Building a Behavioral Research Program in a Medical Setting CHIP - - PowerPoint PPT Presentation

Building a Behavioral Research Program in a Medical Setting

CHIP Lecture Series University of Connecticut October 12, 2006

Terrance L. Albrecht, PhD Program Leader, Communication & Behavioral Oncology Scientific Director, Behavioral/Field Research Core Professor, Dept. of Family Medicine WSU School of Medicine

“Somewhere in America today, someone will be told they have cancer…”

– This will be repeated 3,500 times throughout a standard work day.

• the equivalent of more than seven new cancer

diagnoses every minute spent at the office.

Factors to Consider in Building a BehavioralResearch Program

• 1. Context:

– Community, Economic Environment – Medical Setting (Primary Care? Specialty Care?) – If cancer institute, what type? Expectations for social and behavioral research?

• 2. Expectations for Scientific Collaborations: Intra-/Inter-

/Programmatic

– “transdisciplinary” – “translational” – “community-based participatory research” – “primary/secondary prevention” – “treatment/clinical trials” (phase 1,2,3 testing of therapeutic agents/procedures) – “diagnostic techniques/innovation” (e.g., imaging) – “survivorship”

1. In Our Case: The Urban Context of Detroit Population 911,402

(see Dying Before Their Time) African American 81.6% Persons below poverty level (%) 26.1% under age 18 (34.5%)

• ver age 65

(18.6%) Illiteracy rate 47.0% Children born to single mothers 72.0%

Older African Americans in Detroit (compared to Whites):

• Colorectal, lung, prostate and breast cancers:

– Incidence rates higher than national SEER data – Initial diagnosis at advanced stage – Higher mortality rates

• Colorectal, lung and prostate cancers:

– Higher incidence rates for tri-county area

• Colorectal, breast and lung cancers:

– Lower 5-year survival rates for same site/stage at diagnosis

• All sites

347.3 245.5 1.4

• Prostate

70.4 28.8 2.4

• Larynx

5.4 2.3 2.3

• Stomach

13.3 5.8 2.3

• Myeloma

9.0 4.4 2.0

• Oral cavity and pharynx

7.5 3.9 1.9

• Esophagus

11.7 7.4 1.6

• Liver and intrahepatic bile duct

9.3 6.1 1.5

• Small intestine

0.7 0.5 1.4

• Colon and rectum

34.3 24.8 1.4

• Lung and bronchus

104.1 76.6 1.4

• Pancreas

16.0 12.0 1.3

Site

Cancer Sites in which African American Death Rates exceed White Death Rates among U.S. Men (1997-2001)

African American White Ratio of African American/White

Cancer Sites in which African American Death Rates Exceed White Death Rates among U.S. Women (1997-2001)

• All sites

196.5 165.5 1.2

• Myeloma

6.6 2.9 2.3

• Stomach

6.3 2.8 2.3

• Uterine cervix

5.6 2.6 2.2

• Esophagus

3.2 1.7 1.9

• Larynx

0.9 0.5 1.8

• Uterine corpus

6.9 3.9 1.8

• Pancreas

12.8 8.9 1.4

• Colon and rectum

24.5 17.1 1.4

• Liver and intrahepatic bile duct

3.8 2.7 1.4

• Breast

35.4 26.4 1.3

• Urinary bladder

2.9 2.3 1.3

• Oral cavity and pharynx

2.0 1.6 1.3

African American White Ratio of African American/White

Site

2. In Our Case: Collaborations

• Intra—

– Epidemiology – Genetic Epi (linkage studies of disease)

• Inter—

– Developmental Therapeutics – Clinical/Disease sites: Breast, Prostate, Lung, Colorectal, Gastrointestinal, Ovarian – Gerontology – Pediatrics – Symptom Control (e.g., pain, fatigue)

Where We Started: Our Selected Research Areas

I. Medical Communication and Treatment Decision Making II. Pediatric Cancer Coping and Survivorship

• III. Cancer Health Disparities: Unequal Burden

Communication and Treatment Decision Making

• NIH/NCI 7RO1CA75003 Effects of Physician

Communication on Patient Accrual

• NIH/NCI 1R21CA113220-01 Biopsychosocial Processes and

Treatment Decision Making

• Strategic Research Initiative Grant, Karmanos Cancer

Institute The Use of Real Time Patient Data to Define “Clinical

Benefit” in Phase I Clinical Trials.

Sample Publications

• Eggly, et al. (2006). Discussing “bad news” in the outpatient
• ncology clinic: Rethinking current communication guidelines.

Journal of Clinical Oncology.

• Eggly, et al. (in press). Information seeking during “bad news”
• ncology interactions; Question-asking by patients and their
• companions. Social Science and Medicine.

Physician Communication and Patient Accrual to Cancer Clinical Trials Collaborators:

John Ruckdeschel, M.D. Susan Eggly, Ph.D. Louis Penner, Ph.D. Rebecca Cline, Ph.D. Tanina Foster, M. Ed. Felicity Harper, Ph.D. Elizabeth Heath, M.D.

Patient Physician Experimental Treatment Decision Study Protocol Family Member

Simplified Structural Model

• f Patient Decision Making

MD-PT Communication

Data Collection

• Pre-Test (T1)

– Self-reports – Patient/Family Member Sociodemographics

• Physician-Patient-Family Member Interaction (T2)

– Observational Analysis – Real-Time Video Recording of Clinic Encounter

• Post-Test (T3)

– Self-reports – Phone Interviews with Patients Regarding Decisionmaking (1-2 Weeks After Clinic Encounter)

Observational Coding

• N = 241 video recorded interactions at two

comprehensive cancer centers

• Coding System

– Karmanos Accrual Assessment System (KAAS)

• Content checklist (informed consent/informed refusal)
• Ratings of:

– Quality of Relationship MD-PT; MD-FM (e.g., closeness, trust, rapport) – Physician Actions/Behaviors (e.g., talk overs, dominance, discussion of consent)

(“Front Stage; Back Stage”)

Reliability/Validity Issues

Riddle, D.L., Albrecht, T.L., Coovert, M.D., Penner, L.A., Ruckdeschel, J.C.,et al. (2002). Differences in audiotaped versus videotaped physician-patient interactions. Journal of Nonverbal Behavior, 26, 219-240 Albrecht, T. L., Ruckdeschel, J. C., Ray, F.L., et al. (2005) A portable, unobtrusive device for video recording clinical

• interactions. Behavior Research Methods, 37(1) 165-169

Penner, L.A., Orom, H., et al. (in press). Camera-related behaviors during video recorded medical interactions. Journal of Nonverbal Behavior.

Same Physician; Different Protocols Offered…

Altruism-Based Decision (for clinical trial) Spousal Role; Distortion; Resistance (against clinical trial)

Same Physician, Same Protocol

Visit Type (Observed) N (%) CT Not Mentioned 100 (43%) 27 (12%) 48 (20%) 42 (18%) 17 (7%) 234 (100%) CT Mentioned, MD Reject CT Mentioned, PT Reject CT Discussed CT Discussed, Offer Made Other (e.g., Iressa) Total

Clinic Visit Typology

(Karmanos and Moffitt Cancer Centers)

Patient Misperceptions

• Thirty-nine percent of patients who
• nly discussed a trial, said they were
• ffered one
• Fourteen percent of patients who were
• ffered a trial said they were not
• ffered one

Importance of Decision Factors Accrued vs. Not Accrued (T3 report)

Factor

Accrued Did Not Accrue MD IMPACT:

MD Answered Questions**

2.84 (0.46) 1.83 (0.75) FAMILY INFLUENCE: Opinion of FM Present* 2.36 (0.85) 1.40 (0.89)

Trust in MD**

2.78 (0.51) 2.00 (0.82)

MD Recommendation*

2.48 (0.70) 1.50 (0.55) LIFE/SURVIVAL:

Extend Life**

2.74 (0.53) 1.67 (0.52)

Fighting Cancer*

2.69 (0.55) 2.00 (0.82)

MD Supportive*

2.69 (0.63) 1.83 (0.75)

Quality of Life**

2.59 (0.69) 1.57 (0.53) ALTRUISM:

Help Others*

2.46 (0.69) 1.57 (0.53) PERCEIVED COSTS:

Costs Manageable/Insured*

2.19 (0.85) 1.33 (0.82) *p < .05; **p < .001

Observed Quality of Interaction and Importance of Decision Factors (T2T3)

KAAS Relationship Quality Ratings (Observed) Perceived Costs Family Influence Altruism

• .52*
• .54

Hierarchical Rapport

• .49*

Closeness/Connectedness

• .54*
• .44

Trust

• .63**

Observed Physician Behaviors and Importance of Decision Factors

Physician Behaviors Perceived Costs Life /Survival Family Influence MD Directed Toward Consent .64* Amt Info from MD

• .61*

MD Manner

• .55*

MD Data-Based Info

• .60*
• .40*

KAAS Relationship Quality Ratings (Observed) Trust/Confidence in MD Comfort with Decision .44* .54* .41* Hierarchical Rapport Trust

Observed Quality of Interaction and Reactions to MD/Decision

• Obs. Physician Behaviors

Trust/Confidence in MD Comfort with Decision

MD Code .35* .36*

• .65*

.35* MD Talkovers/Interruptions

• .53

MD Conv. Control/Dominance

• .41

MD Data-Based Information Code Match .40* MD Directed Toward Consent .50*

Observed Physician Behaviors and Reactions to Decision/MD

Family Member

• Obs. Quality

Relationship /MD Behaviors Perceived Costs Family Influence Life /Survival Altruism Hierarchical Rapport

• .76**
• .61*

Closeness /Connectedness

• .57*

Md Responsive to Questions/Concerns

• .52*
• .44*

MD Data Based Information

• .44
• .50*
• .45

FM Code .64* Code Match

• .54*
• .63*
• .66*

Family Member

• Obs. Quality

Relationship /MD Behaviors PT Trust /Confidence in MD PT Comfort with Decision Hierarchical Rapport .44* MD Talkovers/Interruptions

• .46*
• .42*

MD Provide Hope .39* .53*

2 2.5 3 3.5 4

Cure No Cure

0.5 1 1.5 2

Yes Maybe No Impression after Interaction Intercept 3.3883 Cure 0.192 p<.05

Do Patient Expectations about Cure Affect Decisions and Patient Impression after Interaction? (T1T3)

Impr. After Inter.

Decision Intercept 1.6434 Cure

• 0.0115

Decision

Patient expectations about cure has no effect on patients’ decision to follow recommended treatment N = 96 Patients who expect a cure report that their interactions with their doctor were more positive. N = 95

0.5 1 1.5 2

4.57 5.285 6.57

Yes Maybe No

Decision

Do Observers’ ratings of Patient/MD interactions moderate the effect of Patient Expectations about Cure

• n Decisions?

Moderation is present: Patients’ who expect a cure are more likely to follow treatment recommendations when MD/PT interactions are judged as being less positive, and are less likely to follow treatment recommendations when MD/PT interaction are judged as more positive. N = 96

Decisions Intercept 1.1391 cure 2.5177 PTMDobs 0.0966 cure*PTMDobs

• 0.4774

p<.05 p<.061

• Obs. Ratings of PT/MD

Future Research: NIH/NCI Roadmap Project*

• Goal:

– To bring together medical and behavioral scientists to study biopsychosocial processes in treatment decision making

• Pilot Projects:

– Identify biobehavioral markers and devices for measuring stress responses – Assess daily diary methods for assessing recall of treatment information, perceived risks, costs, benefits of clinical trials, family influences – Integrate biobehavioral data with verbal/nonverbal observational data – Analyze interdependent data across multiple levels of analysis/time points

*Joint collaboration between Karmanos Cancer Institute and

American College of Surgeons Oncology Group

• NIH/NCI 5R01CA100027-01 Parental Role in Pediatric

Cancer Pain and Survivorship (Albrecht, PI)

• NIH/NICHD U10HD050096 Collaborative Pediatric

Critical Care Research Network Pediatric Intensivist-Parental

Postmortem PICU Interviews: A Randomized Intervention (Meert, PI)

• NIH/NCI R01 The Burdens of Pediatric Cancer and Cancer

Treatment: A Longitudinal Study (Penner & Cline, Co-PIs)

Communication and Pediatric Cancer/Survivorship

Collaborators: Louis A. Penner, PhD Rebecca W. J. Cline, PhD Ellen Jones, MA Amy Peterson, BA Felicity Harper, PhD Jeffery Taub, MD Kathleen Meert, MD Deborah Ellis, PhD John C. Ruckdeschel, MD

• Cline, R., et al. (2006). Parents’ communication patterns during painful

pediatric oncology procedures. Social Science and Medicine.

• Peterson, A., et al. (in press). Parents’ nonverbal immediacy behaviors during

painful pediatric oncology procedures. Journal of Nonverbal Behavior

• Penner, L., et al. (2006). Dispositional and affective correlates of pain among

pediatric cancer patients. (under review).

• Pediatric cancer patients often report that pain

associated with treatment worse than disease.

• Negative reactions to treatment place children at

subsequent risk for biopsychosocial problems.

The Problem

Research Overview

• Parents/patients consent for self-report measures, and

video recording.

• Parents complete extensive battery of questionnaires on

self and children: personality/temperament.

• Parents report emotional states immediately before

treatment.

• Video record treatment session; obtain pain/distress

ratings.

• Two to three weeks later, parents provide health history

for children.

• Judges use video recordings to code parents’

communication patterns and nonverbal behavior.

Measures

• Medical/Demographic Information:

– Child’s age, gender, and ethnicity; diagnosis, length of treatment, number of treatments

• verall and in last two months, type of

procedure, parents’ age

• Pain/Distress: Ratings from parents,

nurses, children, independent observers using Faces scale

Measures

• Emotional States

– Negative: State Anxiety (Spielberger, 1977)-- unpleasant emotional arousal in face of threatening demands or dangers (e.g. “anxious”, “distressed”, “upset”)- 20 items – Positive: Empathic Concern (Batson, 1991)—an other-oriented emotional response, associated with concern for the well being of a person in distress (e.g., “warm”, “tender”, “compassion”)- 5 items

Correlations Between Emotional States and Pain/Distress

• 0.6
• 0.5
• 0.4
• 0.3
• 0.2
• 0.1

0.1 0.2 0.3 0.4

Pain (parents) Distress (parents) Pain (nurses) Distress (nurses) Pain (child) Distress (observers)

State Anxiety Empathic Concern

* Correlation Coefficients

Dispositions, Emotions, and Pain/Distress

Child Resilience Parents’Empathic Concern Distress

B = -.108 (p < .05) B =

• .

1 4 , p < . 1 B = .367, p < ..05 B = -.044 (p > .39)

Parents’ Communication Patterns

• Supportive: Parent is active and protective

partner; verbally and nonverbally attentive to child’s needs; offers comfort.

• Distancing: Parent is uninvolved bystander or
• bserver leaves the situation physically and/or

emotionally.

• Normalizing: Parent engages in activities

encountered in everyday life.

• Invalidating: Parent communication denies the

validity or reality of child experience, or child as credible source in defining situation (ridicule, criticize) (19%; n=6)

Supportive

Distancing

Normalizing

Invalidating

Emotional States of Invalidators vs. Others

2 4 6 8 10 12 14 16 18 Empathic Concern Invalidators Others

Emotional States *

Pain/Distress Ratings: Invalidators vs. Others

1 2 3 4 5 6

Pain (parents) Distress (parents) Pain (nurses) Distress (nurses) Pain (child) Distress (observers)

Invalidators Others

Pain Ratings (All differences significant)

• NIH/NCI 1U01CA114583-01 Community Network Program

for Older, Underserved African American Adult (Albrecht, Lichtenberg, co-PIs)

Pilot Study 1: “Improving Understanding of Racial/Ethnic Differences in Prostate Cancer Treatment in Metropolitan Detroit” (combining Medicare and SEER Registry databases) PIs:W. Underwood, MD; T. Albrecht, PhD

• NIH/NICHD R21HDO50450 Using Doctor-Patient Teams to

Reduce Health Disparities (L. Penner, PI)

• CDC H75/CCH524709-01 National Center for Vermiculite and

Asbestos Related Cancers. Exploring Psychosocial Issues Related to Vermiculite/Asbestos Exposure: A Community-Based Focus Group Study (Montana and Detroit) (J. Ruckdeschel, PI)

Communication and Health Disparities/Unequal Outcomes

Community Partners

• AARP Michigan
• ACS-Great Lakes Division
• Adult Well-Being Services
• Cancer Control Division-MDCH
• City of Detroit Dept. of Health

and Wellness

• Catholic Archdiocese, Detroit
• City of Detroit Senior Citizens

Department

• Detroit Area Agency on Aging
• Detroit Area Parish Nurses

Network

• Detroit Design Team
• Greater Detroit Area Health

Council, Inc.

• Hannan Foundation
• KCI Community Advisory

Committee

• Michigan Cancer Consortium
• Michigan Society of Hematology

and Oncology

• Neighborhood Services

Organization

• National Council on Alcoholism

and Drug Dependence

• Oakland University School of

Nursing

• ProLiteracy Detroit
• Pfizer, Inc.
• State Senator Buzz Thomas -

Detroit District Office

• Sisters Network, Inc.
• Interfaith Health Council of Detroit

Network relationships without CNP/KCI/IOG

Key Pink Lines – reported relationship Blue Lines – reciprocated Relationship Yellow Dots – Aging Groups Green Dots – Cancer Concerned Groups Blue Dots – General Health Groups Red Dots – Other Groups

Network relationships with CNP in place

Key Pink Lines – reported relationship Blue Lines – reciprocated Relationship Yellow Dots – Aging Groups Green Dots – Cancer Concerned Groups Blue Dots – General Health Groups Red Dots – Other Groups