Transcript Patient-Centered Communication in Cancer Care: Pathways to

How Clinician-Patient
Communication Can Improve
Health Outcomes
Richard L. Street, Jr.
Texas A&M University
June 8, 2010
The problem: How does one
explain these findings?

Kaplan et al (1988)
◦ Patient efforts to exert control—lower blood pressure and
lower A1c 8-12 weeks after the consultation
◦ More equal dr-pt floortime—fewer functional limitations 812 weeks after visit

Orth et al (1987)
◦ lower blood pressure at 2 weeks related to:
◦ Proportion of physician talk that was
explaining/describing/reporting—
◦ Frequency of patient talk that expressed concerns/described
symptoms—lower blood pressure at 2 weeks

Stewart et al (2000)
◦ Patient perception of dr patient-centeredness (e.g,
exploring problem, finding common ground)—
better emotional well-being 2 months after visit

Street et al (1993)
◦ Less nurse controlling behavior—better A1c
control at 2 months

Ward et al (2003)
◦ More active patient participation among lupus
pts—less organ damage at three years
But then what about these findings?
Amount of dr. information—more functional
limitations and lower self-reported health (Kaplan et
al, 1988)
 Pt effort to participate in decision-making not
related to subsequent A1c (Rost et al, 1991)
 Dr. patient-centered communication (partnershipbuilding, supportive) not related to lupus patients’
outcomes (Ward et al, 2003)
 Negative affect related to better A1c (Kaplan et al.,
1988) and poorer A1c (Street et al, 1993)

Or these?

Observer coded PCC (exploring pt. problem, finding
common ground) (Stewart et al., 2000):
◦ was not related to pt’s emotional well-being
◦ only modestly correlated with pt. perception of PCC

Active pt. participation in deciding breast cancer
treatment (Street et al, 1995):
◦ did not predict breast cancer patients’ emotional and
functional well-being at 1 year following treatment
◦ did predict patient assuming responsibility for decision at 1
year which in turn predicted well-being

PCC trained doctors and diabetes outcomes
(Kinmonth et al, 1998)
◦ With trained drs, patients were happier and fatter
Theoretical pathways
Clinician-Patient
Patient Health
Communication
Outcomes
Conceptual/Measurement
challenges
7
Patient-Centered COMMUNICATION:
Six overlapping functions
Epstein MR and Street RL. Patient-centered communication in cancer care:
Promoting healing and reducing suffering. NCI, NIH publication #07-6225, Bethesda MD, 2007
8
http://www.outcomes.cancer.gov/areas/pcc/communication
Proximal Outcomes
*understanding
*satisfaction
*clinician-patient
agreement
*trust
*feeling ‘known’
*patient feels involved
*rapport
*motivation
Indirect
(mediated)
path
Intermediate Outcomes
*access to care
*quality medical decision
*commitment to
treatment
*trust in system
*social support
*self-care skills
*emotional management
Clinician-Patient
Communication
Processes
Health outcomes
Direct path
*survival
*cure/remission
*less suffering
*emotional well-being
*pain control
*functional ability
*vitality





Identify the health outcome of interest
Identify the mechanism for improved health
Model the pathway through which
communication can lead to improved health
Select appropriate measures for
communication variables, proximal outcomes,
and intermediate outcomes
Develop intervention to target
communication process to activate that
mechanism


Outcome: Better pain control at 2 weeks
Pathway:
◦ Effective use of pain medication

Proximal outcomes
◦ Change in pain medication (new medication, change in
dose)

Communication variables
◦ Active patient participation about pain (asking
questions, expressing concern, being assertive)

Intervention: patient activation coaching intervention
◦ Emphasize importance of patient involvement, educate
about pain management, provide opportunities for practice
talking to the doctor and feedback on performance
Coaching
intervention
Active
Patient
Participation
Change in
pain meds
Pain control
at 2 weeks
Research settings
◦ Patient of UC Davis Cancer Center, Kaiser
Permanente oncology clinics in Sacramento and
Roseville, Sacramento VA Health Care System
 Research participants
◦ Age 18-80
◦ Cancers: breast, prostate, lung, head/neck, pancreatic,
colorectal, esophageal
◦ Worst pain of 4 or greater or worst pain of 3 with
impairment of daily activities
◦ Not in hospice
◦ No more than one pain management consultation


Frequency measures (Street, 2001)
◦ Patient involvement-pain specific (frequency of patients’
questions, assertive statements, and concerns specific
about pain) overall and pain-specific

Ratings measures
◦ Coders’ ratings of participatory decision-making
(Kaplan et al, 1995)

Change in pain medication
◦ Patient self-report to question, “During the visit you just
completed, did the physician recommended any change
in your pain medication? “ (Yes--new medicine,;Yes-change in dose or amount of a medicine; No)
Intervention
Control
P
N
77
71
Mean age (yrs)
60
57
<.03
% women
78%
84%
ns
% white
67%
75%
ns
% college degree
37%
46%
ns
Predictors of Active Patient Participation (Street et al., 2010)
Total Patient
Participation
(N = 148)
Predictor
P
Estimate
P
(SE)
0.88 (1.43)
.54
(SE)
-0.27 (0.61)
.65
Patient’s age
-0.20 (0.08)
.01
-0.03 (0.03)
.29
Patient’s education = HS or
-1.59 (1.48)
.29
0.38 (0.64)
.56
0.71 (1.75)
.69
0.66 (0.76)
.39
Patient baseline pain
1.01 (0.36)
.005
0.48 (0.15)
.002
Physician participatory
0.26 (0.09)
.006
0.13 (0.040
.001
0.15 (1.60)
.93
0.36 (0.69)
.60
0.54 (1.32)
.68
1.51 (0.57)
0.009
Patient race = Caucasian
Estimate
Pain-Specific
Participation
(N = 148)
(ref = non-Caucasian)
less
(ref = some college plus)
Patient gender = female
(ref = male)
decision-making
Accompanied = yes
(ref = no)
Education session =
(ref = control)
Predictors of Pain Medication Adjustment
Predictor
Model 1
Model 2
(N = 134)
(N = 134)
P
Estimate
P
(SE)
-0.03 (0.09)
.ns
(SE)
-0.01 (0.09)
ns
Patient’s age
-0.01 (0.00)
.02
-0.01 (0.00)
.02
Patient’s education = HS or less
-0.01 (0.10)
.29
-0.05(0.09)
ns
-0.05 (0.11)
ns
-0.11 (0.11)
ns
Patient baseline pain
0.03 (0.03)
ns
0.01 (0.02)
ns
Accompanied = yes
-0.03 (0.11)
ns
0.36 (0.69)
.60
0.27 (0.9)
0.003
0.19 (0.08)
0.04
0.06 (0.01)
<.0001
Patient race = Caucasian
Estimate
(ref = non-Caucasian)
(ref = some college plus)
Patient gender = female
(ref = male)
(ref = no)
Education session =
(ref = control)
Active patient participation
(pain-specific)
Coaching
intervention
Active
Patient
Participation
Change in
pain meds
Pain control
at 2 weeks
_
Baseline
pain

Outcome: Better pain control and function

Pathway
◦ Placebo effect through beliefs about acupuncture

Proximal outcomes
◦ Patient expectations that acupuncture will improve pain
and function

Communication Intervention
◦ Acupuncturists communicating high vs. neutral
expectations for treatment success
The Houston Arthritis and Acupuncture Treatment Study (HAATS)
(PI, M. Suarez-Almazor)
For the first half of the trial half of the acupuncturists were randomly allocated to the high expectations
communication style and the other half to the neutral style. In the second half the acupuncturists
switched communication styles.

High expectations
◦ “I’ve had a lot of success with patients with your kind of
knee pain”
◦ “I’m optimistic this is going to work for you.”
◦ “You should start feeling better within 4 or 5
treatments”
◦ “Sometimes it just takes a little longer to work, but it
usually does”

Neutral expectations
◦ “It works for some patients and not for others.”
“I am not sure if it’ll work. We just have to wait and see
if it helps you”
◦ “It may be that it’s not working, maybe it will after a few
more treatments”
Communication expectation measure

◦ Coders’ rating on 100mm scale at baseline visit
◦ Based on what you heard, rate this practitioner’s communication style when
discussing treatment expectations for improvement.
◦ Neutral ___________________________________High
Patient expectations (assessed at 4 weeks)

◦
For each condition (e.g., knee pain, stiffness), please choose how you would
expect it to be 2 MONTHS from now after you have completed the
Acupuncture Treatment. (Much worse, Worse, The same. Better, Much Better)
Pain and Function measures

◦
◦
◦
WOMAC scale response at + 6 weeks and + 3 months
Pain—How much pain do you have ‘walking on a flat surface,’ ‘going and down
stairs,’ etc
Function—What degree of difficulty do you have with ‘ascending stairs,’ ‘getting
in/out of car,’
Baseline demographic and clinical characteristics
High Expectations
TCA
Sham
N=75
N=151
Neutral Expectations
TCA
Sham
N=78
N=151
p+
GENDER, females – n (%)
AGE
ETHNICITY – n (%) Whites
51 (68.0%) 97 (64.2%) 50 (64.1%) 98 (64.9%)
63.5 (10.4) 65.3 (9.0)
65.5 (7.8)
63.7 (9.1)
52 (69.3%) 103 (68.2%) 55 (70.5%) 101 (66.9%)
African American
16 (21.3%)
23 (15.2%)
13 (16.7%)
31 (20.5%)
4 (5.3%)
14 (9.3%)
8 (10.3%)
14 (9.3%)
11 (7.3%)
2 (2.6%)
5 (3.3%)
5 (3.3%)
79 (52.3%)
67 (44.4%)
3 (3.8%)
35 (44.9%)
40 (51.3%)
3 (2.0%)
72 (48.0%)
75 (50%)
> 0.20
10.0 (11.7)
8.4 (9.6)
8.4 (7.9)
8.8 (10.4)
0.14
4.4 (1.2)
4.3 (1.3)
4.5 (1.3)
4.6 (1.2)
0.12
WOMAC Pain
43.3 (18.2)
44.8 (18.7)
45.6 (18.6)
45.2 (17.8)
> 0.20
WOMAC Function
41.5 (19.3)
45.1 (18.5)
44.2 (18.8)
44.1 (17.6)
> 0.20
Hispanic
Other
3 (4.0%)
EDUCATIONAL LEVEL – n (%)
Less than high school 2 (2.7%)
High School diploma/some 34 (45.3%)
college 39 (52.0%)
College degree
DURATION OF KNEE PAIN,
years
J-MAP
> 0.20
> 0.20
> 0.20
Results of Path Analysis
Expectations
Communicated
.
11
(baseline)
Patient
Baseline
Expectations
Patient
Expectations
Acupuncture
Effectiveness
.37
-.32
Pain
6wk
(4 weeks)
All effects were significant, P < .05
good overall model fit: chi-square=5.2,
p=.39, RMSEA=.01
.77
Pain
3 months
Another pathway? Shared Mind
Epstein and Peters (2009)
“Collaborative cognition depends on the
physician being mindful not only of the
patient’s values, thoughts, and feelings but also
his or her own. Research exploring shared
deliberation and shared mind must bridge
cognitive science, decision research, and
communication skills training and evaluate
communication processes as well as patients’
experience of care” (p. 197).


Shared mind is a process:
◦ A way of thinking
◦ A way of talking
◦ A way of collaborating

Shared mind is also an outcome
◦ Shared understanding of the problem
◦ Understanding the others’ perspective
◦ Reaching agreement on a course of action
Types of shared understanding

Similar beliefs
◦ What physician believes coincides with what
patient believes

Perceived agreement (fantasy)
◦ What physician believes coincides with what
physician believes patient believes

Understanding of the other (perspectivetaking)
◦ What physician believes the patient believes
coincides with what the patient believes
PT
DR
The CONNECT study (PI, P. Haidet)
Determine how well physicians understand
their patient’s health belief models
 Identify predictors of greater understanding
(more accurate perspective taking)

◦ Communication factors—more active patient
participation (asking questions, expressing
concerns, stating opinion and preferences)
◦ Relationship factors—number of previous visits,
racial/gender concordance
◦ Cultural factors—as assessed by physician-patient
demographics
The CONNECT instrument: Six domains
of health beliefs (*Haidet et al, Patient Education and
Counseling 2008)
To what extent:
 is there a biological cause to the patient’s health
condition
 is the patient at fault for his/her health condition
 does the patient have control over his/her health
condition
 does the condition have meaning for the patient
 can the patient’s health condition benefit from
natural/alternative remedies
 does patient want a partnership with the doctor
 Study
Design: cross-sectional using
pre- and post-visit surveys, audiotapes
of primary care encounters
 Setting: 10 primary care clinics in
Houston, TX
 Subjects: 272 adult patients receiving
care from 29 physicians
The CONNECT Instrument*

Six domains of explanatory models (cause, fault, control,
meaning, alternative treatments, relationship)
Respondent/Focus
Instrument Item
Patient Own:
“I have control over my illness”
Doctor own:
“The patient has control over their illness”
Doc Understanding:
The patient thinks he/she has control over their
illness”
*Haidet et al, Patient Education and Counseling 2008
Data Analysis
Summary scores for each CONNECT
domain
 Compared domain sum scores between
patient own and physician understanding
 Used multivariate models to examine
predictors of greater physician
udnerstanding
 Accounted for effect of patients nested
within physicians

Results: Study Population
Patients (n=272)
Physicians (n=29)
Mean age, yrs
56 (15)
Mean age, yrs
43 (9)
Female (%)
60
Female (%)
58
AA (%)
50
AA (%)
26
Caucasian (%)
39
Caucasian (%)
32
Hispanic (%)
11
Hispanic (%)
3
Asian (%)
39
Differences between patients’ model , physician models, and physician
beliefs about the patients’ models
A
B
C
Physician
Physician Belief
Patient
Difference
Difference
Model
about Patient
Model
A vs. B
B vs. C
(P-value)
(P-value)
Domain
Model
Cause-Biological
71.4
74.4
67.4
.059
0.009
Patient at Fault
49.1
49.1
55.0
ns
<0.001
Patient has Control
78.4
75.8
82.8
0.018
<0.001
Meaning to the Patient
71.0
72.1
79.6
ns
<0.001
50.7
50.8
56.7
ns
<0.001
82.5
81.3
89.7
ns
<0.001
Value of Alternative
Treatments
Patient wants to be
Partner in Relationship
Associations With Better Physician
Understanding*

Greater Patient Participation
◦ Meaning, Control, Alternative Treatments,
Relationship (p = 0.01-0.05)

Race Concordance
◦ Control (p = 0.02)

Lower Patient Education
◦ Control (p = 0.008)

Male Physician
◦ Control, Relationship (p = 0.004, 0.05)
*Lower absolute difference between
patient own and physician understanding score
Association With Poorer Physician
Understanding*

African American Patients (compared w/
Caucasian)
◦ Relationship (p = 0.02)

Hispanic patients (compared with
Caucasian)
◦ Meaning (p = O.06)
*Greater absolute difference between
patient own and physician understanding score
Conclusion
Significant gaps in physician understanding
of patient perspectives in routine primary
care settings
 Certain demographic factors may impact
physician understanding
 Active patient participation improved
understanding in multiple domains

Questions?
References

Epstein RM, Street RL, Jr. Patient-Centered Communication in Cancer Care: Promoting
Healing and Reducing Suffering. Bethesda, MD: National Cancer Institute; 2007. Report No.:
NIH Publication No. 07-6225.

P. Haidet, K.J. O’Malley, B.F. Sharf, A.P. Gladney, A.J. Greisinger, & R.L. Street Jr. (2008)
Characterizing explanatory models of illness in healthcare: Development and validation of
the CONNECT instrument. Patient Education and Counseling, 73, 232-239.

Kaplan SH, Greenfield S, Ware JE, Jr. Assessing the effects of physician-patient interactions
on the outcomes of chronic disease. Med Care 1989; 27:S110-S127.

Street, R. L., Jr. (2001). Active patients as powerful communicators. In W. P. Robinson, & H.
Giles (Eds.), The new handbook of language and social psychology (pp.541-560). New York:
John Wiley & Sons.

R.L. Street, Jr., C. Slee, D. K. Kalauokalani,D.E. Dean, D. J. Tancredi, & R. L. Kravitz (2010)
press) Improving physician-patient communication about cancer pain with a tailored
education-coaching intervention. Patient Education and Counseling, 80, 42-47.

M.M. Ward, S. Sundaramurthy, D. Lotstein, T. Bush, C.M. Neuwelt, & R.L. Street, Jr. (2003).
Participatory patient-physician communication and morbidity in patients with systemic
lupus erythematosus. Arthritis & Rheumatism, 49, 810-818.

HS Gordon, RL Street, Jr., BF Sharf, & J Souchek. (2006) Racial differences in doctors’
information-giving and patients’ participation. Cancer, 106, 1313-1320.

Street, R. L., Jr. (2001). Active patients as powerful communicators. In W. P. Robinson, & H.
Giles (Eds.), The new handbook of language and social psychology (pp.541-560). New York:
John Wiley & Sons.

R.L. Street, Jr, M.N. Richardson,V. Cox, and M.E. Suarez Almazor. (2009)
(Mis)Understanding in patient-health care provider communication about total knee
replacement. Arthritis Care and Research, 61, 100-107.