Participation of Underrepresented Populations:in Cancer-related Trials

Jean G. Ford, M.D.Associate Professor of EpidemiologyJohns Hopkins Bloomberg School of Public Health

Objectives

■ Discuss results of evidence-based review

■ Highlight future directions for research

Background

■ 1993 NIH Revitalization Act called for inclusion of women & minorities in all human subjects research

■ NCI budget nearly doubled from 1993 to 2002– Trial accrual increased– Not clear if all populations benefited

■ Some populations remain under-represented in NCI-funded clinical trials

Key Questions

■ What are barriers & promoters for participation of underrepresented populations in cancer trials?

■ What effects do healthcare providers have on recruitment of under-represented populations?

■ What recruitment strategies are efficacious?■ What measures of recruitment success have

been used?■ What methods have been used to study

recruitment strategies?

Conceptual Framework

Awareness

Opportunity

Acceptance/Refusal

AwarenessBarriers/

Promoters

Interventions

OpportunityBarriers/

Promoters

Moderators/Sociodemographic

Factors

Measuresof

Success

Acceptance/RefusalBarriers/

Promoters

StudyDesign

Ford JG Howerton MW, Powe NR, Bass EB, et al., Cancer. 2008; 112:228-42.

Summary of Search

Electronic Databases

Retrieved5257

Title Review4928

Abstract Review1194

Article Review270

Hand Searching123

Duplicates341

Excluded3734

Excluded924

Excluded205

65 Articles Eligible

Barriers to Enrollment

Barriers to & Promoters of Enrollment

Key Findings from 65 studies

Study Characteristics

■ Dates published– < 1996: 10– 1996 – 2000: 14– 2001 – 2005: 41

■ Study Design:– Observational/Experimental: 9– Descriptive (registry reviews, surveys): 33– Qualitative (focus groups, semi-structured interviews): 24

■ Setting:– Hospital inpatient or outpatient: 43– Community: 33– Other: 7

Study Characteristics (II)

■ Type of population:– African-American: 35– Older adults: 19– Latino/Hispanic: 12– Asian/Pacific Islander: 6– Native American / Alaskan Native: 4– Adolescent: 4– Rural: 2

■ US-based: 40

Barriers to Awareness(N = 18 studies)

0

2

4

6

8

10

12

14

Education re.clinical trials

Culturallyappropriateinformation

Cancerknowledge

Physicianawareness

Barriers to Opportunity(N = 34 studies)

0

5

10

15

20

25

Logistic

s

Eligib

ility

Lack o

f dis

semin

atio

n

Costs

Provi

der re

latio

nship

Provi

der c

harac

teris

tics

Comm

unicat

ion

Lack o

f pro

toco

ls

Barriers to Acceptance(N = 36 studies)

02468

101214161820

Mis

trust

Fear

Culture

Family

Discom

fort

Provi

der-re

late

dCost

s

Transp

ortatio

nTim

e

Relig

ious b

elie

fs

Low h

ealth

lite

racy

Barriers stastically associated with accrual to cancer trials

■ Awareness (n = 3)– Lack of education about clinical trials– Lack of knowledge about origins of cancer

■ Opportunity (n = 16)– Sociodemographic characteristics– Lack of health insurance– Exclusion criteria (age, comorbid conditions)– Provider referral– Managed care

■ Acceptance (n = 3)– Perceived harm– Loss of control– Nature of intervention– Time commitment– Loss of income– Transportation

Ford JG, et al., Cancer. 2008; 112:228-42.

Other Key Findings onBarriers & Promoters

■ Available evidence mostly about accrual to therapeutic trials

■ Barriers to opportunity frequently reported for both prevention & treatment trials

■ Limited data on Latinos/Hispanics, Asian/Pacific Islanders, American Indians/Alaska Natives, older adults, & adolescents

■ Barriers differed across populations

Relation of Barriers & Promoters to Conceptual Framework

■ Awareness:– 8 barriers

■ Opportunity to participate:– 88

■ Acceptance/refusal of participation:– 44

Healthcare Provider Effects

Key Findings from 18 studies

Howerton MW, et al., Cancer. 2007; 109:465-76.

Studies Reporting on Provider Barriers to- and Promoters of Clinical Trial Enrollment

0

2

4

6

8

10

12

14

16

18

Knowledge Attitudes Communication

Studies Reporting on Provider Attitudes as Barriers to Clinical Trial Enrollment (n=15)

0

1

2

3

4

5

6

7

8

9

ProtocolAdherence

Patient Mistrust Patient Costs Data CollectionCosts/Burden

Eligibility Clinical Trials Patient (culturalcompetence)

Studies Reporting on Provider Promoters of Clinical Trial Enrollment (n=7)

00.5

11.5

22.5

33.5

44.5

Studies Reporting on Study Design Barriers (n=11)

0

1

2

3

4

5

6

7

8

9

Eligibility Protocol Length of study/visitstructure

Healthcare System Barriers

■ Lack of minority investigators/personnel (n=1)

■ Lack of dissemination of study opportunities to providers (n=2)

■ Lack of cultural competence among providers and/or staff (n=1)

■ Lack of access to institutions conducting cancer trial (n=1)

Efficacy of Recruitment Strategies

Key findings from 7 studies

Lai GY, et al. Clinical Trials. 2006; 3: 133-141.

Types of Promoters (I)

■ Most frequently reported:– patient incentives (n = 6)– provider incentives (n = 3)– altruism (n = 6)– culturally relevant education about trials (n = 4)

■ 59 distinct promoters of enrollment into cancer clinical trials– 36 therapeutic trials only– 14 prevention trials only– 17 therapeutic and prevention trials

Types of Promoters (II)

■ Promoters in relation to conceptual framework– 6 awareness– 29 opportunity– 25 acceptance

■ Levels at which promoters occur– 40 patient level– 12 provider level– 6 study design level– 2 healthcare system level

Figure D: Studies reporting on promoters of the opportunity for underrepresented populations to participate in cancer-related trials (n =

21)

0

1

2

3

4

5

6

7

Cultural competence /communication

Inclusion criteria Logistics Healthcare context Other factors Demographic factors

# o

f st

ud

ies

All studies

Studies w ith Af rican American

Studies w ith Elderly

Figure F: Studies reporting on promoters of acceptance of participation by underrepresented populations in cancer-re lated trials (n = 17)

0

1

2

3

4

5

6

7

8

9

10

Treatment /Treatment Process

Altruism Perceived benef its Familyconsiderations

Trust Other factors

# o

f st

ud

ies

All studies

Studies w ith Af rican American

Studies w ith Elderly

Recruitment Strategies/Interventions

Author, Year

Target Population Recruitment Intervention

Moinpour, 2000

African Americans (Sites A-D); Latino/Hispanic (Site E)

■Minority recruiter

Linnan, 2002

Workers in a manufacturing company

■Active recruitment arm –Employees signed up to participate

■Passive at worksites–Company provided list of employee names & home phone #s

Brewster, 2002

Latina/Hispanic women

■Clinic registry vs. media campaign

Recruitment Strategies/Interventions (II)

Author, Year Target Population

Recruitment Intervention

Ford, 2004 African Americans

4 increasingly intensive strategies ■Enhanced mailings ■Minority interviewers ■Letter & phone reminders■Church-based project sessions

Mandelblatt, 2005

Latinas ■5-10 minute educational counseling delivered by non-physician study staff + informational brochure■Informational brochure only

Recruitment Strategies/Interventions (III)

Author, Year Target Population Recruitment Intervention

Paskett, 2002 Rural physicians in North Carolina & South Carolina

■Tumor-reporting system■Nurse facilitator■Quarterly newspapers■Health educator

Kimmick, 2005 Older adults

CALGB Physicians

■Standard information (control)■Educational intervention

–Educational symposium–Geriatric oncology educational materials–Monthly mailings & e-mails for 1 year–Case discussion seminar

Results

Author Descriptor of Point Estimate Point Estimate

Moinpour, 2000

Change in Site A enrollment, 1995-1996

Change in Site B enrollment, 1995-1996

Change in Site C enrollment, 1995-1996

Change in Site D enrollment, 1995-1996

Change in Site E enrollment, 1995-1996

-0.3%

-0.5%

-1.8%

-0.6%

0.5%

Results (II)

Author Descriptor of Point Estimate

Point Estimate(p value)

Linnan, 2002

Difference in enrollment 36.6% (p<0.0001)

Brewster, 2002

Odds ratio of women presenting in clinic

Odds ratio of women screened by telephone

3.00 (2.38, 3.78)

2.97 (2.52, 3.51)

Ford, 2004

Difference in enrollment (Arm C vs. Arm D)

1.0% (p<0.01)

Results (III)

Author Descriptor of Point Estimate

Point Estimate(p value)

Mandelblatt, 2005

Intent to enroll in the STAR trial, intervention vs. control

■77% intent vs. 67% in intervention vs. control groups (P=0.03)

Kimmick, 2005

% enrollment of older adults, Intervention

% enrollment of older adults, Control

■Baseline: 40% (P=0.40)■Year 1: 36% (P=0.35)■Year 2: 31% (P=0.83)

■Baseline: 36%■Year 1: 32%■Year 2: 31%

Results (IV)

Author Descriptor of Point Estimate Point Estimate

Paskett, 2002

Enrollment change of breast cancer patients, 1991-1996 (NC)

Enrollment change of colorectal cancer patients, 1991-1996 (NC)

Enrollment change of breast cancer patients, 1991-1996 (SC)

Enrollment change of colorectal cancer patients, 1991-1996 (SC)

-9%

1%

44%

-5%

Recruitment Strategies/Interventions

UyBico SJ et al., JGIM 2007;22:852–863

Limitations of Studies of Efficacy of Recruitment Interventions

■ Few studies available■ All published studies focused on

accrual to prevention trials■ Limited generalizability■ Intensity needed to modify behavior ■ Varying quality of study methods

Summary

Application of Evidence to Conceptual Framework

Awareness

Opportunity

Acceptance/Refusal

AwarenessBarriers/Promoters

Lack of education about CTs Lack of knowledge about

origins of cancer Culturally relevant education

about CTs Lack of physician awareness Health literacy

Interventions

Opportunity Barriers/Promoters

Protocol characteristics Eligibility Costs to patient Provider knowledge Provider attitudes/beliefs Lack of dissemination of

study opportunities topatient/provider

Data collection costs/burden Provider’s communication/

method of presentation

Moderators/Sociodemographic

FactorsRace/Ethnicity

AgeGender

GeographyLanguage

IncomeSES

EducationCulture

Measures of Success

Acceptance/RefusalBarriers/Promoters

Trust in sponsor/investigator Perceived harms/benefits Fear Family Costs Time Transportation Altruism Religious / spiritual beliefs Incentives Provider-related Stress

Study Design

Overall Summary

■ More barriers to opportunity than to awareness or acceptance– More evidence on barriers than on promoters– Mistrust a common theme

■ Provider barriers at level of professionals, study design & healthcare system

■ Recruitment goals rarely reported a priori

Limitations of Evidence■ Heterogeneity

– study design– data quality

■ Relationship between barriers & promoters unclear

■ Overlap between underrepresented populations■ More evidence from therapeutic trials than

prevention trials■ Excluded studies other than clinical trials ■ Recruitment experience not always reported

Conclusions

■ In order to improve the knowledge base on accrual of underrepresented populations to clinical trials, hypothesis-driven studies are needed, with adequate power to detect differences.

■ The quality of the evidence can be improved by nesting recruitment studies within plans for conducting trials.

Principal Investigators:Jean G. Ford, MD

Mollie W. Howerton, PhD, MPH

Co-Investigators:Tiffany L. Gary, PhD, Gabriel Y. Lai, MHS.

Shari Bolen, MD, MPH, Jon Tilburt, MDM. Chris Gibbons, MD, MPH, Charles Baffi, PhD, MPH

Renee F. Wilson, MS, Carolyn J. FeuersteinPeter Tanpitukpongse, Neil R. Powe, MD, MPH, MBA

Eric B. Bass, MD, MPH

The EPC Team