Does Race Influence Decision Making for Advanced Heart FailureTherapies?Khadijah Breathett, MD, MS; Erika Yee, BS; Natalie Pool, PhD, RN; Megan Hebdon, DNP, RN; Janice D. Crist, PhD, RN; Shannon Knapp,PhD; Ashley Larsen, MS; Sade Solola, MD; Luis Luy; Kathryn Herrera-Theut, BS; Leanne Zabala, MD; Jeff Stone, PhD; Marylyn M. McEwen,PhD, RN; Elizabeth Calhoun, PhD; Nancy K. Sweitzer, MD, PhD
Background-—Race influences medical decision making, but its impact on advanced heart failure therapy allocation is unknown.We sought to determine whether patient race influences allocation of advanced heart failure therapies.
Methods and Results-—Members of a national heart failure organization were randomized to clinical vignettes that varied bypatient race (black or white man) and were blinded to study objectives. Participants (N=422) completed Likert scale surveys ratingfactors for advanced therapy allocation and think-aloud interviews (n=44). Survey results were analyzed by least absoluteshrinkage and selection operator and multivariable regression to identify factors influencing advanced therapy allocation, includinginteractions with vignette race and participant demographics. Interviews were analyzed using grounded theory. Surveys revealedno differences in overall racial ratings for advanced therapies. Least absolute shrinkage and selection operator regression selectedno interactions between vignette race and clinical factors as important in allocation. However, interactions between participantsaged ≥40 years and black vignette negatively influenced heart transplant allocation modestly (�0.58; 95% CI, �1.15 to �0.0002),with adherence and social history the most influential factors. Interviews revealed sequential decision making: forming overallimpression, identifying urgency, evaluating prior care appropriateness, anticipating challenges, and evaluating trust while makingrecommendations. Race influenced each step: avoiding discussing race, believing photographs may contribute to racial bias,believing the black man was sicker compared with the white man, developing greater concern for trust and adherence with theblack man, and ultimately offering the white man transplantation and the black man ventricular assist device implantation.
Conclusions-—Black race modestly influenced decision making for heart transplant, particularly during conversations. Becauseadvanced therapy selection meetings are conversations rather than surveys, allocation may be vulnerable to racial bias. (J AmHeart Assoc. 2019;8:e013592. DOI: 10.1161/JAHA.119.013592.)
Key Words: decision making • healthcare delivery • healthcare disparities • heart failure • heart transplant
Racial disparities persist in heart failure (HF).1 Blacks
have the highest rates of HF and greatest mortalitycompared with other ethnicities.1 Compared with whites,blacks are less likely to receive HF medications2 and devicetherapies3 and less likely to receive care by a cardiologist.4
Insurance broadening has contributed to increased accessto heart transplants among blacks but has not fullyeliminated disparities.5 Ventricular assist device (VAD)
implantation rates also remain below expected rates forblacks.6
Healthcare professionals’ decision-making processes maycontribute to racial disparities in HF. A recent meta-synthesis ofmedical qualitative studies concluded that physicians’ clinicaldecisions were influenced by patient ethnicity.7 However, therole of race in the decision-making process for advanced HFtherapies, such as heart transplants and VAD, is unknown.
From the Division of Cardiovascular Medicine, Department of Medicine, Sarver Heart Center (K.B., N.S.), Sarver Heart Center, Clinical Research Office (E.Y, A.L.), College ofNursing (N.P., M.H., J.D.C., M.M.M.), Statistics Consulting Lab, Bio5 Institute (S.K.), Departments of Medicine (S.S.) and Psychology (J.S.), and Center for Population HealthSciences (E.C.), University of Arizona, Tucson, AZ; University of Rochester, Rochester, New York, U.S. (L.L.); and University of Arizona Medical School, Tucson, AZ (K.H.-T., L.Z.).
Accompanying Tables S1 through S4 and Figure S1 are available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.119.013592
This work was presented at the American Heart Association Scientific Sessions, November 16–18, 2019, in Philadelphia, PA.
Correspondence to: Khadijah Breathett, MD, MS, Sarver Heart Center, University of Arizona, 1501 N Campbell Ave, PO Box 245046, Tucson, AZ 85724. E-mail:kbreathett@shc.arizona.edu
Received June 12, 2019; accepted August 9, 2019.
ª 2019 The Authors. Published on behalf of the American Heart Association, Inc., by Wiley. This is an open access article under the terms of the Creative CommonsAttribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 1
ORIGINAL RESEARCH
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
The decision-making process for advanced therapies for HFis complex. Healthcare professionals must determine the besttherapy for patients while also carefully assessing for con-traindications for advanced HF therapies. Most contraindica-tions to advanced therapies exist on a spectrum, and the pointat which advanced therapies should not be pursued issubjective.8,9 Subjectivity at these decision points createsvulnerability for racial bias.10
We sought to determine whether race influences thedecision-making process for advanced HF therapies. Quanti-tative surveys and qualitative think-aloud interviews wereperformed to understand the following: (1) general decisionprocess, (2) how race influences the decision process, and (3)which factors have the greatest influence on the decisionprocess when considering patients of black and white race.
Methods
Study Design, Sample, and RecruitmentA simultaneous mixed-methods approach was performed fromSeptember 2018 to February 2019 among members of theHeart Failure Society of America, a national organization of HFprofessionals, including postdoctoral trainees, nurses, phar-macists, pharmaceutical representatives, nurse practitioners,and physicians. Eligible members (N=1432) included allhealthcare professionals who participate in the decision-making process for advanced HF therapies in adults in theUnited States. Participants were identified through member-ship directory (N=1929), but advanced therapy allocation roleswere not available and exclusion estimates may be underes-timated. Self-report during consent (n=180), inaccurate e-mailaddress (n=49), and electronic search (n=268) identified a total
of 497 ineligible participants. A census approach was used, inwhich electronic surveys were e-mailed to all members inrepeated waves through Qualtrics. Among invitations, 422participated (29% response rate), which is consistent withsurvey responses for healthcare professionals11,12 and higherthan cardiovascular society survey response rates.13 This wouldalso provide a priori 99% power to detect 0.9-unit difference inscores (1–10) with 1-unit SD for supporting advanced therapiesin black versus white men.14 Simultaneously, in-person orvideoconference think-aloud interviews were performed withsurveying among a purposeful sample of members representingdiverse ethnicities, sexes, position, and geographic institution.Snowball sampling was used to help identify 44 professionalsfor interviews.15 Trained research assistants (E.Y., A.L.)performed, audio recorded, and collected field notes for allinterviews. Study participants provided verbal consent andreceived incentives worth $10 (US dollars). This study wasapproved by the University of Arizona Institutional ReviewBoard. The data that support the findings of this study areavailable from the corresponding author on reasonable request.To minimize the possibility of unintentionally sharing informa-tion that can be used to reidentify private information, only asubset of data will be available.
VignetteParticipants were blinded to study objectives until participationwas complete and were randomized 1:1 to white man or blackman. Race was indicated using text, photograph, and ethnic-sounding names.16 Photographs were selected from a normal-ization study conducted by the primary investigator’s (K.B.)team (S.S., L.L., K.H.T., L.Z.). The 2 study photographs differedonly by race, having similar hairstyle, clothing, and physicalbuild; they were similarly rated for age, attraction, intelligence,health, facial expression, and trustworthiness (Table S1). Thevignettes were identical with the exception of race. The vignettedescribed a patient with end-stage HF with a complex history,including multiple relative contraindications for advancedtherapies, such as reduced social support, reduced treatmentadherence, financial instability, and history of remote drug use,obesity, and mild levels of diabetes mellitus, kidney dysfunc-tion, and peripheral vascular disease (Table S2).
Survey InstrumentParticipants were asked to rate on a Likert scale (1–10,strongly disagree to strongly agree) how well individual factorsmake the patient suitable for advanced HF therapies(Figure 1). Final recommendations for heart transplant, bridgeto transplant VAD (future candidate for transplant), destina-tion VAD (not a candidate for transplant), and no advancedtherapies were individually elicited on a Likert scale. Write-in
Clinical Perspective
What Is New?
• Interviews revealed that the decision-making process wasbased on medical appropriateness, comorbidities, andstrength of the social and emotional support system.
• Compared with the white man, black race negativelyinfluenced the allocation of heart transplant modestly.
• The racial differences in the decision-making process forheart transplant were more pronounced during interviewsthan during surveys.
What Are the Clinical Implications?
• Race modestly influences the decision-making process foradvanced heart failure therapies.
• Further investigation should address tactics to reduce racialbias in decision making for advanced heart failure therapies.
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 2
Heart Failure Decision Making and Race Breathett et alORIG
INALRESEARCH
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
responses were invited to describe key reasons for finaldecisions and additional pertinent comments. Demographicinformation was also collected.
Interview GuideThink aloud, a form of cognitive interviewing, is an establishedmethod of qualitatively evaluating the decision-making pro-cess.17,18 The interview guide was framed on the cognitive
interview guide by Shafer and Lohse, which includes probes toassist with thinking aloud.19 Participants were promptedabout thoughts on patient candidacy for advanced therapiesas they read through each section of the vignette (Figure 1).Participants were asked for a final recommendation for typeof advanced therapy and about whether the participanttrusted the patient. Participants were also asked to detail howthe vignette compared with patient presentations during theirown selection meetings.
Figure 1. Interview guide and survey questions. Warm-up questions and helping participant with thinkaloud were from Shafer and Lohse instructions on cognitive interviewing.19
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 3
Heart Failure Decision Making and Race Breathett et alORIG
INALRESEARCH
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
Statistical Analysis
Surveys were considered complete if all 18 survey factorvariables plus need for additional testing (Figure 1) and 4outcome variables (heart transplant, bridge to transplant VAD,destination therapy VAD, and no advanced therapies) weregiven. The exception was for the paper surveys from interviewsubjects for which only the 4 outcome variables were requiredfor inclusion (excluded 5 participants). Simple imputation(median values) was used for missing values from the factorvariables (7 values from 5 participants). Multiple imputationwas used for missing values for participant-demographicquestions (57 missing values from 21 participants). Predictivemean matching was used with the R package “mice.”20 For allanalyses, ordinary least squares regression was used on atransformed response variable to account for nonnormality.The t-test was used to compare 4 outcome results by vignetterace. A 2-stage approach was used to analyze the survey data.This was performed in the interest of avoiding overparame-terization with the large number of variables, as well asbecause of the use of multiple imputation on the participantdemographics. First, least absolute shrinkage and selectionoperator (LASSO)21 was used to perform variable selection onthe 19 factors and the interaction of each of those factorswith vignette race (a total of 38 parameters) for eachresponse variable. The tuning parameter for LASSO (k) wasselected via 10-fold cross validation. For each responsevariable, the value of k giving mean-squared error within 1 SDof the minimum was used. The R package “glmnet” was usedfor the LASSO analysis.22 Last, once a baseline model ofimportant factors was established for each response, alldemographic variables and the interactions of all demographicvariables with vignette race were added. Model estimates andSEs were based on pooled values from 20 imputations.
Think-aloud interviews underwent thematic analysis usinggrounded theory while blinded to the vignette patient’s race.Then, results were unblinded and categories were comparedaccording to race. The stepwise process included thefollowing: (1) open coding: identifying themes (patterns) untilreaching saturation (absence of new themes that designatesappropriate sample size23), (2) central phenomenon: exploringthemes to arrive at a central category, (3) axial coding:connecting categories, and (4) selective coding: generating amodel representing the decision-making process.24 Rigor wasestablished through credibility (validation of interviewresponse with survey response), transferability (debrief withadvanced HF cardiologist [N.S.]), and confirmability (includingtrained interviewers [E.Y., A.L.] and 2 independent analysts[N.P., M.H.] who performed the entire qualitative analyseswith differences arbitrated by an independent qualitativeexpert [J.C.] and the primary investigator [K.B.]).25 An audittrail and codebook were maintained throughout the study.
ResultsParticipants were randomized to white (N=204 survey, n=22interview) and black (N=218 survey, n=22 interview) manpatient vignettes similarly across demographics (Table 1,Table S3). Half or more of participants who completed surveysand interviews were aged ≥40 years, men, non-Hispanic white,cardiologist or cardiothoracic surgeons, with <11 years pasttraining. Participants were dispersed similarly throughout theUnited Network for Organ Sharing regions (Figure S1), with theexception of interview participants missing from region 6.
Survey ResultsThe favorability for heart transplant (mean rating: white, 7.08[95% CI, 6.74–7.43]; black, 7.28 [95% CI, 6.97–7.58]), bridgeto transplant VAD (white, 7.61 [95% CI, 7.28–7.93]; black,7.74 [95% CI, 7.44–8.05]), destination VAD (white, 6.84 [95%CI, 6.47–7.21]; black, 7.17 [95% CI, 6.82–7.53]), and noadvanced HF therapies (white, 2.39 [95% CI, 2.15–2.63];black, 2.23 [95% CI, 1.99–2.47]) was similar for white andblack man vignettes. Results for interview participants werealso quantitatively similar for the white and black manvignettes (Table S4). Variables selected via LASSO for hearttransplant that had a positive effect on allocation included thefollowing (in order of greatest to least magnitude): adherence,social history, other medical and surgical history, laboratorytests, history of present illness, and cardiac diagnostic testing(Table 2). Need for additional testing or consultation had anegative effect on transplant allocation. Variables selected viaLASSO for bridge to transplant VAD only had a positive effecton allocation. In order of greatest to least magnitude, theseincluded the following: social history, other medical andsurgical history, laboratory tests, adherence, and cardiacdiagnostic testing. For destination therapy VAD, no factorswere selected as important by LASSO. For allocation to noadvanced therapy, the LASSO selected factors included thefollowing: need for additional testing or consultation as apositive effect (supporting no advanced therapies) and thefollowing factors as negative effects on allocation (supportingadvanced therapy): history of present illness, adherence,other medical or surgical history, laboratory tests, cardiacdiagnostic testing, social history, and height/weight/bodymass index. The interactions with vignette race and eachsurvey factor were not selected as important in the LASSOanalysis.
After LASSO and multivariable regression, significantparticipant demographic factors negatively influenced alloca-tion to heart transplant, including ≥11 years of past training(�0.49 [95% CI, �0.92 to �0.06]) and interaction of age≥40 years by black vignette (�0.58 [95% CI, �1.15 to�0.0002]). Overall among healthcare professionals aged
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 4
Heart Failure Decision Making and Race Breathett et alORIG
INALRESEARCH
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
≥40 years, the white vignette trended more favorably fortransplant over the black vignette but was not significantlydifferent (white vignette, 0.39 [95% CI, �0.04 to 0.83]; black
vignette, �0.18 [95% CI, �0.56 to 0.19]). Bridge to transplantVAD was negatively influenced by ≥11 years of past training(�0.51 [95% CI, �0.96 to �0.07]). No participant demo-graphic factors significantly influenced the decision fordestination therapy VAD. Advising against candidacy foradvanced therapies was positively influenced by ≥11 years ofpast training (0.49 [95% CI, 0.10-0.88]).
Think-Aloud Interview ResultsUsing grounded theory, a 3-pronged central phenomenonguiding decision making for advanced HF therapies emerged:Is the heart sick enough? Is the body well enough? Is thereenough social/emotional support to make it through theprocess? The decision-making process was sequentiallydirected by 5 themes, with 2 themes occurring simultane-ously: (1) forming an overall impression, (2) identifyingurgency, (3a) evaluating the appropriateness of prior care,(3b) anticipating challenges, and (4) evaluating trust andmaking the ultimate recommendation (Table 3, Figure 2).Themes were further characterized by 12 subthemes describ-ing racial similarities and differences affecting the decision-making process. Exemplar quotes illustrate the descriptionsof the 5 major themes in the following section.
Forming an overall impression
Participants developed an overall impression of the patientfrom the photograph. Some participants suggested thephotograph contributed to negative bias about race andsocioeconomic position for both the white and blackvignettes. Others felt that the photograph had little impacton their decisions and found race irrelevant with theexception of making decisions about genotyping. Severalparticipants found the photographs to be more detrimentaltoward the white than the black vignette.
“. . .a few times a staff member has given the presentation and
decided to actually put the photo [in], and I decided that that was
not good, because you have an opinion that is now not based on
facts on paper, but whether you thought the person was lovely or
not.” (white participant/white vignette)
Identifying urgency
The second step in decision making included figuring out thepatient’s severity of illness and need for advanced HFtherapies. All participants surmised that both the white andblack vignette patients had end-stage HF on the basis of thehistory of present illness. Participants believed that develop-ing consensus for urgency was an important step in thedecision-making process. However, participants displayedmore consistent concern for higher illness acuity within theblack than the white vignette.
Table 1. Participant Demographics
Demographics
White Man Black Man
P Value*Vignette(N=204)
Vignette(N=218)
Age, y 0.76
<40 82 (40.2) 82 (37.6)
40 119 (58.3) 129 (59.2)
Unknown 3 (1.5) 7 (3.2)
Sex 0.34
Men 107 (52.5) 125 (57.3)
Women 92 (45.1) 87 (39.9)
Unknown 5 (2.5) 6 (2.8)
Ethnicity 0.83
Minority 59 (28.9) 67 (30.7)
Non-Hispanic white 137 (67.2) 145 (66.5)
Unknown 8 (3.9) 6 (2.8)
Position 0.83
Noncardiologist 59 (28.9) 65 (29.8)
Cardiologist orcardiothoracicsurgeon
142 (69.6) 146 (67.0)
Unknown 3 (1.5) 7 (3.2)
Past training, y 0.95
<11 113 (55.4) 118 (54.1)
11 89 (43.6) 90 (41.3)
Unknown 2 (1.0) 10 (4.6)
UNOS region
1 14 (6.9) 13 (6.0) 0.30†
2 21 (10.3) 21 (9.6)
3 13 (6.4) 13 (6.0)
4 16 (7.8) 12 (5.5)
5 33 (16.2) 34 (15.6)
6 3 (1.5) 7 (3.2)
7 29 (14.2) 25 (11.5)
8 16 (7.8) 12 (5.5)
9 14 (6.9) 13 (6.0)
10 25 (12.3) 22 (10.1)
11 17 (8.3) 39 (17.9)
Unknown 3 (1.5) 7 (3.2)
Data are given as number (percentage). Interviewed participants meeting exclusioncriteria for survey analysis because of missing values for therapy allocation are notincluded in this table (white man vignette n=2, black man vignette n=3). UNOS indicatesUnited Network for Organ Sharing.*The v2 test for P value excludes unknowns because small values of unknown provideinaccurate approximation.†The UNOS region P value approximation may be inaccurate.
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 5
Heart Failure Decision Making and Race Breathett et alORIG
INALRESEARCH
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
Table 2. Factors Influencing Decision Making for Advanced Therapies From Stage 1 LASSO and Stage 2 Multivariable RegressionModels
Stage 1 LASSO Regression Model
Coefficient
Heart Transplant BTT VAD DT VAD Not Candidate
HPI 0.032 ��� ��� �0.074Age ��� ��� ��� ���Race or ethnicity ��� ��� ��� ���Sex ��� ��� ��� ���Height/weight/BMI ��� ��� ��� �0.0003Insurance ��� ��� ��� ���Blood type and PRA class ��� ��� ��� ���Cardiac history ��� ��� ��� ���NYHA functional class and vital signs ��� ��� ��� ���Medications ��� ��� ��� ���Other medical history and surgical history 0.058 0.080 ��� �0.047Social history 0.085 0.137 ��� �0.016Adherence 0.123 0.020 ��� �0.055Cardiac diagnostic testing 0.027 0.009 ��� �0.031Pulmonary studies ��� ��� ��� ���Laboratory tests 0.054 0.040 ��� �0.043Infectious disease ��� ��� ��� ���Cancer screening ��� ��� ��� ���Additional testing/consultation needed �0.003 ��� ��� 0.003Black vignette ��� ��� ��� ���Stage 2 DemographicFactors’ MultivariableRegression Model
ParameterEstimate (95% CI) P Value
ParameterEstimate (95% CI) P Value
ParameterEstimate (95% CI) P Value
ParameterEstimate (95% CI) P Value
White non-Hispanic �0.13 (�0.47 to 0.21) 0.44 0.18 (�0.18 to 0.55) 0.32 0.49 (�0.02 to 1.00) 0.06 �0.22 (�0.53 to 0.10) 0.18Women 0.09 (�0.29 to 0.46) 0.65 0.19 (�0.20 to 0.58) 0.34 0.27 (�0.27 to 0.82) 0.33 0.06 (�0.29 to 0.40) 0.75Aged ≥40 y 0.39 (�0.04 to 0.83) 0.08 0.25 (�0.20 to 0.71) 0.27 0.03 (�0.62 to 0.68) 0.93 �0.17 (�0.57 to 0.23) 0.40≥11 y Past training �0.49 (�0.92 to �0.06)* 0.02 �0.51 (�0.96 to �0.07)* 0.03 0.12 (�0.52 to 0.77) 0.71 0.49 (0.10 to 0.88)* 0.01Cardiologist 0.04 (�0.37 to 0.46) 0.84 0.12 (�0.31 to 0.55) 0.58 0.01 (�0.59 to 0.61) 0.98 �0.16 (�0.54 to 0.22) 0.41Black vignette 0.20 (�0.53 to 0.93) 0.58 0.15 (�0.61 to 0.92) 0.69 0.76 (�0.33 to 1.86) 0.17 �0.05 (�0.74 to 0.64) 0.89White non-Hispanicand black vignetteinteraction
�0.04 (�0.51 to 0.42) 0.86 �0.35 (�0.84 to 0.15) 0.17 �0.46 (�1.17 to 0.26) 0.21 0.42 (�0.03 to 0.86) 0.06
Women and blackvignette interaction
0.00 (�0.50 to 0.50) 0.99 0.10 (�0.42 to 0.62) 0.71 �0.19 (�0.94 to 0.55) 0.61 �0.24 (�0.71 to 0.23) 0.32
Aged ≥40 y andblack vignetteinteraction
�0.58 (�1.15 to �0.0002)* 0.0499 �0.25 (�0.85 to 0.36) 0.42 0.28 (�0.59 to 1.15) 0.53 �0.06 (�0.62 to 0.49) 0.83
≥11 y Past trainingand black vignetteinteraction
0.52 (�0.05 to 1.09) 0.08 0.50 (�0.10 to 1.10) 0.10 �0.35 (�1.21 to 0.51) 0.42 �0.19 (�0.74 to 0.36) 0.49
Cardiologist andblack vignetteinteraction
�0.14 (�0.69 to 0.41) 0.61 �0.11 (�0.69 to 0.46) 0.70 �0.27 (�1.10 to 0.55) 0.52 �0.09 (�0.61 to 0.44) 0.74
Positive values denote support for decision, and negative values denote disapproval for decision; ellipses indicate 0 coefficient and no influence on decision making. Each factor was alsoincluded as an interaction with black vignette. These interactions were zero coefficients in stage 1 of LASSO regression and do not include participant demographics. Stage 2 multivariablemodel included factors with nonzero coefficient from LASSO model to determine influence of participant demographics. BMI indicates body mass index; BTT, bridge to transplant; DT,destination therapy; HPI, history of present illness; LASSO, least absolute shrinkage and selection operator; NYHA, New York Heart Association; PRA, panel reactive antibody; VAD,ventricular assist device.*Multivariable regression P value factors with significance <0.05.
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 6
Heart Failure Decision Making and Race Breathett et alORIG
INALRESEARCH
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
Table 3. Themes and Subthemes With Illustrative Quotations
Central Phenomenon: Is the Heart Sick Enough? Is the Body Well Enough? Is There Enough Social/Emotional Support to Make It Through the Process?
Themes
Illustrative Quotations (Participant Race)Subthemes Vignette Type
Forming an overall impression
Black “. . .he looks clean and well groomed. You can tell that his beard has been combed. He is ofdarker skin so non-Caucasian; I think it’s actually difficult to tell what ethnicity this patient isbased on the picture anyhow. . . That’s about all I notice about the gentleman.” (white)
“Because they look well-kept, it gives me a sense that they may be compliant with medicaltherapy. Since they take the time to trim their beard, they probably are detail-oriented. That’smy initial impression” (minority)
White “It looks like a middle-aged man. He is not smiling, and he’s bald. . . He’s got a goatee, well-trimmed. Hard to tell what his teeth look like. . .for me, a big part of the physical exam is alwaysthe oral exam. . ..I would not lean one way or another just based on how he looks.” (minority)
“He doesn’t look that sick when I look at him. He’s not heavy and I don’t see any JVD standing up sohe’s got good fill-out of his face and temple area and stuff. Looks young. Lots of options, justlooking at him” (white)
Avoiding explicitly discussing race,except for physiologic issues
Black “African-American, I would probably send genetic testing for amyloid” (minority)“One of the few things that his race would affect would be I guess interpreting PSA, so, it’s goodthat his PSA is negative. . .” (white)
White “I’m looking at a white Caucasian male which I know affects survival and risk. . .” (white)“His picture matches his age and the other description, right non-Hispanic. So that’s about it.Doesn’t look malnourished, doesn’t look cachectic. . . It’s just a head shot. This photo’sneutral.” (white)
Believing photographs may contribute tonegative racial bias
Black “. . .we don’t usually say what the ethnicity of the person is. . . Yeah, we don’t have a picture ofthe person usually either. . .” (white)
White “I am certain that there is a bias, especially when it comes to ethnic minorities and women ingeneral. . .” (minority)
“. . .I just think it’s interesting, because I do think people are swayed by that [knowing race]. . .And I just think that’s wrong. I should not know the sex, anything else.” (white)
Developing some negative reactions tothe white but not black man
Black “Looks like a typical gentleman that you would see, stated age; doesn’t look terribly ill.” (minority)“. . .well-kept,goodhygiene,neatlytrimmedbeard.That’sall Iwouldsaybasedonthephoto.” (minority)
White “. . .if I look at the face, that scares the heck out of me. I’ll be honest with you, it sways me. . . Ithink he looks scary. Honestly, first judgement when you see him, he looks like a prisoner. . .itlooks like a mugshot photo of somebody.” (white)
“My first reaction was cover the face, because I don’t want that information. . . Because I thinkthat you have a bias right there. . . I think most people wouldn’t like him. He doesn’t look veryfriendly to start with.” (white)
Identifying urgency
Black “So, all of these things are signs to me that things are not going well with his illness and that hecould be or is in the end stages of HF.” (white)
White “There’s no question about the need of more advanced care.” (white)“And then the high number of hospitalizations. . . No matter what we decide at this committeemeeting, we know that this guy is probably not going to be alive for much longer.” (white)
Developing consensus for urgency ofadvanced therapies
Black “. . .things are really starting to mount that he’s got advanced disease.” (white)
White “He needs advanced therapies. . .I will say, I will just list him for transplant. . .he’s veryadvanced.” (white)
“The other things jump out in terms of urgency. So the fact that he’s had four hospitalizations inthe past 6 months, his mortality rate is very high. . . so it’s not as if we have a long time tokind of consider his candidacy.” (minority)
Believing the black man was sicker thanthe white man
Black “Multiple hospitalizations in the past 6 months is a terrible prognostic indicator, moreover, itwould insinuate that he has a terrible quality of life. . . His dizziness and lightheadedness withwalking minimal distance suggests that he is a terribly ill man, and he also [has] ventriculararrhythmia, so he has a very high-risk profile for dying.” (white)
Continued
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 7
Heart Failure Decision Making and Race Breathett et alORIG
INALRESEARCH
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
Table 3. Continued
Central Phenomenon: Is the Heart Sick Enough? Is the Body Well Enough? Is There Enough Social/Emotional Support to Make It Through the Process?
Themes
Illustrative Quotations (Participant Race)Subthemes Vignette Type
“I don’t have too many patients in my clinic like this that are not being actively considered foradvanced therapies, or. . .being referred to hospice or at least a palliative care visit. It’s time tobe thinking about those things, yeah.” (white)
“I think he’s too sick. . .he has a sick heart, but otherwise, he’s favorable for transplant listing orVAD I would say, really. . .he’s going to die unless something is done soon. . .” (white)
“You know given how sick he is and that we need to make a decision quickly on this gentleman.My thought given his blood type is that he’s somebody we’re gonna need to move reallyquickly to an LVAD as a bridge hopefully to transplantation.” (white)
White “My sense is he’s not terribly frail. . .” (white)“He doesn’t seem like he’s critically ill, based on his PMH didn’t tell me that. . .” (white)
Evaluating the appropriateness of prior care
Black “Certainly, if he’s not on good therapy, maybe he would be better if he was on bettertherapy. . .” (white)
White “Adherence, lost to follow-up for a few years when he didn’t have health insurance, we see thatall the time. I think it’s a horrible statement in our country, but we haven’t fixed that.” (white)
“I just want to make sure that someone has tried to get him [on] a good medical therapy andthat he is taking what he’s been able to. . .” (white)
Questioning late presentation andappropriate guideline-directed medicaltherapy
Black “I’d also question whether or not he is adequately treated from a medical standpoint, andwhether or not he has been discharged prematurely, and whether or not he is on optimalmedical therapy in the hopes of precluding another hospitalization.” (white)
White “Does he have, has he been, is he on adequate therapy? Is there anything that could be done,that can be done to improve his trajectory?” (white)
“This is our typical kind of patient. . .We usually get them too late and it doesn’t matter. I mean,we still do exactly what we need to do but it’s just. . . the systems are not sophisticatedenough to capture these patients. For example, this guy was in the hospital 4 times in the last6 months. Could we have gotten him, in an ideal world, after his second one and with hiskidneys and liver in better shape?” (white)
Finding more concerns for appropriatetreatment of black than white man
Black “Has he not been triaged and treated properly?” (minority)“. . .he seems to have a pretty malignant course in that he was just diagnosed a year agoand. . .doesn’t seem to have responded to medical management unfortunately. Lots ofhospitalizations which is a poor prognostic indicator.” (white)
“It sounds like he’s a gentleman despite having significant heart failure. . . probably has notbeen exposed to heart failure specialists previously.” (white)
White “I think it was a pretty thorough workup and presentation. . .” (white)“Typical stuff. It looks like he’s had a reasonable workup.” (white)
Anticipating challenges
Black “So, his size and blood type would infer that if he is indeed a candidate for heart transplant, histime on the wait list would be considerable.” (white)
“The social history is concerning for whether or not he has adequate social and financialresources to tolerate a VAD or a transplant. . .” (white)
White “I think the fact that we can’t put him on medications is a major issue. When we consideradvanced therapies, certainly, taking a pill is a lot easier than going through major cardiacsurgery. We try to max out medications first. . .” (white)
“. . . social support and kind of having backup plans as what to do that like any other thing withheart failure patients is a constant struggle but it can be overcome if you make them think thatway. Prepare, have a backup plan. You know, work on it. Talk to people. Have a wider net ofsupport, not just your wife, not just your brother.” (minority)
Developing similar concerns for socialdeterminant of health, includingcaregiver burnout for both races
Black “So if you have a patient, let’s say, that may not have the [social] support for transplant thenmaybe VAD may be an option because there’s less visits. . .. that’s something I would considerin this patient too, if we find that the support is not adequate for transplant.” (minority)
“I think healthcare literacy is a major issue. I don’t think this is a malignantly noncompliant
Continued
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 8
Heart Failure Decision Making and Race Breathett et alORIG
INALRESEARCH
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
Table 3. Continued
Central Phenomenon: Is the Heart Sick Enough? Is the Body Well Enough? Is There Enough Social/Emotional Support to Make It Through the Process?
Themes
Illustrative Quotations (Participant Race)Subthemes Vignette Type
patient. He lets life overwhelm him as a subjective judgment. So I would say this is somethingwe can help modify. It’s not a deal breaker.” (minority)
White “. . .I don’t want your wife to burn out. She’s maintaining 2 children, household, and taking care ofyou so she needs help. And if he says I can’t find anybody I say I cannot move forward withtransplant or VAD if you can’t come up with a second person. Because it’s not fair to his wife.”(minority)
“But he’s kind of drawing the short straw when it comes to the social determinants of health. Hesounds resource poor, lost his job, doesn’t even have disability yet.” (white)
Developing greater concerns foradherence/trust for black thanwhite man
Black “Adherence. . . Lost to follow-up for a couple of years when he didn’t have healthcareinsurance. . . Something’s not fitting here. Is he lying to us? Was he not working for the PostalService? The social worker will need to sort that out. . .” (minority)
“He’s seeking disability. I don’t like the word seeking. . . It’s been 6 months. What kind offinancial support is he going to have?” (minority)
“It makes me wonder what his outpatient situation has been like and what also his complianceissue has been with his outpatient care. . .” (minority)
White “The fact that he’s had, is married or would have had a significant other that’s working lendscredibility to their ability to take care of the needs that come along with really advancedtherapy. . .” (white)
“I’d say the adherence part doesn’t push me away from advanced therapies. It may push metowards LVAD first and say if you miss an appointment or maybe it’s not quite as a big a deal. . ..itgives me some time to get to know him first before he’s in a stage where he needs to takeimmunosuppression and have it tightly regulated and get labs and be compliant.” (white)
Evaluating trust and making the ultimate recommendation
Black “I don’t have reason not to. . . It sounds like he was a guy who worked all his life. . . I don’t trustanyone really but I trust that he wants to live, and if that’s the case then I’ll give him a shot. . .”(minority)
“I don’t have a longitudinal history with him. . . I don’t like to use the word trust. I would use theword has this person shown sufficient compliance and understanding of liability. . .” (minority)
“He seems pretty clean and straightforward based on what I’m presented. I mean it’s one thing,again, you know meeting somebody and reading about somebody are 2 totally differentthings.” (minority)
White “It goes back to the whole photo. So, if I didn’t have the photo and I’m looking at the facts, wewould [trust him]. The answer is, this patient has tried.” (white)
“Neither one’s on the table until I get to know him better, you know. . .” (white)
Believing the heart is a finite gift andshould not be wasted
Black “If thereisn’t [adequatesocialsupport]andthere’sstillagrayareathenIwouldbepushedalittlemoretowards VAD than transplant. . .there’s a limited number of organs available whereas VADS, there’sno limitation on that. So, we describe it as a precious resource, the heart. . .” (white)
“Wereallydoneedtoshowconsistencyoffollow-upjustgivenourobligationswith transplant, thatwewere good stewards of the organs. . .” (white)
“I know I keep saying social compliance and I think it’s more for respect for an organ honestly. . .”(minority)
White “And if you take a 20-year-old heart and put it in somebody who’s only going to get 5 years outof it, you haven’t really served the donor appropriately. And that might be something thatpushes me more towards going towards a VAD. . .. I don’t want to take a heart and give it tosomebody who’s going to mess it up in 5 years.” (white)
“. . .these therapies are not only very expensive, but are limited, and so, in the world of ethics,and trying to maximize utility for everybody, you really don’t want to use resources onsomebody who is not going to benefit fully from them because of things like nonadherence andvariable compliance.” (minority)
“At the same time we need to pick candidates who are capable of taking care of the vitalresource that they were provided, be it a VAD and/or a transplant, obviously transplant more.And it’s my job ethically to make sure he’s able to do that. If some poor mother’s going to
Continued
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 9
Heart Failure Decision Making and Race Breathett et alORIG
INALRESEARCH
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
“. . .I think he is going to need to have a workup and relatively
expeditious action. I don’t think this is the kind of patient you can
sit on for a long time and hope that they’re going to do well. So I
would say that I would not be dragging my feet. . .” (white
participant/black vignette)
Evaluating the appropriateness of care
In the third step, participants questioned whether the patientshad received appropriate care before referral. Participants
wondered whether guideline-directed medical therapy hadbeen provided correctly and why patients were presenting solate for evaluation for advanced HF therapies.
“. . . If he was my own patient, I would probably try myself to see if
indeed he is intolerant. I would have to do a very thorough review
of his chart to see if he has been given the old college try to
confirm absolute intolerance.” (white participant/black vignette)
Participants found more concerns about appropriate prior treat-
ment for the black than the white vignette.
Table 3. Continued
Central Phenomenon: Is the Heart Sick Enough? Is the Body Well Enough? Is There Enough Social/Emotional Support to Make It Through the Process?
Themes
Illustrative Quotations (Participant Race)Subthemes Vignette Type
donate a 17-year-old’s heart, I need to know that this person’s capable of taking care of it.”(white)
Desiring a road test for white man withinotropes to help him get a transplant
White “. . .we will ask him to come to clinic every week, show compliance to his medication regimen,show that he can come to clinic appointments, bring his social support people with him. . .”(minority)
“I still think he’s a candidate [for transplant]. Maybe he would benefit from some IV inotropes inthe meantime. . .” (white)
“If he has an adequate caregiver plan and can get his diabetes under good control would placeon inotropes/IABP and list for transplant.” (white)
“I’m thinking he should be put on inotropes and we should get to know him. . .” (white)“I would like to have some bridge to show me that he’s compliant. . . Start with an inotrope. . .See if they can handle it.” (minority)
Offering the black man a VAD Black “It’s not clear what the equipoise is here for this but I think most people would VAD him too, atthis point.” (white)
“I would lean a little bit more towards an LVAD just given the issues with the noncompliance. . .”(minority)
“. . .[if he won’t take a hep c heart] he’s probably a candidate for LVAD cuz he’s pretty sick. . .”(white)
“I would say that I don’t know right or wrong, I think sometimes in cases where we don’t havethe ability to wait, we typically give the patients the benefit of the doubt for an LVAD, probablynot for transplant.” (white)
“But I think in this case you could go either way. And then really it’s up to him or his financialplus family support that if he can go straight to transplant because obviously the patient andthe family has to be very, very compliant with therapy. So they have to come to the clinic visitand then also take medications and stuff like that. If there is any concern from thatperspective, then LVAD might be a better option.” (minority)
“. . .this guy absolutely merits inotropic support. And the question is, is the inotropic supportgoing to be a bridge to compliance? Is the inotropic support going to be a bridge tooptimization for LVAD? Or is the inotropic support going to be a bridge to transplant listing? Wedon’t know that yet.” (minority)
“. . .I would lean a little bit more towards an LVAD just given the issues with thenoncompliance. . .cuz we can work with [the A1C] and obesity. . . Those are from what I knownow, fixable problems. . .but I think the noncompliance, that is definitely a hindrance to idealcare.” (minority)
“I think we would definitely consider this man for an LVAD at our center pretty soon. I thinktransplant is a little bit harder to be very enthusiastic about. . .” (minority)
White “I would push him to be a transplant candidate instead of a VAD candidate. . . He wouldpotentially be a great candidate for transplant” (white)
“At this stage, if all those statements that I made are presumed, that there’s nothing elsemissing, that he doesn’t have sarcoidosis, he doesn’t have active inflammation in his heart, Iwould probably recommend that he goes through a transplant.” (minority)
HF indicates heart failure; IV, intravenous; IABP, intra-aortic balloon pump; JVD, jugular venous distention; LVAD, left VAD; PSA, prostate specific antigen; PMH, past medical history; VADs,ventricular assist devices.
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 10
Heart Failure Decision Making and Race Breathett et alORIG
INALRESEARCH
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
Throughout the decision-making process, participantsbelieved that the selection team dynamics impacted deci-sions. Hierarchy in selection team seating, role, and opinionwere described during actual advanced HF therapy selectionmeetings.
“But at an actual selection meeting, it’s such an interesting
sociologic study because you always have the people with the
most power and seniority. . .at the front of the room. They’re the
ones who make the decisions. Then the underlings sit in the back
of the room. . .then somebody who knows the patient might
chime in. . .with their usual strong opinion that may or may not be
completely rational in a given clinical situation.” (white partic-
ipant/white vignette)
Anticipating ChallengesIn this simultaneous third step, participants also begananticipating barriers to advanced HF therapies. Participantswere concerned about social determinants of health and therisk of caregiver burnout for both the white and blackvignettes. Participants developed greater concerns for adher-ence with the black vignette.
“I know we sometimes use the VAD. . .like a test to see if they
would do well with a heart transplant and because of his non-
compliant appearing issues in the past, that would be why I would
lean towards a VAD to foresee if he can be compliant before
giving him an organ. . .it’s more of a test to see if. . .a patient can
respect the organ that’s given to them.” (minority participant/
black vignette)
“Adherence, lost to follow-up for a couple of years when he didn’t
have health insurance. That is a red flag for both therapies. You
know, these are high resource-intense therapies that you really do
need to maintain. . . We really do need to show consistency of
follow-up just given our obligation with transplant, that we were
good stewards of the organs.” (white participant/black vignette)
“This guy was taking care of himself, just his disease outstripped
his ability to manage it. . .” (white participant/white vignette)
Evaluating Trust and Making the UltimateRecommendationThe last step in the decision-making process includedevaluating whether the healthcare selection team trustedthe patient with advanced HF therapy, which contributed tothe final recommendation. Participants believed that the heartis a finite gift and should not be wasted. This contributed tohesitancy with offering a transplant in both the black andwhite vignettes. However, the final recommendation differed
Figure 2. Decision-making process for allocating advanced heart failure therapies. Themes from Grounded Theory of Think-AloudInterviews. GDMT indicates guideline-directed medical therapy; LVAD, left ventricular assist device.
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 11
Heart Failure Decision Making and Race Breathett et alORIG
INALRESEARCH
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
by race. The white vignette was more often offered a “roadtest” with inotropes to see if he could appropriately managehis care and be listed for heart transplant later. The blackvignette was offered a VAD because he was thought to be tooill to wait for transplant.
“I would say that I don’t know right or wrong, I think sometimes in
cases where we don’t have the ability to wait, we typically give
the patients the benefit of the doubt for an LVAD [left VAD],
probably not for transplant.” (white participant/black vignette)
DiscussionIn this first study to robustly address how patient raceinfluences decision making for advanced HF therapies, wefound that black race compared with white race modestlyinfluenced allocation to heart transplant, particularly duringinterviews. Survey ratings for advanced HF therapies weresimilar for black and white patient vignettes, but analysisrevealed that black vignette and healthcare professional aged≥40 years were subtly associated with opposing hearttransplant. Surveys revealed that subjectively assessedfactors, including social history and adherence patterns, weresignificant factors in the decision-making process. Bothfactors were described during interviews as negativelyinfluencing allocation of heart transplant, particularly amongblacks.
Interviews revealed a stepwise decision-making processthat was influenced by race. Race influenced each of thesesteps: avoiding discussing race explicitly, believing pho-tographs may contribute to racial bias, believing the blackman was more ill and possibly undertreated compared withthe white man, developing greater concern for trust andadherence with the black man, and ultimately offering thewhite man a transplant and the black man a VAD.
Clinical uncertainty can contribute to racial bias in clinicaldecision making.10,26 The indications and absolute contraindi-cations for advanced HF therapies are definitive in theguidelines; however, relative contraindications, which areincreasing among advanced therapy candidates,27 are open toprovider interpretation.9,28 As observed in this study, health-care professionals had benevolent goals for the patients in thevignettes but were subtly influenced by patient race.
An extensive literature base has demonstrated thatunconscious bias contributes to racial inequalities in medicalcare.7,29–31 Often, racial and ethnic minority patients areperceived negatively by healthcare professionals.7,29–31 Biasincreases further against minority patients compared withwhite patients when the patient has a lower socioeconomicposition.29 As was illustrated in our study, the black man wasviewed as less adherent to therapy than the white man,despite having the same clinical and social history. Over 2decades ago, similar findings were found among physicians
considering cardiovascular catheterizations for black andwhite patients.14 Although bias was subtly observed in ourstudy, its presence can profoundly affect medical care.32
The stability of advanced therapy programs is based onhaving good patient outcomes and reasonable patientvolumes. Programs may deny candidacy to patients perceivedas high risk for poor outcomes because they may threaten thelongevity of the advanced therapy program.33 Blacks havehigher mortality after heart transplant than whites34 and areknown to not receive equitable access to care.4 This couldcontribute to decisions to offer a heart transplant to the whitepatient over the black patient. However, programs havedemonstrated similar long-term outcomes after heart trans-plant for blacks and whites through comprehensive multidis-ciplinary care.35
Reducing racial inequities in decision making for hearttransplant may require several approaches. The first stepcould include reducing clinical uncertainty by making subjec-tive assessments objective. Subjective assessments ofadherence and social support could be replaced with knownobjective measurements, such as the Morisky MedicationAdherence Scale36 and the Stanford Integrated PsychosocialAssessment for Transplantation.37 The second step includesawareness that conscious and unconscious bias can con-tribute to interpretation of assessments.7 Known methods topromote egalitarian treatment of patients and healthcareteam members include bias training.38 Healthcare profession-als can learn how to recognize biases and work to establishconnections that supersede bias.7,38 Last, patient presenta-tions can be changed so that potentially biasing information isnot presented. As described by participants in this study,programs can eliminate photographs and ethnicity identifiersfrom presentations. This last step would not remove bias fromthose that have a working relationship with the patient butmay prevent other team members involved in decision-makingprocesses from potentially making a biased decision.
LimitationsThe response rate for surveys was 29%. Although this maylimit generalization, participants from all US United Networkfor Organ Sharing regions were represented, with theexception of region 6, in the interviews. Because themembership directory did not distinguish members whoroutinely allocate advanced HF therapies, the number ofeligible respondents is lower than the total number sent thesurvey. Thus, our actual response rate is unknown, but >29%.Over 10% of participants were interviewed and saturation wasachieved, signifying an appropriate sample size. Interviewerswere not blinded to study objectives and could haveemphasized topics unintentionally during interviews. However,steps were taken to avoid biased interviewing, including a 3-
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 12
Heart Failure Decision Making and Race Breathett et alORIG
INALRESEARCH
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
hour training on think-aloud protocol and multiple observedpractice interviews.
ConclusionsWhen provided a patient case vignette, HF professionalsfollowed a stepwise process in clinical decision making foradvanced HF therapies that addressed medical appropriate-ness, comorbidities, and strength of social and emotionalsupport systems. Race modestly influenced the decision-making process. Healthcare providers subtly recommendedheart transplant in white patients over black patients, despiteidentical medical and social history. Racial bias was demon-strated particularly during interviews, which more closelyresemble allocation meetings, than in numerical surveys.Because allocation for advanced HF therapies occurs viaconversations in group settings rather than through surveys,the influence of race on decision making may be significantand should be addressed.
AcknowledgmentsWe acknowledge Sarver Heart Center administrative support fromGilbert Maldonado, Mari Vayre, and Taylor Valenzuela.
Sources of FundingDr Breathett received support from the National Heart, Lung,and Blood Institute K01HL142848, University of ArizonaHealth Sciences, Strategic Priorities Faculty Initiative Grant,and University of Arizona, Sarver Heart Center, Women ofColor Heart Health Education Committee. L. Luy receivedsupport from National Institutes of Health R25HL108837.
DisclosuresNone.
References1. Benjamin EJ, Muntner P, Alonso A, Bittencourt MS, Callaway CW, Carson AP,
Chamberlain AM, Chang AR, Cheng S, Das SR, Delling FN, Djousse L, ElkindMSV, Ferguson JF, Fornage M, Jordan LC, Khan SS, Kissela BM, Knutson KL,Kwan TW, Lackland DT, Lewis TT, Lichtman JH, Longenecker CT, Loop MS,Lutsey PL, Martin SS, Matsushita K, Moran AE, Mussolino ME, O’Flaherty M,Pandey A, Perak AM, Rosamond WD, Roth GA, Sampson UKA, Satou GM,Schroeder EB, Shah SH, Spartano NL, Stokes A, Tirschwell DL, Tsao CW,Turakhia MP, VanWagner LB, Wilkins JT, Wong SS, Virani SS; American HeartAssociation Council on Epidemiology and Prevention Statistics Committee andStroke Statistics Subcommittee. Heart disease and stroke statistics—2019update: a report from the American Heart Association. Circulation. 2019;139:e56–e66.
2. Chan PS, Oetgen WJ, Buchanan D, Mitchell K, Fiocchi FF, Tang F, Jones PG,Breeding T, Thrutchley D, Rumsfeld JS, Spertus JA. Cardiac performancemeasure compliance in outpatients: the American College of Cardiology andNational Cardiovascular Data Registry’s PINNACLE (Practice Innovation AndClinical Excellence) program. J Am Coll Cardiol. 2010;56:8–14.
3. Farmer SA, Kirkpatrick JN, Heidenreich PA, Curtis JP, Wang Y, Groeneveld PW.Ethnic and racial disparities in cardiac resynchronization therapy. HeartRhythm. 2009;6:325–331.
4. Breathett K, Liu WG, Allen LA, Daugherty SL, Blair IV, Jones J, Grunwald GK,Moss M, Kiser TH, Burnham E, Vandivier RW, Clark BJ, Lewis EF, Mazimba S,Battaglia C, Ho PM, Peterson PN. African Americans are less likely to receivecare by a cardiologist during an intensive care unit admission for heart failure.JACC Heart Fail. 2018;6:413–420.
5. Breathett K, Allen LA, Helmkamp L, Colborn K, Daugherty SL, Khazanie P,Lindrooth R, Peterson PN. The Affordable Care Act Medicaid expansioncorrelated with increased heart transplant listings in African-Americans but notHispanics or Caucasians. JACC Heart Fail. 2017;5:136–147.
6. Breathett K, Allen LA, Helmkamp L, Colborn K, Daugherty SL, Blair IV, Jones J,Khazanie P, Mazimba S, McEwen M, Stone J, Calhoun E, Sweitzer NK, PetersonPN. Temporal trends in contemporary use of ventricular assist devices by raceand ethnicity. Circ Heart Fail. 2018;11:e005008.
7. Breathett K, Jones J, Lum HD, Koonkongsatian D, Jones CD, Sanghvi U,Hoffecker L, McEwen M, Daugherty SL, Blair IV, Calhoun E, de Groot E,Sweitzer NK, Peterson PN. Factors related to physician clinical decision-making for African-American and Hispanic patients: a qualitative meta-synthesis. J Racial Ethn Health Disparities. 2018;5:1215–1229.
8. Mehra MR, Kobashigawa J, Starling R, Russell S, Uber PA, Parameshwar J,Mohacsi P, Augustine S, Aaronson K, Barr M. Listing criteria for hearttransplantation: International Society for Heart and Lung Transplantationguidelines for the care of cardiac transplant candidates—2006. J Heart LungTransplant. 2006;25:1024–1042.
9. Feldman D, Pamboukian SV, Teuteberg JJ, Birks E, Lietz K, Moore SA, MorganJA, Arabia F, Bauman ME, Buchholz HW, Deng M, Dickstein ML, El-Banayosy A,Elliot T, Goldstein DJ, Grady KL, Jones K, Hryniewicz K, John R, Kaan A, KusneS, Loebe M, Massicotte MP, Moazami N, Mohacsi P, Mooney M, Nelson T,Pagani F, Perry W, Potapov EV, Eduardo Rame J, Russell SD, Sorensen EN, SunB, Strueber M, Mangi AA, Petty MG, Rogers J. The 2013 International Societyfor Heart and Lung Transplantation guidelines for mechanical circulatorysupport: executive summary. J Heart Lung Transplant. 2013;32:157–187.
10. Unequal treatment: confronting racial and ethnic disparities in health care—Institute of Medicine [Internet]. http://www.iom.edu/Reports/2002/Unequal-Treatment-Confronting-Racial-and-Ethnic-Disparities-in-Health-Care.aspx.Accessed August 21, 2012.
11. Cunningham CT, Quan H, Hemmelgarn B, Noseworthy T, Beck CA, Dixon E,Samuel S, Ghali WA, Sykes LL, Jett�e N. Exploring physician specialist responserates to web-based surveys. BMC Med Res Methodol. 2015;15:32.
12. Morton SMB, Bandara DK, Robinson EM, Carr PEA. In the 21st century, what isan acceptable response rate? Aust N Z J Public Health. 2012;36:106–108.
13. Sarma AA, Nkonde-Price C, Gulati M, Duvernoy CS, Lewis SJ, Wood MJ.Cardiovascular medicine and society: the pregnant cardiologist. J Am CollCardiol. 2017;69:92–101.
14. Schulman KA, Berlin JA, Harless W, Kerner JF, Sistrunk S, Gersh BJ, Dub�e R,Taleghani CK, Burke JE, Williams S, Eisenberg JM, Escarce JJ. The effect of raceand sex on physicians’ recommendations for cardiac catheterization. N Engl JMed. 1999;340:618–626.
15. Granovetter M. Network sampling: some first steps. Am J Sociol.1976;81:1287–1303.
16. Bertrand M, Mullainathan S. Are Emily and Greg more employable than Lakishaand Jamal? A field experiment on labor market discrimination [Internet].National Bureau of Economic Research; 2003. http://www.nber.org/papers/w9873. Accessed December 12, 2016.
17. Li AC, Kannry JL, Kushniruk A, Chrimes D, McGinn TG, Edonyabo D, Mann DM.Integrating usability testing and think-aloud protocol analysis with “near-live”clinical simulations in evaluating clinical decision support. Int J Med Inform.2012;81:761–772.
18. Crist JD, Koerner KM, Hepworth JT, Pasvogel A, Marshall CA, Cruz TP, EffkenJA. Differences in transitional care provided to Mexican American and non-Hispanic white older adults. J Transcult Nurs. 2017;28:159–167.
19. Shafer K, Lohse B. How to conduct a cognitive interview: a nutritionaleducation example [Internet]. http://www.au.af.mil/au/awc/awcgate/usda/cog_interview.pdf. Accessed April 6, 2019.
20. van Buuren S, Groothuis-Oudshoorn K. mice: Multivariate imputation bychained equations in R. J Stat Softw. 2011;45:1–67.
21. Tibshirani R. Regression shrinkage and selection via the Lasso. Journal of theRoyal Statistical Society Series B (Methodological). 1996;58:267–288.
22. Friedman JH, Hastie T, Tibshirani R. Regularization paths for generalized linearmodels via coordinate descent. J Stat Softw. 2010;33:1–22.
23. Curry LA, Nembhard IM, Bradley EH. Qualitative and mixed methods provideunique contributions to outcomes research. Circulation. 2009;119:1442–1452.
24. Strauss A, Corbin JM. Basics of Qualitative Research: Grounded TheoryProcedures and Techniques. Thousand Oaks, CA: Sage Publications, Inc;1990.
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 13
Heart Failure Decision Making and Race Breathett et alORIG
INALRESEARCH
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
25. Creswell JW. Qualitative Inquiry and Research Design: Choosing Among FiveApproaches. Thousand Oaks, California: SAGE Publications; 2012.
26. Balsa AI, McGuire TG. Prejudice, clinical uncertainty and stereotyping assources of health disparities. J Health Econ. 2003;22:89–116.
27. Kilic A, Conte JV, Shah AS, Yuh DD. Orthotopic heart transplantation inpatients with metabolic risk factors. Ann Thorac Surg. 2012;93:718–724.
28. Mehra MR, Canter CE, Hannan MM, Semigran MJ, Uber PA, Baran DA,Danziger-Isakov L, Kirklin JK, Kirk R, Kushwaha SS, Lund LH, Potena L, Ross HJ,Taylor DO, Verschuuren EAM, Zuckermann A; International Society for HeartLung Transplantation (ISHLT) Infectious Diseases, Pediatric and Heart Failureand Transplantation Councils. The 2016 International Society for Heart LungTransplantation listing criteria for heart transplantation: a 10-year update. JHeart Lung Transplant. 2016;35:1–23.
29. van Ryn M, Burke J. The effect of patient race and socio-economic status onphysicians’ perceptions of patients. Soc Sci Med. 2000;50:813–828.
30. Green AR, Carney DR, Pallin DJ, Ngo LH, Raymond KL, Iezzoni LI, Banaji MR.Implicit bias among physicians and its prediction of thrombolysis decisions forblack and white patients. J Gen Intern Med. 2007;22:1231–1238.
31. Hall WJ, Chapman MV, Lee KM, Merino YM, Thomas TW, Payne BK, Eng E, DaySH, Coyne-Beasley T. Implicit racial/ethnic bias among health care profes-sionals and its influence on health care outcomes: a systematic review. Am JPublic Health. 2015;105:e60–e76.
32. van Ryn M, Burgess DJ, Dovidio JF, Phelan SM, Saha S, Malat J, Griffin JM, FuSS, Perry S. The impact of racism on clinician cognition, behavior, and clinicaldecision making. Du Bois Rev. 2011;8:199–218.
33. Felker GM, Milano CA, Yager JEE, Hernandez AF, Blue L, Higginbotham MB,Lodge AJ, Russell SD. Outcomes with an alternate list strategy for hearttransplantation. J Heart Lung Transplant. 2005;24:1781–1786.
34. Morris AA, Kransdorf EP, Coleman BL, Colvin M. Racial and ethnic disparities inoutcomes after heart transplantation: a systematic review of contributingfactors and future directions to close the outcomes gap. J Heart LungTransplant. 2016;35:953–961.
35. PamboukianSV, Costanzo MR,Meyer P, Bartlett L, Mcleod M, Heroux A. Influenceof race in heart failure and cardiac transplantation: mortality differences areeliminated by specialized, comprehensive care. J Card Fail. 2003;9:80–86.
36. Morisky DE, Ang A, Krousel-Wood M, Ward HJ. Predictive validity of amedication adherence measure in an outpatient setting. J Clin Hypertens(Greenwich). 2008;10:348–354.
37. Maldonado JR, Dubois HC, David EE, Sher Y, Lolak S, Dyal J, Witten D. TheStanford Integrated Psychosocial Assessment for Transplantation (SIPAT): anew tool for the psychosocial evaluation of pre-transplant candidates.Psychosomatics. 2012;53:123–132.
38. Stone JA, Moskowitz GB. Non-conscious bias in medical decision making: whatcan be done to reduce it? Med Educ. 2011;45:768–776.
ORIG
INALRESEARCH
DOI: 10.1161/JAHA.119.013592 Journal of the American Heart Association 14
Heart Failure Decision Making and Race Breathett et al
Dow
nloaded from http://ahajournals.org by on F
ebruary 8, 2020
