Transcript Provider Prejudice and Stereotyping in Clinical Decision

Health care disparities
Stereotyping and
unconscious bias
Harry Pomeranz
Mercy College
October 2008
Do you think the average African American is
better off, worse off, or just about as well off as
the average white person in terms of access to
health care?
70
60
50
Better Off
About the Same
Worse Off
No Opinion
40
30
20
10
0
Whites
Source: Morin, 2001
African Americans
Black and White Differences in Specialty
Procedure Utilization Among Medicare
Beneficiaries Age 65 and Older, 1993
Black White
Angioplasty
(procedures per 1,000 beneficiaries per
year)
Coronary Artery Bypass Graft Surgery
(procedures per 1,000 beneficiaries per
year)
Mammography
(procedures per 100 women per year)
Hip Fracture Repair
(procedures per 100 women per year)
Amputation of All or Part of Limb
(procedures per 1,000 beneficiaries per
year)
Bilateral Orchiectomy
(procedures per 1,000 beneficiaries per
year)
Source: Gornick et al., 1996
Black-toWhite Ratio
2.5
5.4
0.46
1.9
4.8
0.40
17.1
26.0
0.66
2.9
7.0
0.42
6.7
1.9
3.64
2.0
0.8
2.45
What are potential sources of
disparities in care?
Health systems-level factors – financing,
structure of care; cultural and linguistic
barriers
 Patient-level factors – including patient
preferences, refusal of treatment, poor
adherence, biological differences
 Disparities arising from the clinical
encounter

Potential Sources of Racial and Ethnic
Healthcare Disparities –
Healthcare Systems-level Factors
Lack of stable relationships with
primary care providers –-- minority
patients, even when insured at the
same level as whites, are more likely
to receive care in emergency rooms
and have less access to private
physicians

Disparities in the Clinical Encounter: The
Core Paradox
How could well-meaning and highly
educated health professionals,
working in their usual circumstances
with diverse populations of patients,
create a pattern of care that appears
to be discriminatory?
Perceptions of Disparities in
Health Care
Figure 18
Generally speaking, how often do you think our health care system treats
people unfairly based on…
Percent Saying “Very/Somewhat Often”
Doctors
72%
70%
Whether or not they have insurance
How well they speak English
The Public
43%
58%
What their race or ethnic
background is
29%
Whether they are male or female
15%
27%
47%
Source: Kaiser Family Foundation, National Survey of Physicians, March 2002 (conducted March-October 2001);
Survey of Race, Ethnicity and Medical Care: Public Perceptions and Experiences, October 1999 (Conducted July –
Sept., 1999)
Disparities in the Clinical Encounter: The
Core Paradox

Uncertainty – a plausible
hypothesis, particularly when
providers treat patients that are
dissimilar in cultural or linguistic
background
Disparities in the Clinical Encounter: The
Core Paradox

Stereotyping – evidence suggests
that physicians, like everyone else,
use these ‘cognitive shortcuts’
Stereotyping: A Definition

Stereotyping can be defined as the
process by which people use social
categories (e.g. race, sex) in acquiring,
processing, and recalling information
about others.
Stereotyping: A Definition

Stereotyping beliefs may serve
important functions - organizing and
simplifying complex situations and
giving people greater confidence in
their ability to understand, predict, and
potentially control situations and
people.
Stereotyping: Risks
Can exert powerful effects on
thinking and actions at an implicit,
unconscious level, even among
well-meaning, well-educated
persons who are not overtly biased.
 Can influence how information is
processed and recalled.

Stereotyping: Risks

Can exert “self-fulfilling” effects, as
patients’ behavior may be affected
by providers’ overt or subtle
attitudes and behaviors.
Stereotyping: When Is It in Action?

Situations characterized by time
pressure, resource constraints, and
high cognitive demand promote
stereotyping due to the need for
cognitive ‘shortcuts’ and lack of full
information.
What is the Evidence that Physician Biases and
Stereotypes May Influence the Clinical Encounter?
study conducted in actual clinical
settings found that doctors are more
likely to ascribe negative racial
stereotypes to their minority
patients.

These
stereotypes were ascribed to
patients even when differences in
minority and non-minority patients’
education, income, and personality
characteristics were considered.
van Ryn and Burke (2000)
What is the Evidence that Physician Biases and
Stereotypes may Influence the Clinical Encounter?
medical
students were more likely
to evaluate a white male “patient”
with symptoms of cardiac disease as
having “definite” or “probable”
angina, relative to a black female
“patient” with objectively similar
symptoms. Rathore et al. (2000)
What is the Evidence that Physician Biases and
Stereotypes may Influence the Clinical Encounter?
mental
health professionals and
trainees were more likely to evaluate
a hypothetical patient more
negatively after being “primed” with
words associated with African
American stereotypes. Abreu (1999)
The Elimination of Health Care
Disparities

In 2002, the Institute of Medicine (IOM)
published Unequal Treatment: Confronting
Racial and Ethnic Disparities in Health
Care,

it reported that racial and ethnic
minorities experience a lower quality
of health care than non-minorities,
even when the patient's insurance
status and income are controlled.
The Elimination of Health Care
Disparities

The study committee also found evidence
that stereotyping, biases, and
uncertainty on the part of health care
providers contribute to unequal
treatment.
The Elimination of Health Care
Disparities

Clinicians may order fewer tests for
racial and ethnic minorities if they do
not understand the patient's
description of symptoms.
The Elimination of Health Care
Disparities

Alternatively, clinicians may order
more diagnostic tests to compensate
for not understanding what their
patients are saying.
Race was noted in 16 of 18 case
presentations by residents, but only 19 of
36 cases involving white patients.
 Race was mentioned in 10 of 10 cases
when the resident described black
patient's unflattering characteristics, but
only four of nine cases where the resident
described unflattering characteristics in
white patients.

African American patients were viewed by
physicians as less intelligent, less
educated, less likely to comply with their
advice and more likely to have problems
with alcohol and drugs.
 Physicians also rated African American
patients as less likely to be the kind of
person whom the physician could have as
a friend.


Using pain-management vignettes in
patients who differed only in race, male
physicians prescribed higher doses of
hydrocodone to whites than to blacks,
while female physicians did the opposite
“Implicit Bias” and
“Unconscious Stereotyping”
Research indicates:
 Implicit biases are pervasive.
 People are often unaware of their implicit
biases
 Ordinary people harbor negative
associations in relation to various groups
“Implicit Bias” and
“Unconscious Stereotyping”
Implicit biases predict behavior
 People differ in levels of implicit bias

Implicit Bias and Clinical Outcomes

Physicians reported no explicit preference
for white versus black patients

Implicit Association Test (IAT) revealed
implicit preference favoring white
Americans
Implicit Bias and Clinical Outcomes

IAT revealed implicit stereotypes of black
Americans as less cooperative with medical
procedures and less cooperative generally

As physicians’ pro-white implicit bias increased,
so did their likelihood of treating white patients
and not treating black patients with thrombolysis
Dual Process Stereotyping
Two distinct methods of stereotyping:
1.
Automatic stereotyping
2.
Goal modified stereotyping
Burgess and van Ryn: Understanding the provider contribution to
race/ethnicity disparities in pain treatment; Pain Med. 2006
Automatic Stereotyping

occurs when stereotypes are automatically
activated and influence judgments/behaviors
outside of consciousness

Occur regardless of their relevance to the
perceivers’ goals
Burgess and van Ryn: Understanding the provider contribution to
race/ethnicity disparities in pain treatment; Pain Med. 2006
Goal Modified Stereotyping

More conscious process, done when
specific needs of clinician arise (time
constraints, filling in gaps in information
needed to make complex decisions
Burgess and van Ryn: Understanding the provider contribution to
race/ethnicity disparities in pain treatment; Pain Med. 2006
Function of Stereotyping

Providers are likely to apply information contained in
racial/ethnic stereotypes to interpret symptoms and
make decisions

Stereotypes likely to be used when stereotypic
information is perceived as clinically relevant, and the
decision is complex
Burgess and van Ryn: Understanding the provider contribution to
race/ethnicity disparities in pain treatment; Pain Med. 2006
Web and Other Resources
“Heads Up!” Website:
http://www.stop-disparities.org/RESOURCES.html
Web and Other Resources
Implicit Association Test:
https://implicit.harvard.edu/implicit/
Project Implicit Information Page:
http://projectimplicit.net/
(Recommended Tests: Race, Arab-Muslim, Gender, Sexuality)
"The Police Officer's Dilemma"
http://home.uchicago.edu/~jcorrell/TPOD.html
and then click on the very bottom link
http://backhand.uchicago.edu/Center/ShooterEffect/