Mispredicting adaptation to adverse outcomes: New evidence from the medical domain

1
(Mis)predicting adaptation to adverse outcomes
New evidence from the medical domain
George Loewenstein(presentation at HDGC 1/22/03)

• Collaborators (partial list)
• Peter Ubel, M.D.
• John Hershey, Ph.D.
• Jonathan Baron, Ph.D.
• David A. Asch, M.D., M.B.A.
• Christopher Jepson, Ph.D.
• Angela Fagerlin, Ph.D.
• Julie Lucas, B.B.A.
• Jason Riis

2
Adaptation

• Material (behavioral)
• Hedonic

3
Predictions of adaptation

• General finding people underpredict their own
  speed of adaptation (both negative and positive)

• Loewenstein Frederick, 1997 (diverse, including
  income)
• Gilbert et al. 1998 (e.g., tenure)
• Schkade Kahneman, 1998 (living in Cal.)
• Sieff, Dawes Loewenstein, 1999 (reaction to HIV
  status)
• Wilson et al, 2000 (win or loss of team)

4
Application to the medical domain

• (Which hopefully sheds light more broadly on
  adaptation, and the accuracy of intuitions about
  adaptation, in diverse domains)

5
Most patients report a high quality of life

• Brickman, Coates, and Janoff-Bulman (1978)
  Surprisingly small difference in self-reported
  happiness (on 5 point scale) between paraplegics
  and matched controls
• paraplegics 2.96
• controls 3.82
• Wortman and Silver (1987) quadriplegics reported
  no greater frequency of negative affect than
  control respondents!
• Tyc (1992) no difference in quality of life or
  psychiatric symptomatology in young patients who
  had lost limbs to cancer compared with those who
  had not.

6
Non-patients dont expect patients to be as happy
as they report being..

• Discrepancy between patients evaluations of
  their own quality of life and non-patients
  evaluations of what their quality of life would
  be if they had the same health conditions
• Chronic dialysis (Sackett and Torrance, 1978)
• Nonpatient predictions .39
• Patient reports .56
• Colostomies
• Nonpatient predictions .80
• Patient reports .92
• ? The discrepancy

7
Many possible causes of the discrepancy

• Explanations that implicate non-patients
• Misconstrual of medical condition?
• 'Focusing illusion'
• Underappreciation of adaptation
• Explanations that implicate patients
• Renorming of scales
• Dissonance reduction
• 'Neutral' explanations
• Mismatch between subject populations?

8
Whether discrepancy is important for medical
policy depends on its cause

• Attempts to rationalize health care delivery
• Nonpatients evaluations of QOL serve as inputs
• Informed consent/ patient decision making
• Individual treatment decisions often based on
  perceptions, by people who do not have
  conditions, of what it would be like to have
  those conditions

9
An illustration

• Slevin et al., 1990 who say they'd accept a
  grueling course of chemotherapy for 3 extra
  months of life
• radiotherapists 0
• oncologists 6
• healthy persons 10
• current cancer patients 42
• whose values to use?

10
Data!
11
Within-subject study of kidney transplant and
dialysis (unpublished)

• (n127 dialysis patients who ultimately received
  transplants all numbers on 0-100 quality of life
  scale)
• Reported well-being
• pre-transplant 64.16
• Predicted well-being
• one year later 91.19
• Reported well-being
• one year later 76.81
• Recalled well-being 47.19
• Notes - all means significantly different from
  one-another
• - those not transplanted over-predicted their
  own misery

12
Evidence of misconstrual
13
bad scales?

• Classic criticism is that patients renorm the
  scales based on their own experiences or on new
  points of social comparison
• But when sufferers and nonsufferers of diverse
  problems rated QoL with anchored or unanchored
  scales, anchored scales produced larger
  discrepancies
• Baron et al., Effect of assessment method on the
  discrepancy between judgments of health disorders
  people have and do not have.

14
Study 2

• Web-based n99 (ages 16-68 median 36 22 male)
• Rated series of health conditions
• With vague or better-defined scale
• Vague e.g., "100 is a very good quality of
  life"
• Better-defined e.g., "100 is as good as that of
  someone with a meaningful job, friends, family,
  and good health"
• For self or other
• Then stated whether they had the condition

• Conditions
• Asthma
• Back pain
• Insomnia
• Shortness
• Overweight
• Nearsightedness
• Acne
• Smoking habit
• Arthritis
• Heart disease

15
Study 2 results

• Self-ratings consistently higher than other
  ratings
• Have/have not discrepancy was larger with
  better-defined scale than with vague scale

16
Self-deception by patients?

• Jason Riis et al. (in progress)
• Palm Pilots given to 60 end stage renal patients
  dialysis 3 times per week.
• 28 matched (age, gender, educ., race) healthy
  controls
• Palms carried for 7 days beeped randomly in each
  90 minute segment of day
• On each beep, respondent asked 12 questions,
  including

17
Please tap the button below that best describes
the mood you were feeling just before the Palm
Pilot beeped

• 2 Very pleasant
• 1 Slightly pleasant
• 0 Neutral
• -1 Slightly unpleasant
• -2 Very unpleasant

• When Palms returned, subjects estimated mood
  distributions on the above scale
• Last Week (during which they carried the palm)
• Typical Week
• Dialysis Scenario
• Controls (Following presentation of a dialysis
  scenario "Imagine that you had dialysis")
• Patients As in the scenario no other health
  problems.
• Other Person (Someone else your age with same
  health)
• Healthy
• Controls In perfect health
• Patients If never had kidney trouble

18
Main results
19
Conclusions so far..

• Discrepancy not due to
• mismatch between populations
• scale renorming
• patient misrepresentation (to self or other)
• Misconstrual may contribute

20
Mispredictions by nonpatients?

• focusing illusion (Kahneman Schkade Wilson,
  Gilbert et al.)
• underprediction of adaptation

21
Tests of focusing illusion

• Subjects in all studies were prospective
  Philadelphia jurors

22
First defocusing task life domains

• How much do you think having a below-the-knee
  amputation would affect
• Your overall health?
• Your standard of living?
• Your work?
• Your love life?
• Your family life?
• Your social life?
• Your spiritual side of your life?
• Your leisure activities, such as hobbies,
  pastimes, travel, and entertainment?

23
Disability Ratings Before and After Defocusing
Exercise

• QoL Rating (0 - 100)
• Disability N Before After P
• Paraplegia 53
• Below-knee 52
• amputation

58.5 78.1
51.8 72.3
0.02 0.01
24
Second defocusing task concrete events

• If you had below the knee amputation/paraplegia,
  what would your experience of these things be
  like compared to now?

• visiting with friends and/or family
• paying bills and taxes
• vacation and travel
• getting caught in traffic
• physical recreational activities
• arguing with family and/or friends
• reading and/or watching TV or movies
• coping with death and/or illness in the family

25
Concrete Events Defocusing Results

• QoL Rating (0 - 100)
• Disability N Before After P
• Paraplegia 50
• Paraplegia 51
• BKA 51
• BKA 51

55 - 71 -
51 45 72 67
.41 .05 .27 .34
26
Third defocusing task time weighted

• Think about the past day, starting from when you
  woke up yesterday to when you woke up this
  morning. What did you do yesterday? In the
  spaces provided, we would like you to list the
  things that took up the most amount of time from
  yesterday when you woke up to today when you woke
  up.
• Subjects were asked to imagine how these five
  activities would be affected if they had the
  disability in question.

27
Time Weighted Defocusing Results

• QoL Rating (0 - 100)
• Disability N Before After P
• Paraplegia 57
• Paraplegia 60
• BKA 53
• BKA 54

.59 .23 .60 .08
51 - 75 -
50 45 74 67
28
Fourth Defocusing Task Changes for Better or
Worse

• To get subjects to think more broadly about
  disabilities
• We asked them to think about aspects of their
  live that would probably
• change for the better
• be unchanged
• change for the worse

29
Changes Results

• QoL Rating (0 - 100)
• Disability N Before After P

Paraplegia 105 53 55
.09 Paraplegia 103 -- 57
.46 BKA 117 75 75
.31 BKA 106 -- 73
.29
30
Are Disability Ratings Influenced by Failure to
Consider Adaptation?

• Adaptation exercise
• Think about one emotionally difficult life
  experience that happened to you at least 6 months
  prior to now
• At the end of those 6 months would say you felt
• Much worse
• About the same
• Much better than you would have predicted

31
Adaptation Results

• QoL Rating (0 - 100)
• Disability N Before After P
• Paraplegia 123
• Paraplegia 56

.003 .001
47 -
52 62
32
Should we not care about environmental change (or
forget about road safety)?

• knowledge of these results has little effect on
  willingness to pay, etc.. (we may not understand
  why, but there may be a good reason)
• happiness/quality of life matters, but doesn't
  include everything we care about..
• quantity and quality of well-being (Skorupski)
• children
• hitchhiking
• mountaineering