et al., 2007; Slovic & Peters, 2006)
Feelings is for doing (Zeelenberg et al., 2008)
 People often make judgments/decisions by asking
“How do I feel about it?” (Schwarz & Clore, 1988)
Feelings vs. Beliefs
 What happens when feelings and beliefs obviously
conflict?


Feelings say “do something!”
Beliefs (and objective fact) say “don’t do that—it’s not going to
work”
 A real life case...
The case of prostate cancer screening
 PSA screening for prostate cancer was broadly
recommended for years
 Now, it is not strictly recommended at all



United States Preventive Services Task Force recommends
against the test
American Cancer Society recommends that men may choose to
get test only after they are informed
Evidence suggests it may save no lives, and causes harm
 There is no replacement test
https://www.harding-center.mpg.de/en
https://www.uspreventiveservicestaskforce.org
Theoretical Perspectives
 Feeling: Cancer is scary!
 Belief: There’s nothing I can do!
 Health behavior models: Protection Motivation
Theory/Extended Parallel Process Model (Rogers,
1975; Witte, 1992):

Fear/anxiety + screening is ineffective = defensive coping
 But...feelings can sometimes influence decisions
directly, independent of beliefs:

Fear/anxiety + screening is ineffective = do it anyway
Observed disconnect between beliefs and
screening preferences
 After being informed about revised recommendation
against prostate cancer screening...(Squiers et al., 2013)


62% of men agreed with recommendation to not get screened
only 13% planned to follow recommendation
 People want screening under other dubious conditions



E.g. 1/2 want test to detect benign cancers, or “pseudo-disease”
E.g. 2/3rds want test when nothing could be done about the cancer
(Schwartz, Woloshin, Fowler & Welch, 2004; Waller, Osborne &
Wardle, 2016)
 What is going on here?


A: Perhaps they still believe there is a chance for benefit
A: Perhaps they understand the low chance of benefit, but feelings—
e.g. anxiety—drive desire to do something to prevent cancer.
The Present Research
 Would anyone want a cancer screening test that
unambiguously had no mortality benefit whatsoever?

Make it completely clear: This test does not reduce the chance
of death from cancer or extend the length of life
 If a substantial proportion of people accept such a
test, what is driving their decisions?


Beliefs: Refusal to believe that screening does not affect death
rate
Feelings: e.g., cancer anxiety
A hypothetical test with no benefit
 Posed a hypothetical scenario to nationally
representative sample of U.S. adults



1606 men and women surveyed online
Age: 40-70, M=54, SD=8.6
Demographics matched U.S. population statistics (race/ethnicity,
education, income)
 Funding source: Time Sharing Experiments for the Social
Sciences (TESS) and the National Science Foundation
A hypothetical test with no benefit
 “Imagine that the following is true: There is a new test to
screen for [breast/prostate] cancer that uses diagnostic
technology to look for abnormalities that may be early stage
cancer.
 “However, years of research have unquestionably shown that
the test does not extend life or reduce the chance of
death from [breast/prostate] cancer. Studies have
shown that when comparing two groups of [women/men],
one who received this test yearly and one who never received
the test, both groups had equal numbers of [women/men]
who died from [breast/prostate] cancer. In these studies, the
group of [women/men] who got the screening test had more
cancers diagnosed and treated, but this did not reduce the
number of cancer deaths or extend the length of life.
A hypothetical test with no benefit
 “One problem with the test is that it may detect cell
abnormalities that will never develop into dangerous cancer.
This can lead to unnecessary treatment. The test can also fail
to detect dangerous cancers. Finally, the most dangerous
cancers grow and spread so quickly that they are more likely
to be detected between screenings from symptoms.”
Method
Baseline measures:
1. Worry about getting cancer, 1=not at
all, 7=extremely
2. Perceived lifetime cancer risk, 1=low
risk, 7=high risk
Read about test with no
mortality benefit
No detailed
harms
Detailed list of
screening harms
Outcome measures
Outcome measures
 Critical outcome: If this was the only test that was
available to you, would you want to get this test, or not?
 What we are really asking: Is getting this test
viewed as better than no test at all?
 Other important outcomes:
What did you believe about this test? Did you believe that
this test saves lives, does not save lives, or were you
unsure?
2. In your opinion, how risky is this test? 1-7, not at all riskyvery risky
1.
Results
100%
90%
80%
70%
60%
50%
Did not want the test
40%
Wanted the test
30%
20%
54%
31%
10%
0%
No detailed harms Detailed harms
Beliefs
 Only 55% of respondents believed that the test does
not save lives (891/1606)
Predictive strength of beliefs vs. feelings
Summary
 A surprising proportion of people want screening in
the clear absence of benefit
 Many wanted screening even though they believed
that the test does not save lives
 Anxiety predicted decisions controlling for perceived
benefits and risks

Suggests that emotions create desire for action despite
acknowledged futility
 Points to the need to “just do something” when it
comes to extreme perceived threats; emotions
motivate action (e.g. Zeelenberg et al., 2008)
Thank you
 Contact: [email protected]
 This study was supported by Time Sharing
Experiments for the Social Sciences (TESS) and the
National Science Foundation
PSA screening data
Moyer, V.A., 2012, https://www.uspreventiveservicestaskforce.org