Use of Role Model Stories to Overcome Barriers to Hospice among African Americans

Abstract

Objectives:

To test a brochure comprising hospice patient role model stories aimed at improving attitudes and knowledge of hospice among older African Americans.

Design:

Pre–post community-based study.

Setting:

Community-based organizations including senior centers, community exercise programs, churches, and senior care management services in greater Los Angeles area.

Participants:

Seventy-one African Americans aged 65 and older.

Intervention:

Hospice brochure containing theoretically driven role model stories portraying African Americans' experience with hospice, their initial attitudes and beliefs about hospice, factors influencing their enrollment in the program, and outcomes following enrollment.

Measurements:

Change from baseline in attitudes toward, knowledge of, and intentions to enroll in hospice.

Results:

Mean knowledge scores (range, 0–9) increased from 5.76 (standard deviation [SD] = 2.12) at pretest to 6.95 (SD = 1.90) at posttest (t = −6.14, p < 0.001). Average attitude toward hospice (range, 12–72) improved 12.5% from 41.16 (SD = 5.20) to 46.29 (SD = 5.45) (t = −7.52, p < 0.001). Intentions to enroll a family member in hospice increased from 84.5% to 92.9% (p = 0.002) and intentions to seek hospice care for themselves increased from 83.1% to 95.8% (p < 0.001).

Conclusion:

This small pre–post pilot study found that exposure to a hospice brochure containing theoretically driven, culturally parallel, role model stories was effective in improving knowledge of and attitudes toward hospice as well as intentions to enroll a family member or self in hospice care. Further study is needed to determine the impact of using this brochure within a clinical setting.

Introduction

Enrollment in hospice has continued to rise, with 1.45 million patients, representing 38.5% of all national deaths, enrolled in 2008.¹ Despite this growth, African Americans continue to be less likely than whites to use hospice,^2–4 with a recent study³ reporting African Americans with heart failure are 41% less likely to receive hospice care than whites. As a result, African Americans are more likely to die in the hospital as compared with whites.^4,5

Numerous barriers to hospice enrollment among African Americans have been identified and include lack of awareness and knowledge of hospice services,^6–8 misconceptions about hospice,^6,7 preference for more aggressive care,^4,9 and lack of trust in the health system and providers.^6,8 Furthermore, a qualitative study⁶ found that although African Americans prefer to have family members as caregivers, there is also a strong sentiment to minimize caregiver burden, a concern that has been identified as a barrier to hospice enrollment.^10,11 Additionally, an expected lack of minority employees within hospice care can also prevent enrollment among minorities.⁸

Since overcoming hospice barriers is an integral part to increasing quality end-of-life care and facilitation of informed decision-making among African Americans, the unanimous recommendation is to increase knowledge and awareness of hospice care through culturally appropriate educational materials and interventions.^{4,6,8,12–15} To the authors' knowledge, no studies exist that examine the use of culturally specific hospice materials or literature.

Use of stories depicting “model” individuals relay a nonthreatening narrative rich in cultural values and cultural relevancy to inform a specific community and target behavior. Widely used in the late 1990s, role model stories have been used to address a variety of issues such as HIV and AIDS risk behaviors,^16–19 cervical cancer education,²⁰ smoking cessation,²¹ and obesity prevention.²² While reporting successful outcomes, studies using role model stories to specifically target African Americans have focused solely on issues such as HIV/AIDS risk,^16–18 other sexually transmitted diseases,²³ and Pap smear screenings.²⁴

The purpose of this study is to describe the development of theoretically driven role model stories, real stories of African Americans' experience with hospice, and conduct a pilot evaluation of the effectiveness of this media in improving knowledge and attitudes toward hospice and increasing intentions to enroll in hospice care.

Methods

Theoretical framework

Two theoretical models were used to guide the development of the role model stories: Social Learning Theory²⁵ and the Theory of Reasoned Action.²⁶ Social Learning Theory posits that individuals learn from observing the behavior of others, whether in person or depicted through media materials. Furthermore, individuals learn from others' experiences in engaging in the behavior, as well as vicariously experience the rewards or punishment related to that behavior. These rewards provide the motivation for the observer to imitate the modeled behavior. In terms of hospice, individuals can learn about hospice through the experiences of others who have received it. Through the stories, readers can observe others' misconception about hospice and the positive outcomes (rewards) associated with hospice enrollment.

The Theory of Reasoned Actions postulates that behavioral intentions, the extent to which an individual plans to engage in the behavior, are predicted by the individual's attitude toward and perceived subjective norms about the behavior. Attitudes are comprised of beliefs about the consequences of the behavior, both positive and negative, which are weighted in the formulation of the attitude toward the behavior. Subjective norms describe how others in their social environment perceive the importance of the behavior as well as the importance of their opinion on that behavior to the individual. For example, an individual may believe that enrolling in hospice will provide comprehensive medical care, but that everyone who enrolls in hospice dies quickly. These negative and positive beliefs are weighed unequally in the development of the overall attitude toward hospice. Social norms may be more strongly influenced by the opinion of their peers than by physician recommendation, so gaining the support of a family member or peer may carry heavier influence. Combined, attitudes and social norms predict intentions to enroll in hospice, which are associated with actual behaviors.

Intervention: Role model brochure development

All participants were provided with a role model brochure titled Questions and Answers about Hospice for You or a Loved One (available at www.picf.org), and asked to read the entire brochure. The brochure was driven by the theoretical framework described above, and developed based on the actual words and stories of African Americans currently receiving hospice and family members of an African American who had received hospice care in the previous two years.

Recruitment of brochure role models

Role models were recruited using flyers distributed to local hospice programs and their staff, who were instructed to provide the flyers to their African American clients and family members and encourage them to contact the project principal investigator (PI) if they had interest in participating. Hospice staff did not directly recruit role models to prevent undue pressure or potential coercion. Individuals contacting the PI were provided detailed information on the purpose of the project and were offered $100 for participating and given the option of using their own name and photo in the story.

Interview protocol and story development

Individuals interested in participating as role models were contacted by the project coordinator, a senior African American social worker, who arranged to interview them in their home. Interviews were conducted using a semistructured protocol aimed at eliciting sufficient personal background information to establish identity with the target population. Other questions were developed to gather information consistent with the theoretical framework driving this project and included interviewee attitudes, knowledge, and concerns about hospice prior to enrollment and identify any change in these following receipt of service. Audio-taped interviews were transcribed verbatim, maintaining actual words and the richness of language used as well as cultural nuances of the language and wording.

Stories were then developed through abstracting segments of the interviews associated with responses to the theoretically driven interview protocol, specifically to illustrate individuals initial attitudes and beliefs about hospice (potential barriers to enrollment), motivating factors influencing their enrollment in the program, and outcomes, both positive and negative, following program enrollment. Stories were formulated at an eighth-grade literacy level to appeal to a broad audience. Word count for each story was limited to 350–450 words. Stories maintained the stated barriers encountered, if any, and the outcomes experienced by the role model. While several barriers were identified among those interviewed for this brochure, including misconceptions about hospice, beliefs that hospice enrollment meant imminent death, and hospice represented suboptimal care, no negative experiences were reported, and thus were not included in the stories. (See Appendix A.)

The brochure also included information about hospice, presented as short text and bullet points to highlight key program components. Specific efforts were made to include points related to commonly held misconceptions such as that most hospice care is provided in the home. An interactive quiz was included on the back cover to engage the reader.

Pretesting

Two focus groups were conducted with 20 African Americans, 10 females and 10 males, 65 and older to solicit feedback on all elements of the brochure including imagery, language and the graphic elements of the brochure to ensure cultural appropriateness and resonance with the target audience. Three mock-ups were obtained from separate graphic designers and reviewed by focus group participants. Responses were recorded through notes taken during the focus groups. Materials were developed based on the feedback received from focus group participants.

Semifinal brochures were pretested among health care professionals working with older adults and hospice patients to gather final input on design and hospice facts. This expert review resulted in changes to increase color contrast and font size to improve legibility.

Study design

A pre/post design was used to test the impact of the role model brochures on attitudes toward and knowledge of hospice and participants' intentions to enroll either oneself or a family member in hospice. The pre-test survey was administered, followed by a 30- to 45-minute period in which the participant was given the role model brochure to read. The posttest survey was administered after participants completed reading the brochure, 30–45 minutes following the baseline survey. This study was reviewed and approved by the Partners in Care Foundation's and University of Southern California's Institutional Review Board.

Pilot study sample eligibility and recruitment

Eligibility criteria included being an African American adult, age 65 and older, plus having two or more chronic conditions (such as heart disease, diabetes, hypertension, etc.) identified through self-reports. We targeted those with two or more chronic conditions to ensure our sample was not comprised of all healthy older adults. Participants were recruited from community-based organizations including senior centers, churches, and community activity groups.

Measures

The survey used in this study was developed specifically to measure knowledge, attitudes, and intentions toward hospice and hospice enrollment as, according to the theoretical framework, these are the core factors associated with behavior. A series of questions were used to measure knowledge (9 questions), attitudes (12 questions) and intentions toward hospice enrollment (Table 1).

Table 1.

Responses to Knowledge and Attitude Questions (n = 71)

		Pre	Post
Knowledge questions		% Correct
1	All adults and children who have an illness that cannot be cured are able to get hospice services, not just those with cancer.	69.0%	91.5%
2	Patients can stop hospice services and start them again at a later time if they want to.	64.8%	85.9%
3	Patients must have health insurance to get hospice services.	54.9%	54.9%
4	A patient must be told by their doctor that they have 6 months to live or less to be allowed to get hospice care.	59.2%	29.6%
5	If a patient on hospice lives more than 6 months, hospice services must be stopped.	80.3%	95.8%
6	Hospice workers are available by phone 24-hours a day, every day.	87.3%	98.6%
7	Hospice provides medical, psychological, and spiritual care for patients.	83.1%	94.4%
8	Most patients that are in hospice receive care at home.	59.2%	94.4%
9	Hospice benefits pay for medications.	50.7%	88.7%

Attitude questions		% Agree
1	I am comfortable talking about hospice with my doctor.	76.1%	85.7%
2	Talking about hospice makes patients and families feel hopelessness and that they are giving up.	26.7%	11.5%
3	Medical care in the home is not as good as the hospital.	31.9%	15.5%
4	When there is no hope for curing a patient and their life will last months instead of years, doctors should always talk about hospice as an option.	76.0%	88.8%
5	Doctors should generally try to keep their patients alive on machines for as long as possible.	25.4%	23.9%
6	Talking about hospice services should be done with patients before they are in the last stages of their disease.	83.1%	95.8%
7	I know what hospice is.	70.4%	90.2%
8	I know how long hospice cares for a patient.	26.7%	63.4%
9	Patients and families do not want to talk about hospice services.	33.8%	21.1%
10	Patients and families do not want to have strangers in their home, even if the strangers are with hospice.	39.4%	21.1%
11	I know the types of services hospice provides.	50.7%	91.6%
12	If my doctor recommended hospice care, I would feel that he/she is giving up on me.	28.2%	12.9%

Knowledge questions included a response set of true and false, and were adapted from a previous study of hospice knowledge²⁷ to reflect commonly held myths about hospice (e.g., hospice location of care, eligibility criteria, and financing). Total scores reflect number of correctly answered questions, ranging from 0 to 12. The attitude questionnaire was adapted from a survey designed to ascertain physician attitudes toward hospice²⁷ and were followed by a Likert scale of strongly disagree, disagree, neutral, agree, and strongly agree, with possible scores ranging from 12 (negative attitude) to 60 (positive attitude). The final two questions on intentions to enroll in hospice, also scored by this same Likert scale, were:

If a family member was extremely ill, I would consider enrolling them in hospice.

If I had a disease that could not be cured, I would want to receive hospice care.

The survey was reviewed for face validity by two palliative care physicians and pretested with an African American over 65 years old.

All surveys were administered by trained social workers. Participants were interviewed individually and instructed to read the brochure and refrain from discussing brochure content with anyone, including the survey administrators, to ensure that participants' attitudes, knowledge and intentions were not being influenced by discussion with others, and to improve the measurement of effects from the brochure itself.

Analysis

Descriptive statistics were used to describe participant characteristics. Knowledge scores were computed by summing the scores from the instrument items for a composite score. Attitude scores were computed by reverse coding negative items and summing all questions for a composite score, with a high score indicating positive attitude toward hospice. Differences between pre and post composite scores were analyzed using paired t tests. Nonparametric statistics were used to compare the change in intention items from pre to post. Analyses were conducted using SPSS 15.0 statistical software package (SPSS Inc., Chicago, IL).

Results

Sample description

A total of 71 African Americans were interviewed (Table 2). Participants' ages ranged from 63 to 91. The majority were retired (72.9%) African American (100%) females (80.3%), and 33.8% were married, 28.2% divorced, and 28.2% widowed. One participant identified her ethnic background as mixed, “black/West Indian.” Educational level of was generally high, with 45.1% having completed some college, and 12.7% being college graduates. Annual incomes were most commonly between $10,000–$19,000 (48.3%) or $20,000–$29,000 (27.6%). More than half (60%) reported knowing someone that had received hospice care.

Table 2.

Participant Demographics (n = 71)

	Frequency	Percent
Gender
Female	57	80.3
Marital status
Single	6	8
Married	24	33.8
Divorced/separated	21	29.6
Widowed	20	28.2
Ethnicity^a
Black	71	100.0
Other	1	1.4
Education level
8th grade or less	1	1.4
Some high school	5	7
High school graduate	19	26.8
Some college	32	45.1
College graduate	9	12.7
Post graduate	5	6.7
Annual Income
Under $10,000	4	6.9
$10,000–$19,000	28	48.3
$20,000–$29,000	16	27.6
$30,000–$39,000	0	0
$40,000–$49,000	6	10.3
$50,000 or more	4	6.9
Work status
Work full-time	7	10
Work part-time	8	11.4
Unemployed	4	5.7
Retired	51	72.9
Know(n) anyone receiving hospice care?
Yes	42	60

One participant reported mixed ethnicity, therefore total exceeds 100%.

Knowledge

Paired t tests resulted in a significant improvement in knowledge of hospice from pretest to posttest. Mean knowledge scores increased from 5.76 (standard deviation [SD] = 2.12) at pretest to 6.95 (SD = 1.90) at post test, a statistically significant improvement (t = −6.14, p < 0.001; Table 3). Knowledge questions demonstrating the greatest improvement from baseline included knowing that hospice covers costs of medications (75% improvement) and that most hospice care is provided in the home (59% improvement). More participants had an incorrect response at posttest to the question regarding needing a physician referral to hospice in the last 6 months of life, declining from 59% to 30% correct (Table 1).

Table 3.

Changes in Attitude, Knowledge, and Intentions (n = 71)

	Pre	Post
	Mean ± standard Deviation		Effect Size, %	p value
Attitude	41.16 ± 5.2	46.29 ± 5.4	12.5%	< 0.001^a
Knowledge	5.76 ± 2.1	6.95 ± 1.9	20.7%	< 0.001^a
Intentions
Self	4.25 ± 0.86	4.54 ± 0.63	N/A^b	0.012^c
Family	4.23 ± 0.85	4.46 ± 0.63	N/A^b	0.005^d

p < 0.001.

Wilcoxon signed ranks test.

p < 0.05.

p < 0.01.

Attitudes

Results of analysis revealed a significant improvement in attitudes toward hospice at posttest. Mean attitude scores increased from 41.16 (SD = 5.20) at pretest to 46.29 (SD = 5.45) at posttest, a statistically significant improvement (t = −7.52, p < 0.001; Tables 1 and 3).

Intentions

Significant change was also observed in stated intentions to enroll a family member or oneself in hospice should they become seriously ill. At pretest 84.5% of respondents either agreed or strongly agreed that they would consider enrolling a family member in hospice which increased to 92.9% after review of the hospice brochure (χ² = 21.37; p = 0.002). In terms of intentions to seek hospice care for oneself, a significant increase was also found, with 83.1% agreeing at pretest, increasing 12.7% to 95.8% of respondents at posttest (χ² = 46.59; p < 0.001; Table 3).

Discussion

Exposure to the hospice brochure resulted in significant improvements in knowledge of and attitudes toward hospice. A specific area of knowledge improvement included a key misconception regarding hospice care: location of care. Following brochure review, more respondents correctly recognized home as the primary site of hospice care. Participants scored lower at follow-up regarding needing a physician referral to hospice in the last 6 months of life. This may be due to one story that implied the protagonist survived more than 6 months; plus this fact was not clearly stated within the brochure, an area for improvement. While attitudes toward hospice improved overall, there was little change in participants' belief that doctors should generally try to keep their patients alive on machines for as long as possible, a belief that may delay and prevent access to hospice. This lack of change may reflect a true preference for life prolonging care rather than a negative attitude toward hospice that can be impacted by this brochure. Although intentions to enroll a family member or oneself in hospice were high at baseline, there was a significant increase in behavioral intentions following brochure review. These findings suggest that the brochure positively impacted overall knowledge, attitudes, and intentions toward hospice enrollment, and are consistent with findings from previous research that reported a positive association between exposure to hospice literature and positive beliefs about hospice.¹⁴ Moreover, African Americans with positive hospice experiences are more likely to consider hospice than those without this exposure.²⁸

In order to ensure that the evaluation reflected the effect of the brochure itself rather than communication engendered by the brochure, efforts were made to restrict conversation between participants. These limitations would not be typical within a clinical or community setting, thus it is anticipated that the brochures could potentially promote conversations on end of life care options, both between patients and physicians and with family members. Anecdotal evidence supporting this was found following completion of the follow-up surveys. One African American community leader was tearful, stating one story reminded him of a deceased family member. Others remarked how they wished a particular family member had had hospice before they died. These comments also attest to the extent to which participants can identify with the role model; a key factor involved in role model story development.¹⁷

Limitations of this study include the limited geographic location of the sample, the small sample size, the lack of comparison group, and the study design that includes a convenience sample of individuals who were not seriously ill. Additionally, the surveyed sample had high intentions to enroll in hospice at baseline; therefore findings from this study may not be generalizable among individuals less positively inclined toward hospice.²⁸ Knowledge and attitude measures were adapted from existing studies, and have not been validated. While theoretical models support the relationship between intention and actual behaviors, the lack of longitudinal data in this study limits the ability to conclude the influence of this brochure in improving hospice enrollment rates. This study tested the impact of the overall brochure and did not test the impact of role model stories, or other components (e.g., facts, interactive quiz about hospice readiness), individually. Thus additional studies employing randomized designs among seriously ill older adults, testing effects of brochure components, with long-term follow-up measures are needed to determine the extent to which this publication influences actual hospice enrollment.

Implication

Studies have documented barriers in physician communications about hospice care for seriously ill patients,^29,30 with a recent national study finding that physicians discussed hospice care with only 53% of their patients with stage IV lung cancer in the 2 months before death.³¹ This rate was even lower among African Americans, with only 49% having these conversations with their physician. Use of media material may improve the likelihood of patients engaging in these conversations with their physicians. This concept is supported by Browning³² in her article on empowerment for effective end-of-life decision-making. She points to the importance of a “discovery continuum” in which a patient or family member gathers and processes information they feel is relevant and enables them to make a unique, contextual decision.³² Browning asserts that indecisiveness is a product of confusing terminology, lack of information, and feelings of oppression, and recommends situational and contextual discussions betweens patients, family members, and providers which may be facilitated by educational materials.

The role model brochures may also stimulate conversation earlier in the disease trajectory. By sharing the brochure with patients earlier in their disease process, physicians can begin discussion over a longer period of time, while simultaneously acknowledging to patients that although hospice may not be something they are ready for now, eventually it is a care option for most individuals. Studies have found that end-of-life conversations with physicians result in longer stays in hospice³³ which in turn, is associated with prolonged lives of patients with particular diseases,^1,35 greater overall satisfaction with end-of-life care, and fewer family concerns.^33,34,36

Conclusion

Use of theoretically driven educational materials, tailored specifically for underserved populations may be effective in improving knowledge about and attitudes toward hospice. Further research is needed to determine if use of this material improves access to end of life care programs among African Americans.

Footnotes

Acknowledgment

This research was supported by a grant from PacifiCare/United Healthcare.

Author Disclosure Statement

No conflicts of interest exist for any of the authors as there are no commercial associations, affiliations, or financial interests to disclose. This study was funded by PacifiCare/United Healthcare, which did not participate in the design or implementation of this grant.

Appendix A. Sample Role Model Story

References

National Hospice and Palliative Care Organization. NHPCO Facts and Figures: Hospice Care in America. 2009:1–15. www.nhpco.org/files/public/Statistics_Research/NHPCO_facts_and_figures.pdf. 2010 December 21.

Enguidanos

, Yip

, Wilber

. Ethnic variation in site of death among Medicaid/Medicare dually eligible older adults. J Am Geriatr Soc, 2005; 53:1411–1416.

Givens

, Tjia

, Zhou

, Emanuel

, Ash

. Racial and ethnic differences in hospice use among patients with heart failure. Arch Intern Med, 2010; 170:427–432.

Kwak

, Haley

, Chiriboga

. Racial differences in hospice use and in-hospital death among Medicare and Medicaid dual-eligible nursing home residents. Gerontologist, 2008; 48:32–41.

Lackan

, Eschbach

, Stimpson

, Freeman

, Goodwin

. Ethnic differences in in-hospital place of death among older adults in California: Effects of individual and contextual characteristics and medical resource supply. Med Care, 2009; 47:138–145.

Born

, Greiner

, Sylvia

, Butler

, Ahluwalia

. Knowledge, attitudes and beliefs about end-of-life care among inner-city African Americans and Latino/Hispanic Americans. J Palliat Med, 2004; 7:247–256.

Chung

, Essex

, Samson

. Does Caregiver knowledge matter for hospice enrollment and beyond? Pilot study of minority hospice patients. Am J Hosp Palliat Med, 2009; 26:165–171.

Washington

, Bickel-Swenson

, Stephens

. Barriers to hospice use among African Americans: A systematic review. Health Soc Work, 2008; 33:267–274.

McKinley

, Garrett

, Evans

, Danis

. Differences in end-of-life decision making among black and white ambulatory cancer patients. J Gen Intern Med, 1996; 11:651–656.

10.

Bullock

, McGraw

, Blank

, Bradley

. What matters to older African Americans facing end-of-life decisions? A focus group study. Journal of Social Work in End-of-Life and Palliative Care., 2005; 1,3:3–19.

11.

Jenkins

, Lapelle

, Zapka

, Kurent

. End-of-Life care and African Americans: Voices from the community. J Palliat Med, 2005; 8:585–592.

12.

Lorenz

, Ettner

, Rosenfeld

, Carlisle

, Liu

, Asch

. accommodating ethnic diversity: A study of California hospice programs. Med Care, 2004; 42:871–874.

13.

Yancu

, Farmer

, Leahman

. Barriers to hospice use and palliative care services use by African American adults. Am J Hosp Palliat Med, 2010; 27:248–253.

14.

Johnson

, Kuchibhatla

, Tulsky

. Racial Differences in self-reported exposure to information about hospice care. J Palliat Med, 2009; 10:921–927.

15.

Rhodes

, Teno

, Welch

. Access to hospice for African Americans: Are they informed about the option of hospice? J Palliat Med, 2006; 9:268–272.

16.

Berkley-Patton

, Goggin

, Liston

, Bradley-Ewing

, Neville

. Adapting effective narrative-based HIV-prevention interventins to increase minorities' engagement in HIV/AIDS services. Health Commun, 2009; 24:199–09.

17.

Corby

, Enguidanos

, Kay

. Development and use of role-model stories in a community-level risk-reduction intervention. Public Health Rep, 1996; 111,Suppl 1:54–58.

18.

King

, Nu'Man

, Fuller

, Brown

, Smith

, Howell

, Little

, Patrick

, Glover

. The diffusion of a community-level HIV intervention for women: lessons learned and best practices. J Womens Health (Larchmt), 2008; 17:1055–1066.

19.

Terry

, Liebman

, Person

, Bond

, Dillard-Smith

, Tunstall

. The Woman and Infants Demonstration Project: An integrated approach to AIDS prevention and research. AIDS Educ Prev, 1999; 11:107–121.

20.

Hou

, Fernandez

, Parcel

. Development of a cervical cancer educational program for Chinese women using intervention mapping. Health Promot Pract, 2004; 5:80–87.

21.

Whittaker

, Maddison

, McRobbie

, Bullen

, Denny

, Dorey

, Ellis-Pegler

, van Rooyen

, Rodgers

. A multimedia mobile phone-based youth smoking cessation intervention: findings from content development and piloting studies. J Med Internet Res, 2008; 10:e49.

22.

Reininger

, Barroso

, Mitchell-Bennett

, Cantu

, Fernandez

, Gonzalez

, Chavez

, Freeberg

, McAlister

. Process evaluation and participatory methods in an obesity-prevention media campaign for Mexican Americans. Health Promot Pract Health Promot Pract, 2010; 11:347–357.

23.

Corby

, Wolitski

. Condom use with main and other sex partners among high-risk women: Intervention outcomes and correlates of reduced risk. Drug Soc (New York), 1996; 9:75–96.

24.

Suarez

, Nichols

, Brady

. Use of peer role models to increase Pap smear and mammogram screening in Mexican-American and black women. Am J Prev Med, 1993; 9:290–296.

25.

Bandura

. Self-efficacy: Toward a unifying theory of behavioral change. Psychol Rev, 1977; 84:191–215.

26.

Ajzen

, Fishbein

. Understanding Attitudes and Predicting Social Behavior. Englewood Cliffs, NJ: Prentice-Hall International Inc., 1980.

27.

Ogle

, Mavis

, Wyatt

. Physicians and hospice care: Attitudes, knowledge, and referrals. J Palliat Med, 2002; 5:85–92.

28.

Ludke

, Smucker

. Racial Differences in the willingness to use hospice services. J Palliat Med, 2007; 10:1329–1337.

29.

Brickner

, Scannell

, Marguet

, Ackerson

. Barriers to hopice care and referrals: Survey of physicians' knowledge, attitudes, and perceptions in a health maintenance organization. J Palliat Med, 2004; 7:411–418.

30.

McGorty

, Bornstein

. Barriers to physicians' decisions to discuss hospice: Insights gained from the United States hospice model. J Eval Clin Pract, 2003; 9:363–372.

31.

Huskamp

, Keating

, Malin

, Zaslavsky

, Weeks

, Earle

, Teno

, Virnig

, Kahn

, He

, Ayanian

. Discussions with physicians about hospice among patients with metastatic lung cancer. Arch Intern Med, 2009; 169:954–962.

32.

Browning

. Empowering family members in end-of-life care decision making in the intensive care unit. Dimens Crit Care Nurs, 2008; 28:18–23.

33.

Wright

, Zhang

, Ray

, Mack

, Trice

, Balboni

, Mitchell

, Jackson

, Block

, Maciejewski

, Prigerson

. Associations between end-of-life discussions, patient mental health, medical care near death, and caregiver bereavement adjustment. JAMA, 2008; 300:1665–1673.

34.

Zhang

, Wright

, Huskamp

, Nilsson

, Maciejewski

, Earle

, Block

, Maciejewski

, Prigerson

. Health care costs in the last week of life: Associations with end-of-life conversations. Arch Intern Med, 2009; 169:480–488.

35.

Connor

, Pyenson

, Fitch

, Iwasaki

. Comparing hospice and nonhospice patient survival among patients who die within a three-year window. J Pain Symptom Manage, 2007; 33:238–246.

36.

Teno

, Shu

, Casarrett

, Spence

, Rhodes

, Connor

. Timing of referral to hospice and quality of care: Length of stay and bereaved family members' perceptions of the timing of hospice referral. J Pain Symptom Manage, 2007; 34:120–125.