[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Table 1.  Participant Demographic Characteristics for Vignettes
Participant Demographic Characteristics for Vignettes
Table 2.  Major Themes With Illustrative Quotations
Major Themes With Illustrative Quotations
Table 3.  Survey Responses
Survey Responses
1.
Breathett  K, Allen  LA, Helmkamp  L,  et al.  The Affordable Care Act Medicaid expansion correlated with increased heart transplant listings in African Americans but not Hispanics or Caucasians.   JACC Heart Fail. 2017;5(2):136-147. doi:10.1016/j.jchf.2016.10.013PubMedGoogle ScholarCrossref
2.
Breathett  K, Allen  LA, Helmkamp  L,  et al.  Temporal trends in contemporary use of ventricular assist devices by race and ethnicity.   Circ Heart Fail. 2018;11(8):e005008. doi:10.1161/CIRCHEARTFAILURE.118.005008PubMedGoogle Scholar
3.
Benjamin  EJ, Muntner  P, Alonso  A,  et al; American Heart Association Council on Epidemiology and Prevention Statistics Committee and Stroke Statistics Subcommittee.  Heart disease and stroke statistics, 2019 update: a report from the American Heart Association.   Circulation. 2019;139(10):e56-e528. doi:10.1161/CIR.0000000000000659PubMedGoogle ScholarCrossref
4.
Farmer  SA, Kirkpatrick  JN, Heidenreich  PA, Curtis  JP, Wang  Y, Groeneveld  PW.  Ethnic and racial disparities in cardiac resynchronization therapy.   Heart Rhythm. 2009;6(3):325-331. doi:10.1016/j.hrthm.2008.12.018PubMedGoogle ScholarCrossref
5.
Mezu  U, Ch  I, Halder  I, London  B, Saba  S.  Women and minorities are less likely to receive an implantable cardioverter defibrillator for primary prevention of sudden cardiac death.   Europace. 2012;14(3):341-344. doi:10.1093/europace/eur360PubMedGoogle ScholarCrossref
6.
Breathett  K, Liu  WG, Allen  LA,  et al.  African Americans are less likely to receive care by a cardiologist during an intensive care unit admission for heart failure.   JACC Heart Fail. 2018;6(5):413-420. doi:10.1016/j.jchf.2018.02.015PubMedGoogle ScholarCrossref
7.
Breathett  K, Yee  E, Pool  N,  et al.  Does race influence decision making for advanced heart failure therapies?   J Am Heart Assoc. 2019;8(22):e013592. doi:10.1161/JAHA.119.013592PubMedGoogle Scholar
8.
Colvin  M, Smith  JM, Hadley  N,  et al.  OPTN/SRTR 2017 annual data report: heart.   Am J Transplant. 2019;19(suppl 2):323-403. doi:10.1111/ajt.15278PubMedGoogle ScholarCrossref
9.
Chang  PP, Chambless  LE, Shahar  E,  et al.  Incidence and survival of hospitalized acute decompensated heart failure in four US communities (from the Atherosclerosis Risk in Communities Study).   Am J Cardiol. 2014;113(3):504-510. doi:10.1016/j.amjcard.2013.10.032PubMedGoogle ScholarCrossref
10.
DeFilippis  EM, Truby  LK, Garan  AR,  et al.  Sex-related differences in use and outcomes of left ventricular assist devices as bridge to transplantation.   JACC Heart Fail. 2019;7(3):250-257. doi:10.1016/j.jchf.2019.01.008PubMedGoogle ScholarCrossref
11.
Hsich  EM.  Sex differences in advanced heart failure therapies.   Circulation. 2019;139(8):1080-1093. doi:10.1161/CIRCULATIONAHA.118.037369PubMedGoogle ScholarCrossref
12.
Breathett  K, Jones  J, Lum  HD,  et al.  Factors related to physician clinical decision-making for African American and Hispanic patients: a qualitative meta-synthesis.   J Racial Ethn Health Disparities. 2018;5(6):1215-1229. doi:10.1007/s40615-018-0468-zPubMedGoogle ScholarCrossref
13.
McMurray  RJ, Clarke  OW, Barrasso  JA,  et al.  Gender disparities in clinical decision making—Council on Ethical and Judicial Affairs, American Medical Association.   JAMA. 1991;266(4):559-562. doi:10.1001/jama.1991.03470040123034PubMedGoogle ScholarCrossref
14.
Sadler  GR, Lee  H-C, Lim  RS, Fullerton  J.  Recruitment of hard-to-reach population subgroups via adaptations of the snowball sampling strategy.   Nurs Health Sci. 2010;12(3):369-374. doi:10.1111/j.1442-2018.2010.00541.xPubMedGoogle ScholarCrossref
15.
O’Brien  BC, Harris  IB, Beckman  TJ, Reed  DA, Cook  DA.  Standards for reporting qualitative research: a synthesis of recommendations.   Acad Med. 2014;89(9):1245-1251. doi:10.1097/ACM.0000000000000388PubMedGoogle ScholarCrossref
16.
Bertrand  M, Mullainathan  S.  Are Emily and Greg more employable than Lakisha and Jamal? a field experiment on labor market discrimination. National Bureau of Economic Research. Published July 2003. Accessed December 12, 2016. https://www.nber.org/papers/w9873
17.
Li  AC, Kannry  JL, Kushniruk  A,  et al.  Integrating usability testing and think-aloud protocol analysis with “near-live” clinical simulations in evaluating clinical decision support.   Int J Med Inform. 2012;81(11):761-772. doi:10.1016/j.ijmedinf.2012.02.009PubMedGoogle ScholarCrossref
18.
Shafer  K, Lohse  B.  How to conduct a cognitive interview: a nutritional education example. National Institute of Food and Agriculture. Published January 21, 2015. Accessed June 17, 2020. https://nifa.usda.gov/resource/how-conduct-cognitive-interview-nutrition-education-example
19.
Strauss  A, Corbin  JM.  Basics of Qualitative Research: Grounded Theory Procedures and Techniques. Sage Publications, Inc; 1990.
20.
Curry  LA, Nembhard  IM, Bradley  EH.  Qualitative and mixed methods provide unique contributions to outcomes research.   Circulation. 2009;119(10):1442-1452. doi:10.1161/CIRCULATIONAHA.107.742775PubMedGoogle ScholarCrossref
21.
Creswell  JW.  Qualitative Inquiry and Research Design: Choosing Among Five Approaches. SAGE Publications; 2012.
22.
Benjamini  Y, Hochberg  Y.  Controlling the false discovery rate: a practical and powerful approach to multiple testing.   J R Stat Soc (Series B). 1995;57(1):289-300. doi:10.1111/j.2517-6161.1995.tb02031.xGoogle Scholar
23.
Schulman  KA, Berlin  JA, Harless  W,  et al.  The effect of race and sex on physicians’ recommendations for cardiac catheterization.   N Engl J Med. 1999;340(8):618-626. doi:10.1056/NEJM199902253400806PubMedGoogle ScholarCrossref
24.
Daugherty  SL, Blair  IV, Havranek  EP,  et al.  Implicit gender bias and the use of cardiovascular tests among cardiologists.   J Am Heart Assoc. 2017;6(12):e006872. doi:10.1161/JAHA.117.006872PubMedGoogle Scholar
25.
Heise  L, Greene  ME, Opper  N,  et al; Gender Equality, Norms, and Health Steering Committee.  Gender inequality and restrictive gender norms: framing the challenges to health.   Lancet. 2019;393(10189):2440-2454. doi:10.1016/S0140-6736(19)30652-XPubMedGoogle ScholarCrossref
26.
American Psychological Association.  Who are family caregivers? Published 2011. Accessed June 3, 2020. https://www.apa.org/pi/about/publications/caregivers/faq/statistics
27.
Parker  K, Horowitz  JM, Stepler  R.  2: Americans see different expectations for men and women. Pew Research Center’s Social and Demographic Trends Project. December 5, 2017. Accessed April 20, 2020. https://www.pewsocialtrends.org/2017/12/05/americans-see-different-expectations-for-men-and-women/
28.
Smedley  B, Stith  A, Nelson  A.  Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Institute of Medicine; 2003. doi:10.17226/12875
29.
Feldman  D, Pamboukian  SV, Teuteberg  JJ,  et al; International Society for Heart and Lung Transplantation.  The 2013 International Society for Heart and Lung Transplantation guidelines for mechanical circulatory support: executive summary.   J Heart Lung Transplant. 2013;32(2):157-187. doi:10.1016/j.healun.2012.09.013PubMedGoogle ScholarCrossref
30.
Mehra  MR, Kobashigawa  J, Starling  R,  et al.  Listing criteria for heart transplantation: International Society for Heart and Lung Transplantation guidelines for the care of cardiac transplant candidates, 2006.   J Heart Lung Transplant. 2006;25(9):1024-1042. doi:10.1016/j.healun.2006.06.008PubMedGoogle ScholarCrossref
31.
Hart  M.  Subjective decisionmaking and unconscious discrimination.   Ala Rev. 2004;56:741-792. Accessed June 29, 2020. https://scholar.law.colorado.edu/cgi/viewcontent.cgi?article=1414&context=articlesGoogle Scholar
32.
Bui  QM, Allen  LA, LeMond  L, Brambatti  M, Adler  E.  Psychosocial evaluation of candidates for heart transplant and ventricular assist devices: beyond the current consensus.   Circ Heart Fail. 2019;12(7):e006058. doi:10.1161/CIRCHEARTFAILURE.119.006058PubMedGoogle Scholar
33.
Dew  MA, DiMartini  AF, Dobbels  F,  et al.  The 2018 ISHLT/APM/AST/ICCAC/STSW recommendations for the psychosocial evaluation of adult cardiothoracic transplant candidates and candidates for long-term mechanical circulatory support.   J Heart Lung Transplant. 2018;37(7):803-823. doi:10.1016/j.healun.2018.03.005PubMedGoogle ScholarCrossref
34.
Levy  N, Harmon-Jones  C, Harmon-Jones  E.  Dissonance and discomfort: does a simple cognitive inconsistency evoke a negative affective state?   Motiv Sci. 2018;4(2):95-108. doi:10.1037/mot0000079Google ScholarCrossref
35.
Stone  J, Moskowitz  GB, Zestcott  CA, Wolsiefer  KJ.  Testing active learning workshops for reducing implicit stereotyping of Hispanics by majority and minority group medical students.   Stigma Health. 2019;5(1):94-103. doi:10.1037/sah0000179Google ScholarCrossref
36.
Carnes  M, Devine  PG, Baier Manwell  L,  et al.  The effect of an intervention to break the gender bias habit for faculty at one institution: a cluster randomized, controlled trial.   Acad Med. 2015;90(2):221-230. doi:10.1097/ACM.0000000000000552PubMedGoogle ScholarCrossref
37.
Alfandre  D.  Clinical recommendations in medical practice: a proposed framework to reduce bias and improve the quality of medical decisions.   J Clin Ethics. 2016;27(1):21-27.PubMedGoogle Scholar
38.
Balsa  AI, McGuire  TG.  Prejudice, clinical uncertainty and stereotyping as sources of health disparities.   J Health Econ. 2003;22(1):89-116. doi:10.1016/S0167-6296(02)00098-XPubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Views 2,025
    Citations 0
    Original Investigation
    Cardiology
    July 21, 2020

    Association of Gender and Race With Allocation of Advanced Heart Failure Therapies

    Author Affiliations
    • 1Sarver Heart Center, Division of Cardiology, Department of Medicine, University of Arizona, Tucson
    • 2Sarver Heart Center, Clinical Research Office, University of Arizona, Tucson
    • 3College of Nursing, University of Arizona, Tucson
    • 4College of Nursing, University of Utah, Salt Lake City
    • 5Statistics Consulting Lab, Bio5 Institute, University of Arizona, Tucson
    • 6Department of Medicine, University of Arizona, Tucson
    • 7University of Rochester, New York
    • 8University of Arizona Medical School, Tucson
    • 9Department of Medicine, University of California, Los Angeles
    • 10Department of Psychology, University of Arizona, Tucson
    • 11Center for Population Health Sciences, University of Arizona, Tucson
    JAMA Netw Open. 2020;3(7):e2011044. doi:10.1001/jamanetworkopen.2020.11044
    Key Points español 中文 (chinese)

    Question  Is bias against a patient’s gender and race associated with the allocation of advanced heart failure therapies?

    Findings  In a qualitative study of 46 health care professionals, there was more bias against women compared with men when evaluating appearance and social support, particularly among African American women. Final recommendations were not different for groups defined by gender or race; all were offered ventricular assist devices rather than transplantation during interviews.

    Meaning  Although gender and race were not associated with allocation of advanced heart failure therapies in this study, there was evidence of bias, which could contribute to delayed and lower allocation of advanced therapies to women.

    Abstract

    Importance  Racial bias is associated with the allocation of advanced heart failure therapies, heart transplants, and ventricular assist devices. It is unknown whether gender and racial biases are associated with the allocation of advanced therapies among women.

    Objective  To determine whether the intersection of patient gender and race is associated with the decision-making of clinicians during the allocation of advanced heart failure therapies.

    Design, Setting, and Participants  In this qualitative study, 46 US clinicians attending a conference for an international heart transplant organization in April 2019 were interviewed on the allocation of advanced heart failure therapies. Participants were randomized to examine clinical vignettes that varied 1:1 by patient race (African American to white) and 20:3 by gender (women to men) to purposefully target vignettes of women patients to compare with a prior study of vignettes of men patients. Participants were interviewed about their decision-making process using the think-aloud technique and provided supplemental surveys. Interviews were analyzed using grounded theory methodology, and surveys were analyzed with Wilcoxon tests.

    Exposure  Randomization to clinical vignettes.

    Main Outcomes and Measures  Thematic differences in allocation of advanced therapies by patient race and gender.

    Results  Among 46 participants (24 [52%] women, 20 [43%] racial minority), participants were randomized to the vignette of a white woman (20 participants [43%]), an African American woman (20 participants [43%]), a white man (3 participants [7%]), and an African American man (3 participants [7%]). Allocation differences centered on 5 themes. First, clinicians critiqued the appearance of the women more harshly than the men as part of their overall impressions. Second, the African American man was perceived as experiencing more severe illness than individuals from other racial and gender groups. Third, there was more concern regarding appropriateness of prior care of the African American woman compared with the white woman. Fourth, there were greater concerns about adequacy of social support for the women than for the men. Children were perceived as liabilities for women, particularly the African American woman. Family dynamics and finances were perceived to be greater concerns for the African American woman than for individuals in the other vignettes; spouses were deemed inadequate support for women. Last, participants recommended ventricular assist devices over transplantation for all racial and gender groups. Surveys revealed no statistically significant differences in allocation recommendations for African American and white women patients.

    Conclusions and Relevance  This national study of health care professionals randomized to clinical vignettes that varied only by gender and race found evidence of gender and race bias in the decision-making process for offering advanced therapies for heart failure, particularly for African American women patients, who were judged more harshly by appearance and adequacy of social support. There was no associated between patient gender and race and final recommendations for allocation of advanced therapies. However, it is possible that bias may contribute to delayed allocation and ultimately inequity in the allocation of advanced therapies in a clinical setting.

    Introduction

    Heart failure therapies are inequitably allocated to minority racial groups in the US.1,2 Despite African American individuals representing the highest racial risk group for heart failure incidence and mortality,3 they are less likely to receive defibrillators4,5 and care by a cardiologist compared with white individuals.6 Racial disparities are not fully explained by socioeconomic factors,4-6 and may be related to bias. A 2019 study7 suggested that patient race is associated with the allocation process for advanced heart failure therapies, heart transplantations, and ventricular assist devices (VADs). Among vignettes of men patients with identical clinical and social histories, white men were favored over African American men for allocation to heart transplants.7 Variability in clinicians’ assessments of social support and adherence contributed to the racial bias.7

    Advanced heart failure therapies are also disparately allocated by gender.2,8 Although women have a higher prevalence of heart failure than men,3 women receive less than one-quarter of heart transplants and VADs.2,8 This may be related to a higher prevalence of heart failure with preserved ejection fraction in women than in men,3 which is rarely an indication for transplantation and not an indication for VADs. However, more than 40% of heart failure hospitalizations in white women are attributed to heart failure with reduced ejection fraction,3,9 and African American women have higher rates of heart failure with reduced ejection fraction than preserved ejection fraction.9 In the US, women receive fewer advanced heart therapies.10,11 The reason for the gender disparity in advanced heart failure therapies is unknown.

    Patient race and gender are known to be associated with clinical decision-making.12,13 With complex decision-making processes like the allocation of advanced heart failure therapies, it is unknown whether the intersection of gender and race is associated with decision-making among health care professionals. It is unclear how gender and race are associated with subjective assessments in the evaluation for advanced therapies. This study used interviews and supplemental surveys to examine the decision-making process in a controlled setting with identical clinical vignettes that varied only by gender and race. This study addressed the overriding hypotheses that African American women would be evaluated more harshly and would be less likely to be offered heart transplantation than white women or men.

    Methods
    Study Design, Sample, and Recruitment

    In a simultaneous mixed-methods study performed in April 2019, US members of the International Society for Heart and Lung Transplantation (ISHLT) were interviewed at ISHLT scientific sessions and administered supplemental surveys. ISHLT is composed of clinicians, allied health professionals, scientists, and trainees. Eligible members included US-based health care professionals who engage in clinical decision-making for the allocation of advanced heart failure therapies. Participants were identified through the annual scientific sessions program and ISHLT directory. Participants were purposefully selected for diversity of participant demographic characteristics (ie, gender, race, geography, and years past training). Snowball sampling was used to meet an enrollment goal of 44 participants because it was sufficient to reach 1:1 thematic saturation in a prior study7 and snowball sampling is an established method of recruiting specific hard-to-reach subgroups.14 Study participants provided verbal informed consent and received monetary incentives worth US $10 for participating. Trained research assistants performed, audio recorded, and collected field notes for all interviews. This study followed the Standards for Reporting Qualitative Research (SRQR) reporting guideline.15 The University of Arizona institutional review board approved this study.

    Vignettes

    Participants were masked to study objectives until participation was complete. Participants were randomized 1:1 to vignettes of an African American woman or a white woman to compare with a prior study including only vignettes of African American men and white men.7 On each of the 3 interview days, 2 participants were randomized to a vignette of a male patient as a data check to compare with a prior study using only vignettes of men, in which patients with similar body mass index (calculated as weight in kilograms divided by height in meters squared) differed slightly in height. Gender and race were indicated by photographs, text, and names associated with ethnic or gender identity.16 The 4 study photographs included an African American woman, a white woman, an African American man, and a white man. Within each gender, the hairstyle and physical build were similar. Across all 4 photographs, clothing was similar, and photos were previously rated similarly for age, attractiveness, intelligence, health, facial expression, and trustworthiness in a normalization study (eTable 1 in the Supplement and prior work7). Clinical vignettes were otherwise identical with the exception of gender-based evaluations (normal mammogram and shorter height for women and normal prostate specific antigen for men; eTable 2 in the Supplement). In brief, vignettes described patients with end-stage heart failure with complex history, including multiple relative contraindications for advanced therapies.

    Interviews

    Interviews were conducted using the think-aloud method, which elicits verbalization of conscious and unconscious thoughts.17 Think-aloud is an established method of understanding the decision-making process.17 Participants were directed to articulate their thoughts as they read through each section of the vignette during a 30-minute interview. Prompts from the Shafer and Lohse cognitive interviewing guide18 were used to aid participants in verbalizing their thoughts, such as “Tell me what you are thinking. What thoughts are going through your mind right now?” Participants were asked how each section of the vignette influenced their recommendation for advanced heart failure therapies. On conclusion of discussing the vignette, participants were asked to share any thoughts that they had not previously shared, provide a final recommendation for the patient in the clinical vignette, and describe their reasons for the recommendation. Then participants were asked whether they trusted the patient in the clinical vignette and how the vignette compared with patient presentations at their respective centers. The entire interview guide was duplicated from a previously published study7 because the other guide had been rigorously tested and provided opportunity for comparison of studies.

    Survey Instrument

    Supplemental surveys were provided to participants after completing interviews. Using a Likert scale (1-10, strongly disagree to strongly agree), participants were asked to rate how each section of the vignette influenced the patient’s suitability for advanced heart failure therapies. Then participants were asked to rate the suitability for each advanced heart failure therapy (eg, heart transplantation, bridge-to-transplantation VAD, destination VAD, no advanced therapy) and need for additional testing or consultation (1 indicating strong disagreement that a therapy would be appropriate and 10, strong agreement). Participants were also asked to provide write-in responses that supported their decisions. The demographic characteristics of the participants were also collected. The survey questions were duplicated from a previously published study.7

    Statistical Analysis

    Thematic analysis of think-aloud interviews was performed using grounded theory while masked to patient gender and race. Results were then unmasked and categorized according to gender and race. The stepwise process of grounded theory19 included: (1) open coding, ie, identifying tentative themes until reaching saturation (no new ideas)20; (2) identifying a central phenomenon, ie, combining themes to form a central category; (3) axial coding, ie, identifying associations between themes; and (4) selective coding, ie, supporting themes with specific codes. The result of this approach is a model that represents the decision-making process. Rigor was established through credibility and triangulation (ie, validation of interview response with survey response), transferability (ie, debrief with an advanced heart failure cardiologist [N.S., a white woman]), confirmability (including trained interviewers with bachelor’s degrees [E.Y., an Asian American and white woman; R.H.Y., an Asian American and white man] and 2 independent faculty analysts with doctorate degrees [N.P., a white woman and M.H., a white woman] who performed the entire qualitative analyses with differences arbitrated by an independent qualitative expert [J.C., a white woman] and the primary investigator, an advanced heart failure cardiologist [K.B., an African American woman]).21 Neither the interviewers nor the analysts had relationships with study participants. An audit trail and codebook were maintained throughout the study.

    Participant survey demographic characteristics were compared across race within each gender using the Fisher exact test for categorical data. Mean results of the individual survey questions were compared within vignettes of women patients using a 2-tailed Wilcoxon test to compare with overall findings from qualitative analyses and adjusted for multiple comparisons with false-discovery rate procedure, setting significance at P < .05.22 For hypothesis generation, a secondary analysis was performed to examine the means and standard errors of this study that included a prior study7 conducted by the principal investigator (K.B.) examining the responses of 44 participants who were randomized 1:1 to vignettes of an African American man and a white man in interviews and surveys. Groups were not statistically compared, as some individuals participated in both studies (12 participants in the survey; 6 participants in the study interview and survey).

    Results

    Among US-based members of an international heart transplantation organization, 46 participants (24 [52%] women, 20 [43%] racial minority) were randomized and assigned to clinical vignettes of an African American woman (20 participants [43%]), a white woman (20 participants [43%]), an African American man (3 participants [7%]) and a white man (3 participants [7%]) (Table 1). Most participants were in the cardiology (31 participants [67%]) and cardiothoracic surgery (7 participants [15%]) fields. Participants represented 10 of the 11 US regions for heart transplantation. There were no statistically significant differences in participant demographic characteristics within the patient vignette subsamples of women and men patients.

    Think-Aloud Interview Results

    The allocation decision-making process of participants was found to be defined by a central phenomenon: is the heart sick enough, is the body well enough, and is there enough social and/or emotional support to receive advanced therapies? Participants allocated therapies by sequential steps following major themes of implicit bias: (1) forming an overall impression: critiquing appearance of women more harshly; (2) identifying urgency: believing the African American man had more severe illness than the white man; (3) evaluating appropriateness of prior care: believing the African American woman’s care was less appropriate than the white woman; (4) anticipating challenges: believing similar forms of social support were adequate for men and inadequate for women, particularly African American women; and (5) evaluating trust and making an ultimate recommendation: recommending a VAD over transplantation for all subgroups (Table 2). Subthemes varied by race and gender of the clinical vignette including positive, negative, and neutral observations (eTable 3 in the Supplement).

    Forming an Overall Impression: Critiquing Appearance of Women More Harshly

    Vignette photographs were not welcomed by all participants. Some participants avoided dwelling on the photograph and instead developed clinical assessments from the histories in addition to photographs. Others had strong reactions to the photos, particularly those of women patients. Negative impressions were focused on the visual appearance of the women in the vignettes, including age, weight, hair, makeup, and facial expression.

    “When I see her, I don’t think of her as totally friendly, I see someone who’s a little unkempt to some extent, and yet she’s probably fine and normal. That was my immediate reaction of not wanting the photo. Because I’m now making a decision about whether she is capable of taking care of herself or a machine based on her hair. Maybe there’s value there, but that’s not how we do it. And so, I was upset to see the photos. So anyway, that’s my feeling” (white woman participant/patient vignette of white woman).

    Identifying Urgency: Believing the African American Man Had More Severe Illness Than Other Patients

    Participants believed that the patients had clinical histories that warranted urgent evaluations for heart failure therapies. However, participants believed that the African American man had more severe illness than patients in the other vignette categories. Both the white and African American women were perceived as equally sick.

    “His risk of death in the next few months is very high. Again, still can’t decide his eligibility, but it means it may be indicated in an urgent setting for transplantation down the road” (white man participant/patient vignette of African American man).

    Evaluating Appropriateness of Prior Care: Believing the Setting for the African American Woman’s Care Was Less Appropriate Than the White Woman’s

    Participants were concerned about whether appropriate treatment was provided prior to referral for advanced heart failure therapies. Prior care was perceived to be inadequate for the vignettes of women patients. Concerns for inappropriate care were greater for the African American woman compared with the vignette of the white woman.

    “It’s a shame that this lady was only diagnosed 2 years ago. I mean I get angry about that. I mean particularly being a [minority] provider, I see that many patients that are referred to me regardless of their race tend to be referred late from a heart failure standpoint. I find that my minority patients, particularly my African American patients, are referred even later.… Many times it’s because their symptoms were going unrecognized by the people that were taking care of them … or their symptoms weren’t believed.… They tell me many stories and I’m hoping that this isn’t the case for her but unfortunately if you see it enough times … it starts to dishearten you” (minority man participant/patient vignette of African American woman).

    Anticipating Challenges: Believing Similar Forms of Social Support Were Adequate for Men and Inadequate for Women, Particularly African American Women

    Participants were concerned about medical comorbidities, such as peripheral arterial disease and chronic kidney disease, irrespective of patient vignette category. However, the assessment of social support differed by vignette gender. Participants believed that the women patients were the primary caregivers for children and found the spouse an inadequate form of support; men patients were perceived as having more adequate caregiver support. Children were considered liabilities, particularly for the vignette of the African American woman. Participants also had more concerns regarding family dynamics and the adequacy of finances when examining the vignette of the African American woman.

    “Is the husband willing to help? She’s married, that certainly doesn’t mean that he’s necessarily going to be the person bringing her to appointments or helping.… I mean, just making sure that there’s the social support to make sure that she’s successful, and that we’re not hurting her more by doing anything else” (white woman participant/patient vignette of African American woman).

    “Because she’s African American, it sounds like her socioeconomic status is not the greatest—1 car, she’s working, disability; those sorts of things that make me … think that she’s probably not as socioeconomically stable as other patients. Especially, since she lacked health care insurance a couple of times” (white woman participant/patient vignette of African American woman).

    “I’m just worried about the fact that she’s taking care of little kids. She does have a spouse who will be her caregiver and so she does have support but I mean thus far nothing is a red flag, it’s just concern that those 2 kids need to be taken care of but she’s in the hospital” (minority man participant/patient vignette of African American woman).

    Evaluating Trust and Making the Ultimate Recommendation: Recommending VADs Over Transplantation for All Subgroups

    Trust was based on interpersonal interactions and patient behavior. Adherence had less of a role in evaluating trust. Adherence was considered a social determinant of health, and participants felt that the women participants deserved the benefit of doubt. Overall, participants supported providing inotropes to patients to improve kidney function with the intention of offering VADs. A VAD instead of transplantation was recommended in vignettes irrespective of race or gender category.

    “It’s just really challenging. I mean, you meet a person and they’re underinsured. They have a family, they’re a regular person, right? It’s easier when someone doesn’t have anyone; they’re divorced or they’ve never had kids and … they failed completely on the social side, but she’s got enough soft things that will make it challenging, and in our program, we would say, ‘We’ll take a chance on her for a VAD but not for a heart’” (white man participant/patient vignette of white woman).

    Supplemental Survey Results

    When comparing results only for the vignettes of women patients, final recommendations for each treatment option were similar for the African American woman and white woman (mean [standard error] Likert score rating, heart transplantation: 7.53 [0.59] vs 6.53 [0.61]; P = .56; bridge-to-transplant VAD: 8.11 [0.58] vs 7.78 [0.57]; P = .79; destination therapy: 7.00 [0.77] vs 7.48 [0.63]; P > .99; no therapies: 1.40 [0.18] vs 1.43 [0.25]; P > .99) (Table 3). There were no significant differences in participants’ ratings of all clinical and social factors fully adjusted for multiple comparisons (Table 3).

    In the secondary analysis, survey results were combined from this study (46 participants) and a prior study of 44 health care professionals who were randomized to either a vignette of an African American man or a white man and who completed interviews (eTable 4 in the Supplement).7 Recommendations for heart transplantation were lower for white women than for other groups (eg, 6.53 [0.61] vs for African American man, 7.35 [0.49]). VAD recommendations were higher for African American men than other groups (eg, bridge-to-transplant VAD: 8.21 [0.34] vs for white man, 7.70 [0.50]; destination therapy: 7.68 [0.56] vs 6.80 [0.57]). When isolating factors of race/ethnicity and gender, support for advanced therapies were higher for African American women than other groups (eg, no therapy: 1.40 [0.18] vs for white man, 1.96 [0.26]).

    Discussion

    In a national study of clinicians, both patient gender and race were associated with bias in the decision-making process for heart transplantation allocation despite patients having identical case presentations. Compared with the vignettes of men, women patients were judged more harshly according to their appearance. The African American man was thought to have more severe illness than the white man and the women of both races. Social support assessments were more critical when the patient was a woman, particularly an African American woman. Spouses were often deemed inadequate forms of support when the patient was a woman. Children were considered liabilities when the patient was a woman. Ultimately, VADs were recommended over transplantation for patients irrespective of patient race and gender during the interviews and no significant differences were observed in treatment recommendations with surveys, but differences were found between vignettes that might negatively affect the delivery of care. To our knowledge, this is the first national study to find that subjective assessments can be influenced by gender and race biases in the allocation of advanced heart failure therapies.

    This contemporary study aligns with a historical body of reports and studies, while extending these findings to decision-making in advanced heart failure. In 1991, the Association of American Medical Colleges Council on Ethical and Judicial Affairs13 identified gender bias as an etiology of variability in clinician decision-making and provided a call to action for multiple leading medical organizations. Approximately a decade later, a study of physicians23 revealed reduced likelihood of offering a cardiac catheterization to African American women compared with white men despite having identical clinical histories. A 201724 study observed a significant correlation between level of implicit bias and likelihood of a cardiologist recommending a cardiac catheterization to a woman with the same clinical history as a man. While gender bias has been increasingly addressed as a public health issue, little progress has been achieved.25

    Ultimate recommendations for advanced heart therapies differed from a 2019 study7 that favored transplantation over VAD in white men compared with African American men with identical clinical histories. In this study, which focused on African American and white women patients, clinicians recommended VADs more than transplantation for all patients during interviews. While the secondary analysis of survey results from the prior study combined with this study revealed no major differences in treatment recommendations for any gender or race group, recommendation differences in the interview conversations are noteworthy. The allocation of advanced heart failure therapies is distinct from cardiac catheterization referral processes, which can be recommended by an individual health care professional. Similar to other organ replacement therapies, a multidisciplinary team of health care professionals decide who is appropriate for advanced heart therapy. Conversations rather than surveys guide these allocation meetings, and more dominant members may lead the conversations and control the final decision.7 In addition, the differences in concerns by gender and race could contribute to a delay in treatment, which could worsen outcomes.

    The differences between reality and bias are important philosophical questions raised by this study. Assumptions raised by health care professionals in this study reflect what is observed and valued in society. According to the Family Caregiver Alliance,26 between 53% and 68% of caregivers are women. According to a 2017 study,27 the most valued traits in the US are appearance in women and honesty in men. Health care professionals were more critical of the idea of a male caregiver and of the appearance of women. It is easy to infer that participants were providing assessments based on their reality, but the danger lies in applying a stereotype to a population or individual. Stereotypes often yield bias and lead to health inequity.28 These gender and racial biases may be the underlying reasons for unequitable allocation of advanced heart failure therapies by gender and race.

    There is an urgent need to standardize assessment of subjective criteria for advanced heart therapies. Guidelines provide objective criteria for determining who has severe enough illness to consider advanced heart therapies.29,30 However, criteria for many of the relative contraindications for advanced therapies are subjective and variable.29,30 More subjective relative contraindications, namely social support and adherence, are sources of potential gender and racial biases in the allocation of advanced heart therapies.12,13,31 Subjective rather than objective assessments may be used partially because of limited data regarding these contraindications.32 A number of different objective measures for social support and adherence exist, but most evidence comes from single-center studies and other organ replacements.32 Thus, the governing bodies for heart and lung transplantation and social work recommend that standardized psychosocial evaluations be routinely performed among patients being considered for advanced heart therapies, but do not elaborate which measures of social support should be used or the data supporting their use.33 Important areas for future investigation include generating multicenter data on the use of objective social assessments for advanced heart therapy allocation.

    Additional steps can be taken to reduce the influence of bias during advanced heart therapy allocation. As demonstrated in this study, health care professionals had altruistic plans for each patient, but patient gender and race influenced the decision-making process. Bias reduction training may create a culture of equity for patients with advanced heart failure, particularly among individuals with a desire to parallel their behavior with their belief systems.34 Bias training has been effective in reducing implicit bias and changing behavior among health care professionals.35,36 Small systemic changes may also improve health equity.12 Participants in this study identified patient photographs and descriptions of children as potential sources of bias. Steps can be taken to avoid including these data during routine presentations of candidates for advanced therapies. Finally, this study does not address shared decision-making between the patient and health care professional; assuring that patients are well educated about their options can assist the health care professional in advocating on their behalf.37

    Limitations

    This study was subject to several limitations. First, the single clinical vignette used does not represent the full spectrum of possible clinical presentations. However, these vignettes represented patients with multiple relative contraindications for advanced heart therapies, which uniformly presented a picture of advanced disease while creating uncertainty about the appropriateness of advanced therapies. Such a vignette is appropriate for understanding how patient gender and race may influence health care professionals’ decisions.38 Second, this study included a small number of participants focused on vignettes of men patients and was not designed to have an equal number of vignettes by gender. This limits the power to evaluate significant differences in the supplementary quantitative surveys, including the interaction of participant demographic characteristics, and limits generalizability, which is inherent to qualitative design. However, this sample was large enough for the qualitative analysis, representing more participants than typically used in qualitative studies.21 Sample size was appropriate by thematic saturation analysis. Furthermore, some of the participants previously participated in a similar study evaluating decision-making for African American and white men patients. They may have suspected the objectives of this study and may have been subject to the Hawthorne effect by providing socially desirable results under observation; however, that would bias results toward the null of no differences in the evaluation of differences in gender and race. No participants were aware of results of the prior study of only men patients.

    Conclusions

    In a study of US health care professionals, bias related to patient gender and race was present in the process of allocating advanced heart failure therapies. Women patients, particularly African American women patients, were judged more harshly for their appearance and degree of social support than men of both races with identical clinical and social histories. Final recommendations in these interviews supported VADs over transplantation, irrespective of patient gender and race. Bias related to gender and race could lead to delayed allocation and inequity in patient outcomes. Further investigation and implementation of bias reduction strategies are needed.

    Back to top
    Article Information

    Accepted for Publication: April 28, 2020.

    Published: July 21, 2020. doi:10.1001/jamanetworkopen.2020.11044

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2020 Breathett K et al. JAMA Network Open.

    Corresponding Author: Khadijah Breathett, MD, MS, Sarver Heart Center, Division of Cardiology, Department of Medicine, University of Arizona, 1501 N Campbell Ave, PO Box 245046, Tucson, AZ 85724 (kbreathett@shc.arizona.edu).

    Author Contributions: Dr Breathett had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Breathett.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Breathett.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Breathett, Pool, Hebdon, Crist, Knapp.

    Obtained funding: Breathett.

    Administrative, technical, or material support: E. Yee, R. Yee, Solola, Luy, Herrera-Theut, Zabala.

    Supervision: Sweitzer, Crist, Stone, Calhoun.

    Conflict of Interest Disclosures: None reported.

    Funding/Support: Dr Breathett received support from grants K01HL142848, L30HL148881, and NHLBI 2R25HL126146-05 Subaward 11692sc from National Heart, Lung, and Blood Institute; the University of Arizona Health Sciences, Strategic Priorities Faculty Initiative Grant; and support from the University of Arizona, Sarver Heart Center, Women of Color Heart Health Education Committee. Dr Hebdon received support from grant T32NR013456 from National Institutes of Health. Mr Luy received support from grant R25HL108837 from the University of Arizona’s Summer Institute on Medical Ignorance.

    Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    References
    1.
    Breathett  K, Allen  LA, Helmkamp  L,  et al.  The Affordable Care Act Medicaid expansion correlated with increased heart transplant listings in African Americans but not Hispanics or Caucasians.   JACC Heart Fail. 2017;5(2):136-147. doi:10.1016/j.jchf.2016.10.013PubMedGoogle ScholarCrossref
    2.
    Breathett  K, Allen  LA, Helmkamp  L,  et al.  Temporal trends in contemporary use of ventricular assist devices by race and ethnicity.   Circ Heart Fail. 2018;11(8):e005008. doi:10.1161/CIRCHEARTFAILURE.118.005008PubMedGoogle Scholar
    3.
    Benjamin  EJ, Muntner  P, Alonso  A,  et al; American Heart Association Council on Epidemiology and Prevention Statistics Committee and Stroke Statistics Subcommittee.  Heart disease and stroke statistics, 2019 update: a report from the American Heart Association.   Circulation. 2019;139(10):e56-e528. doi:10.1161/CIR.0000000000000659PubMedGoogle ScholarCrossref
    4.
    Farmer  SA, Kirkpatrick  JN, Heidenreich  PA, Curtis  JP, Wang  Y, Groeneveld  PW.  Ethnic and racial disparities in cardiac resynchronization therapy.   Heart Rhythm. 2009;6(3):325-331. doi:10.1016/j.hrthm.2008.12.018PubMedGoogle ScholarCrossref
    5.
    Mezu  U, Ch  I, Halder  I, London  B, Saba  S.  Women and minorities are less likely to receive an implantable cardioverter defibrillator for primary prevention of sudden cardiac death.   Europace. 2012;14(3):341-344. doi:10.1093/europace/eur360PubMedGoogle ScholarCrossref
    6.
    Breathett  K, Liu  WG, Allen  LA,  et al.  African Americans are less likely to receive care by a cardiologist during an intensive care unit admission for heart failure.   JACC Heart Fail. 2018;6(5):413-420. doi:10.1016/j.jchf.2018.02.015PubMedGoogle ScholarCrossref
    7.
    Breathett  K, Yee  E, Pool  N,  et al.  Does race influence decision making for advanced heart failure therapies?   J Am Heart Assoc. 2019;8(22):e013592. doi:10.1161/JAHA.119.013592PubMedGoogle Scholar
    8.
    Colvin  M, Smith  JM, Hadley  N,  et al.  OPTN/SRTR 2017 annual data report: heart.   Am J Transplant. 2019;19(suppl 2):323-403. doi:10.1111/ajt.15278PubMedGoogle ScholarCrossref
    9.
    Chang  PP, Chambless  LE, Shahar  E,  et al.  Incidence and survival of hospitalized acute decompensated heart failure in four US communities (from the Atherosclerosis Risk in Communities Study).   Am J Cardiol. 2014;113(3):504-510. doi:10.1016/j.amjcard.2013.10.032PubMedGoogle ScholarCrossref
    10.
    DeFilippis  EM, Truby  LK, Garan  AR,  et al.  Sex-related differences in use and outcomes of left ventricular assist devices as bridge to transplantation.   JACC Heart Fail. 2019;7(3):250-257. doi:10.1016/j.jchf.2019.01.008PubMedGoogle ScholarCrossref
    11.
    Hsich  EM.  Sex differences in advanced heart failure therapies.   Circulation. 2019;139(8):1080-1093. doi:10.1161/CIRCULATIONAHA.118.037369PubMedGoogle ScholarCrossref
    12.
    Breathett  K, Jones  J, Lum  HD,  et al.  Factors related to physician clinical decision-making for African American and Hispanic patients: a qualitative meta-synthesis.   J Racial Ethn Health Disparities. 2018;5(6):1215-1229. doi:10.1007/s40615-018-0468-zPubMedGoogle ScholarCrossref
    13.
    McMurray  RJ, Clarke  OW, Barrasso  JA,  et al.  Gender disparities in clinical decision making—Council on Ethical and Judicial Affairs, American Medical Association.   JAMA. 1991;266(4):559-562. doi:10.1001/jama.1991.03470040123034PubMedGoogle ScholarCrossref
    14.
    Sadler  GR, Lee  H-C, Lim  RS, Fullerton  J.  Recruitment of hard-to-reach population subgroups via adaptations of the snowball sampling strategy.   Nurs Health Sci. 2010;12(3):369-374. doi:10.1111/j.1442-2018.2010.00541.xPubMedGoogle ScholarCrossref
    15.
    O’Brien  BC, Harris  IB, Beckman  TJ, Reed  DA, Cook  DA.  Standards for reporting qualitative research: a synthesis of recommendations.   Acad Med. 2014;89(9):1245-1251. doi:10.1097/ACM.0000000000000388PubMedGoogle ScholarCrossref
    16.
    Bertrand  M, Mullainathan  S.  Are Emily and Greg more employable than Lakisha and Jamal? a field experiment on labor market discrimination. National Bureau of Economic Research. Published July 2003. Accessed December 12, 2016. https://www.nber.org/papers/w9873
    17.
    Li  AC, Kannry  JL, Kushniruk  A,  et al.  Integrating usability testing and think-aloud protocol analysis with “near-live” clinical simulations in evaluating clinical decision support.   Int J Med Inform. 2012;81(11):761-772. doi:10.1016/j.ijmedinf.2012.02.009PubMedGoogle ScholarCrossref
    18.
    Shafer  K, Lohse  B.  How to conduct a cognitive interview: a nutritional education example. National Institute of Food and Agriculture. Published January 21, 2015. Accessed June 17, 2020. https://nifa.usda.gov/resource/how-conduct-cognitive-interview-nutrition-education-example
    19.
    Strauss  A, Corbin  JM.  Basics of Qualitative Research: Grounded Theory Procedures and Techniques. Sage Publications, Inc; 1990.
    20.
    Curry  LA, Nembhard  IM, Bradley  EH.  Qualitative and mixed methods provide unique contributions to outcomes research.   Circulation. 2009;119(10):1442-1452. doi:10.1161/CIRCULATIONAHA.107.742775PubMedGoogle ScholarCrossref
    21.
    Creswell  JW.  Qualitative Inquiry and Research Design: Choosing Among Five Approaches. SAGE Publications; 2012.
    22.
    Benjamini  Y, Hochberg  Y.  Controlling the false discovery rate: a practical and powerful approach to multiple testing.   J R Stat Soc (Series B). 1995;57(1):289-300. doi:10.1111/j.2517-6161.1995.tb02031.xGoogle Scholar
    23.
    Schulman  KA, Berlin  JA, Harless  W,  et al.  The effect of race and sex on physicians’ recommendations for cardiac catheterization.   N Engl J Med. 1999;340(8):618-626. doi:10.1056/NEJM199902253400806PubMedGoogle ScholarCrossref
    24.
    Daugherty  SL, Blair  IV, Havranek  EP,  et al.  Implicit gender bias and the use of cardiovascular tests among cardiologists.   J Am Heart Assoc. 2017;6(12):e006872. doi:10.1161/JAHA.117.006872PubMedGoogle Scholar
    25.
    Heise  L, Greene  ME, Opper  N,  et al; Gender Equality, Norms, and Health Steering Committee.  Gender inequality and restrictive gender norms: framing the challenges to health.   Lancet. 2019;393(10189):2440-2454. doi:10.1016/S0140-6736(19)30652-XPubMedGoogle ScholarCrossref
    26.
    American Psychological Association.  Who are family caregivers? Published 2011. Accessed June 3, 2020. https://www.apa.org/pi/about/publications/caregivers/faq/statistics
    27.
    Parker  K, Horowitz  JM, Stepler  R.  2: Americans see different expectations for men and women. Pew Research Center’s Social and Demographic Trends Project. December 5, 2017. Accessed April 20, 2020. https://www.pewsocialtrends.org/2017/12/05/americans-see-different-expectations-for-men-and-women/
    28.
    Smedley  B, Stith  A, Nelson  A.  Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Institute of Medicine; 2003. doi:10.17226/12875
    29.
    Feldman  D, Pamboukian  SV, Teuteberg  JJ,  et al; International Society for Heart and Lung Transplantation.  The 2013 International Society for Heart and Lung Transplantation guidelines for mechanical circulatory support: executive summary.   J Heart Lung Transplant. 2013;32(2):157-187. doi:10.1016/j.healun.2012.09.013PubMedGoogle ScholarCrossref
    30.
    Mehra  MR, Kobashigawa  J, Starling  R,  et al.  Listing criteria for heart transplantation: International Society for Heart and Lung Transplantation guidelines for the care of cardiac transplant candidates, 2006.   J Heart Lung Transplant. 2006;25(9):1024-1042. doi:10.1016/j.healun.2006.06.008PubMedGoogle ScholarCrossref
    31.
    Hart  M.  Subjective decisionmaking and unconscious discrimination.   Ala Rev. 2004;56:741-792. Accessed June 29, 2020. https://scholar.law.colorado.edu/cgi/viewcontent.cgi?article=1414&context=articlesGoogle Scholar
    32.
    Bui  QM, Allen  LA, LeMond  L, Brambatti  M, Adler  E.  Psychosocial evaluation of candidates for heart transplant and ventricular assist devices: beyond the current consensus.   Circ Heart Fail. 2019;12(7):e006058. doi:10.1161/CIRCHEARTFAILURE.119.006058PubMedGoogle Scholar
    33.
    Dew  MA, DiMartini  AF, Dobbels  F,  et al.  The 2018 ISHLT/APM/AST/ICCAC/STSW recommendations for the psychosocial evaluation of adult cardiothoracic transplant candidates and candidates for long-term mechanical circulatory support.   J Heart Lung Transplant. 2018;37(7):803-823. doi:10.1016/j.healun.2018.03.005PubMedGoogle ScholarCrossref
    34.
    Levy  N, Harmon-Jones  C, Harmon-Jones  E.  Dissonance and discomfort: does a simple cognitive inconsistency evoke a negative affective state?   Motiv Sci. 2018;4(2):95-108. doi:10.1037/mot0000079Google ScholarCrossref
    35.
    Stone  J, Moskowitz  GB, Zestcott  CA, Wolsiefer  KJ.  Testing active learning workshops for reducing implicit stereotyping of Hispanics by majority and minority group medical students.   Stigma Health. 2019;5(1):94-103. doi:10.1037/sah0000179Google ScholarCrossref
    36.
    Carnes  M, Devine  PG, Baier Manwell  L,  et al.  The effect of an intervention to break the gender bias habit for faculty at one institution: a cluster randomized, controlled trial.   Acad Med. 2015;90(2):221-230. doi:10.1097/ACM.0000000000000552PubMedGoogle ScholarCrossref
    37.
    Alfandre  D.  Clinical recommendations in medical practice: a proposed framework to reduce bias and improve the quality of medical decisions.   J Clin Ethics. 2016;27(1):21-27.PubMedGoogle Scholar
    38.
    Balsa  AI, McGuire  TG.  Prejudice, clinical uncertainty and stereotyping as sources of health disparities.   J Health Econ. 2003;22(1):89-116. doi:10.1016/S0167-6296(02)00098-XPubMedGoogle ScholarCrossref
    ×