Abstract
Background:
Heart failure affects >6 million people in the United States alone and is most prevalent in Black patients who suffer the highest mortality risk. Yet prior studies have suggested that Black patients are less likely to receive advanced heart failure therapy. We hypothesized that Black patients would have decreased rates of durable left ventricular assist device (LVAD) implantation within our expansive heart failure program.
Methods:
A retrospective single-center cohort study was conducted at a single high-volume academic medical center. Patients between 18 and 85 years admitted with a diagnosis of cardiogenic shock or congestive heart failure between 1, 2013 and 12, 2017 with a left ventricular ejection fraction < 30% and inotropic dependence or need for mechanical circulatory support were included. Patients with contraindications to durable LVAD were excluded. An adjusted logistic regression model for durable LVAD implantation within 90 days of the index admission was used to determine the effect of race on durable LVAD implantation.
Results:
Among the 702 study patients (60.9% White, 34.1% Black), durable LVAD implantation was performed within 90 days of the index admission in 183 (26%) of the cohort. After multivariate analysis, Black patients were not found to have a statistically significant difference in durable LVAD implantation rates compared to White patients in our study (OR: 0.68 [95% confidence interval: 0.45–1.04; p: .074]).
Conclusions:
Black patients in our study did not have a statistically significant difference in the rate of durable LVAD implantation compared with White patients after adjustments were made for age, sex, socioeconomic, and clinical covariates. Larger prospective studies are needed to validate these findings.
Keywords: gender, heart failure, left ventricular assist device, race
1 |. INTRODUCTION
Heart failure affects approximately 6.6 million people in the United States. This number is expected to rise in coming years due to the aging of the population, and it is estimated that over 8 million people in the United States will carry this diagnosis by 2030.1 Heart failure is more prevalent in the Black population and carries the highest mortality risk of all racial groups.2,3 Implementation of innovative intensive therapies in cardiovascular disease has been delayed in Black patients.4,5 Black patients are less likely than White patients to receive invasive cardiac procedures despite equivalent disease severity and access to care.4–6
Durable left ventricular assist device (LVAD) implantation improves survival in patients with advanced cardiac failure compared to optimal medical therapy.7–9 However, evidence is emerging that Black patients with advanced heart failure are less likely to receive LVAD therapy compared to White patients.10 The role of durable LVAD implantation for the treatment of advanced heart failure will likely only continue to expand in the future, therefore it is imperative to determine if there is equitable access to this therapy in large academic center settings, where the majority of these procedures are performed.
We hypothesized that Black patients would have decreased rates of LVAD implantation compared with White patients within our expansive heart failure program.
2 |. METHODS
We conducted a retrospective observational cohort study at Duke University Hospital. Patients admitted to Duke University Hospital between January 2013 and December 2017 with a primary diagnosis of cardiogenic shock or congestive heart failure with a left ventricular ejection fraction (LVEF) < 30% and inotropic dependence or need for mechanical circulatory support were included in the study. Mechanical circulatory support included intra-aortic balloon pump (IABP), extracorporeal membrane oxygenation (ECMO), or extracorporeal ventricular assist device. Exclusion criteria included: age <18 or >85 years of age, advanced kidney disease (defined as dialysis-dependent or Cr > 3.5), hypertrophic cardiomyopathy, cardiac amyloidosis, pregnancy, primary pulmonary hypertension, cor pulmonale, Eisenmenger’s, chronic thromboembolic pulmonary hypertension, pulmonary vascular resistance >8 WU, prior orthotopic heart transplant, cirrhosis, uncorrected thyroid disease, pericardial disease, active myocarditis, severe chronic obstructive pulmonary disease (COPD) with FEV1 < 1.5 L/m, BMI > 40 and current substance abuse. International Classification of Diseases Codes recorded in the electronic medical record on admission were used to identify patients admitted with a diagnosis of congestive heart failure or cardiogenic shock. Patient demographics including age, sex, race, insurance type, median zip code associated income, and medical comorbidities were obtained from the electronic medical record. Fourteen patients did not have available median zip code income, necessitating the use of median county income. Race was self-reported by study participants. Race categories were defined by investigators based on the revisions to the US Office of Management and Budget Standards for the classification of federal data on race and ethnicity.
Ethical approval was obtained from the Duke University Health System Institutional Review Board.
2.1 |. Statistical analysis
Baseline demographic variables and patient characteristics were reported as number and percentage for categorical variables and median and interquartile range for continuous variables. Categorical characteristics were compared between groups using chi-squared tests or Fisher exact tests, as appropriate, and continuous characteristics were compared using Kruskal–Wallis tests. Durable LVAD implantation within 90 days of the index admission was the a priori outcome in our statistical analysis plan. The index admission was defined as the first admission for heart failure that met study inclusion criteria between January 2013 to December 2017. Subsequent admissions were excluded. Based on institutional practices, 90 days was identified to be a sufficient duration of time to determine whether a patient who met study inclusion criteria would receive a device if determined to be eligible, considering the evaluation process and patient’s high risk for mortality. The rate of LVAD implantation was compared across racia groups using a logistic regression model with adjustment for a priori specified patient characteristics and comorbidities which included insurance status, employment status, age, sex, median zip-code income, BMI, tobacco use, COPD, peripheral vascular disease, diabetes, hypertension, chronic kidney disease, and coronary artery disease. Results are reported as odds ratios with 95% confidence intervals (CIs), as well as p-values. A p-value less than .05 was considered statistically significant. All statistical analysis was carried out using R version 3.5.0.
3 |. RESULTS
A total of 8181 patients were admitted to Duke University Hospital with a primary diagnosis of cardiogenic shock or congestive heart failure between January 2013 and December 2017. After applying the inclusion criteria of LVEF < 30% and inotropic or mechanical support dependence and previously described exclusion criteria, a study population of 702 patients was identified of which 34.1% were Black, 60.9% were White, and 5.1% were categorized as Other which included American Indian or Alaska Native 8 (1.1%), Asian 8 (1.1%), multiracial 3 (0.43%), Native Hawaiian or Pacific Islander 1 (0.14%), other 10 (1.4%), not reported/declined: 4 (0.57%). The decision to classify these patients as Other was due to their small numbers limiting meaningful statistical analysis. Gender representation in the study population was disproportionate, with 72% men. The majority of the patients had health insurance with fewer than 1% of patients being uninsured; 32.8% of patients had private insurance, 10.3% had Medicaid and 56.4% had Medicare (Table 1).
TABLE 1.
Patient characteristics
| Black or African American (n = 240; 34.2%) | Caucasian/White (n = 427; 60.8%) | Othera (n = 35; 5.0%) | p-value | |
|---|---|---|---|---|
| Age | 59 [49, 67] | 65 [56, 71] | 60 [51, 66] | <.001 |
| Gender | - | - | - | .024 |
| Female | 81 (33.8%) | 104 (24.4%) | 12 (34.3%) | - |
| Male | 159 (66.3%) | 323 (75.6%) | 23 (65.7%) | - |
| Ethnicity | - | - | - | <.001 |
| Hispanic | 1 (0.4%) | 2 (0.5%) | 9 (25.7%) | - |
| Non-Hispanic | 236 (98.3%) | 419 (98.1%) | 24 (68.6%) | - |
| Not reported/declined/unavailable | 3 (1.3%) | 6 (1.4%) | 2 (5.7%) | - |
| BMI | 28.3 [23.7, 32.8] | 27.6 [23.9, 32.0] | 25.5 [22.0, 28.6] | .348 |
| Employment status | - | - | - | <.001 |
| Employed | 24 (10.0%) | 78 (18.3%) | 5 (14.3%) | - |
| Not employed | 40 (16.7%) | 41 (9.6%) | 9 (25.7%) | - |
| Retired | 89 (37.1%) | 223 (52.2%) | 9 (25.7%) | - |
| Disabled | 87 (36.3%) | 85 (19.9%) | 12 (34.3%) | - |
| Insurance status | - | - | - | <.001 |
| Private | 67 (27.9%) | 156 (36.5%) | 15 (42.9%) | - |
| Medicare | 130 (54.2%) | 248 (58.1%) | 12 (34.3%) | - |
| Medicaid | 41 (17.1%) | 23 (5.4%) | 7 (20.0%) | - |
| Uninsured | 2 (0.8%) | 0 (0.0%) | 1 (2.9%) | - |
| Median zip code incomea | 46,900 [36,600, 57,900] | 50,800 [40,400, 61,800] | 46,900 [33,700, 70,500] | .001 |
| ECMO during admission | 8 (3.3%) | 36 (8.4%) | 6 (17.1%) | .003 |
| Inotropes during admission | 240 (100.0%) | 416 (97.4%) | 34 (97.1%) | .042 |
| IABP during admission | 34 (14.2%) | 70 (16.4%) | 5 (14.3%) | .732 |
Abbreviation: ECMO, extracorporeal membrane oxygenation.
Other includes American Indian/Alaska Native, Asian, two or more races, Native Hawaiian/Pacific Islander, not reported/declined.
There were several noteworthy differences in patient characteristics between racial groups. Black patients were younger (median age 59 vs. 65) with a higher proportion of women (33.8% vs. 24.4%). Median zip code-associated income was slightly higher in White patients at $50,800 compared to $46,900 in Black patients. Black patients were less likely to have private insurance (27.9% vs. 36.5%) and Medicare (54.2% vs. 58.1%) but were more likely to have Medicaid compared to White patients (17.1% vs. 5.4%). Comorbidities were similar among racial groups with the only significant difference present in the prevalence of coronary artery disease, which was higher in the White group at 58.1% versus 36.7% (Table 2). The incidence of chronic kidney disease was similar among the racial groups, but Black patients were noted to have a higher creatinine (creatinine = 1.4 vs. 1.3 in White patients) on admission compared to other racial groups (p < .001). Black patients in the study had higher mean pulmonary artery pressures and pulmonary vascular resistance on right heart catheterization data. Black patients in the study were most likely to require inotropic support during admission (100% vs. 97.4% in White patients). There was a significant difference in the utilization of ECMO as a mode of mechanical circulatory support, with the highest use in patients classified as Other at 17.1% and the lowest use in Black patients at 3.3%. The use of ECMO in White patients during the index admission was 8.4%. However, the use of IABP as a mode of mechanical circulatory support was similar among racial groups.
TABLE 2.
Patient comorbidities
| Black or African American (n = 240; 34.2%) | Caucasian/White (n = 427; 60.8%) | Other (n = 35; 5.0%) | p-value | |
|---|---|---|---|---|
| Tobacco use | 21 (8.8%) | 34 (8.0%) | 2 (5.7%) | .889 |
| COPD | 27 (11.3%) | 52 (12.2%) | 5 (14.3%) | .811 |
| Peripheral vascular disease | 20 (8.3%) | 50 (11.7%) | 2 (5.7%) | .299 |
| Diabetes | 6 (2.5%) | 10 (2.3%) | 1 (2.9%) | .917 |
| Hypertension | 125 (52.1%) | 192 (45.0%) | 13 (37.1%) | .103 |
| Chronic kidney disease | 72 (30.0%) | 104 (24.4%) | 6 (17.1%) | .148 |
| Coronary artery disease | 88 (36.7%) | 248 (58.1%) | 18 (51.4%) | <.001 |
| Baseline creatinine | 1.4 [1.1, 1.8] | 1.3 [1.0, 1.7] | 1.1 [0.9, 1.5] | <.001 |
| PVR | 3.5 [2.3, 5.1] n = 41 | 2.7 [1.7, 3.6] n = 58 | 5.5 [4.9, 6.1] n = 2 | .007 |
| Fick cardiac output index | 1.95 [1.63, 2.47] n = 107 | 2.13 [1.72, 2.59] n = 194 | 1.98 [1.77, 2.76] n = 17 | .515 |
| DUHSPA mean | 39 [33, 45] n = 107 | 36 [28, 42] n = 195 | 40 [35, 48] n = 17 | .014 |
| DUHSPCW mean | 25 [20, 30] n = 106 | 23 [18, 29] n = 195 | 27 [21, 30] n = 17 | .289 |
| DUHSRA mean | 13 [8, 18] n = 106 | 13 [9, 17] n = 195 | 15 [10, 19] n = 17 | .776 |
Abbreviation: COPD, chronic obstructive pulmonary disease.
Of the 702 patients included in the study, 26% (183) underwent durable LVAD implantation within 90 days of the index admission. 7.2% (51) received a heart transplant, 1.28% (9) received a durable LVAD and a heart transplant, and 18.6% (131) expired within 90 days of admission. Unadjusted analysis of durable LVAD implantation determined that 22.9% of Black patients in the study received a durable LVAD compared to 28.8% of White patients (p = .066) (Table 3) (Figure 1). Ninety-day mortality rates among the three groups studied were similar 20% (Black patients) versus 17.6% (White patients) versus 22.9% (American Indian/Alaska Native, Asian, Multiracial, Native Hawaiian/Pacific Islander, other, and not reported/declined (p = .737). INTERMACS scores were available for 177 of the 183 patients. The average and median INTERMACS scores were similar between the Black and White patients in the study (Table 4). Of the 183 patients who received a durable LVAD, 53 were bridge to transplant, 127 were destination therapy (DT), and 3 could not be determined. The proportion of patients receiving durable LVAD therapy as DT was similar between the Black and White groups. 65.5% of Black patients underwent LVAD implantation for DT as compared to 70.7% of White patients (p: .722) (Table 5). Age was more frequently listed as a reason for DT LVAD for White patients in the study compared to Black patients. Psychiatric concerns were more commonly listed as a reason for LVAD DT in Black patients compared to White patients in the study (Table 6).
TABLE 3.
Outcomes at 90 days from index admission
| Black or African American (n = 240; 34.2%) | Caucasian/White (n = 427; 60.8%) | Other (n = 35; 5.0%) | p-value | |
|---|---|---|---|---|
| Heart transplant within 90 days of admission | 17 (7.1%) | 32 (7.5%) | 2 (5.7%) | .968 |
| LVAD within 90 days of admission | 55 (22.9%) | 123 (28.8%) | 5 (14.3%) | .066 |
| Death within 90 days of admission | 48 (20.0%) | 75 (17.6%) | 8 (22.9%) | .737 |
Abbreviation: LVAD, left ventricular assist device.
FIGURE 1.
Ninety-day outcomes by race
TABLE 4.
INTERMACS
| Black or African American (n = 54) | Caucasian/White (n = 118) | Other (n = 5) | |
|---|---|---|---|
| INTERMACS score | |||
| 1 | 5 (9.3%) | 19 (16.1%) | 1 (20.0%) |
| 2 | 20 (37.0%) | 37 (31.4%) | 4 (80.0%) |
| 3 | 21 (38.9%) | 31 (26.3%) | 0 |
| 4 | 8 (14.8%) | 31 (26.3%) | 0 |
| Average ± SD | 2.6 ± 0.9 | 2.6 ± 1.0 | 1.8 ± 0.4 |
| Median (IQR) | 3 (2, 3) | 3 (2, 4) | 2 (2, 2) |
TABLE 5.
Destination versus bridge to transplant LVAD implantation
| Black or African American (n = 55) | Caucasian/White (n = 123) | Other (n = 5) | |
|---|---|---|---|
| Bridge to transplant | 18 (32.7%) | 34 (27.6%) | 1 (20%) |
| Destination therapy | 36 (65.5%) | 87 (70.7%) | 4 (80%) |
| Unknown | 1 (1.8%) | 2 (1.6%) | 0 |
Abbreviation: LVAD, left ventricular assist device.
TABLE 6.
Reasons for DTT LVAD
| Black (n = 36) | White (n = 87) | Other (n = 4) | p-value | |
|---|---|---|---|---|
| Age | 6 (16.7%) | 35 (40.2%) | 1 (25.0%) | .031 |
| Allosensitization | 1 (2.8%) | 1 (1.1%) | 0 (0%) | .532 |
| Comorbidities | 12 (33.3%) | 21 (24.1%) | 0 (0%) | .349 |
| Infection | 1 (2.8%) | 2 (2.3%) | 0 (0%) | >.999 |
| Malignancy | 2 (5.6%) | 7 (8.0%) | 1 (25.0%) | .394 |
| Nonadherence | 1 (2.8%) | 1 (1.1%) | 0 (0%) | .532 |
| Obesity | 5 (13.9%) | 4 (4.6%) | 0 (0%) | .218 |
| Patient preference | 0 (0%) | 2 (2.3%) | 0 (0%) | >.999 |
| Previous sternotomy | 1 (2.8%) | 6 (6.9%) | 0 (0%) | .737 |
| Psychiatric | 3 (8.3%) | 0 (0%) | 0 (0%) | .033 |
| Pulmonary hypertension | 8 (22.2%) | 15 (17.2%) | 1 (25.0%) | .467 |
| Social determinants of health | 6 (16.7%) | 3 (3.4%) | 0 (0%) | .066 |
| Substance or tobacco use | 6 (16.7%) | 10 (11.5%) | 1 (25.0%) | .301 |
| No documented reason | 4 (11.1%) | 9 (10.3%) | 0 (0%) | .840 |
Abbreviation: LVAD, left ventricular assist device.
After adjustment for comorbidities, age, sex, employment, median income, and type of insurance, Black patients did not have a statistically significant difference in rates of durable LVAD implantation (OR: 0.68 [95% CI, 0.45–1.04; p: .074]). In this multivariable analysis, women were less likely than men to receive LVAD implantation (OR: 0.63 [95% CI, 0.41–0.97; p: .035]). Patients with Medicaid were less likely to receive durable LVAD implantation compared to patients insured with Medicare (OR: 0.29 [95% CI, 0.12–0.66; p: .003]). Private insurance was associated with an increased likelihood of receiving a durable LVAD compared to Medicaid (OR: 2.64 [95% CI, 1.16–6.02; p: .02]). Increasing age decreased the likelihood of durable LVAD implantation (OR: 0.97 [95% CI, 0.95–0.99; p: .015]) (Table 7). Patients with a history of chronic kidney disease were more likely to receive a durable LVAD (OR: 1.99 [95% CI, 1.34–2.96; p: .001]) (Table 7).
TABLE 7.
Adjusted logistic regression model for 90-day LVAD utilization
| Odds Ratio (95% CI) | p-value | |
|---|---|---|
| Race | ||
| Black versus White | 0.68 (0.45, 1.04) | .074 |
| Other versus White | 0.50 (0.18, 1.40) | .190 |
| Insurance status | ||
| Medicaid versus Medicare | 0.29 (0.12, 0.66) | .003 |
| Private versus Medicare | 0.76 (0.47, 1.20) | .235 |
| Private versus Medicaid | 2.64 (1.16, 6.02) | .021 |
| None versus Medicare | N/A | N/A |
| Employment status | ||
| Disabled versus retired | 1.22 (0.72, 2.07) | .471 |
| Employed versus retired | 0.97 (0.50, 1.88) | .918 |
| Not employed versus retired | 0.76 (0.36, 1.62) | .480 |
| Age (per year) | 0.97 (0.95, 0.99) | .015 |
| Gender (female vs. male) | 0.63 (0.41, 0.97) | .035 |
| Median zip-code income (per $1000) | 1.00 (0.99, 1.01) | .915 |
| BMI (per point) | 1.04 (1.00, 1.07) | .046 |
| Tobacco use | 0.47 (0.21, 1.05) | .064 |
| COPD | 1.29 (0.75, 2.21) | .360 |
| Peripheral vascular disease | 0.89 (0.48, 1.65) | .704 |
| Diabetes | 1.60 (0.53, 4.85) | .405 |
| Hypertension | 1.26 (0.86, 1.84) | .244 |
| Chronic kidney disease | 1.99 (1.34, 2.96) | .001 |
| Coronary artery disease | 1.19 (0.79, 1.77) | .406 |
Abbreviations: CI, confidence interval; COPD, chronic obstructive pulmonary disease; LVAD, left ventricular assist device.
An additional analysis was then performed comparing rates of durable LVAD implantation between White patients to the Black and Other patient groups combined with adjustment for comorbidities and socioeconomical variables. This multivariable analysis revealed that Black, American Indian/Alaska Native, Asian, Multiracial, Native Hawaiian/Pacific Islander, other, and not reported/declined patients when collapsed into one group, were less likely to receive durable LVAD implantation compared to White patients (OR: 0.66 [95% CI: 0.44–0.99; p: .043]) (Table 8).
TABLE 8.
Adjusted logistic regression model for 90-day LVAD utilization
| Odds Ratio (95% CI) | p-value | |
|---|---|---|
| Race | ||
| Non-White versus White | 0.66 (0.44, 0.99) | .043 |
| Insurance status | ||
| Private versus Medicare | 0.75 (0.47, 1.19) | .224 |
| Medicaid versus Medicare | 0.28 (0.12, 0.66) | .003 |
| Medicaid versus private | 0.38 (0.17, 0.86) | .021 |
| None versus Medicare | N/A | N/A |
| Employment status | ||
| Disabled versus retired | 1.21 (0.71, 2.06) | .484 |
| Employed versus retired | 0.96 (0.49, 1.86) | .899 |
| Not employed versus retired | 0.75 (0.35, 1.60) | .461 |
| Age (per 10 years) | 0.77 (0.62, 0.95) | .014 |
| Gender (female vs. male) | 0.63 (0.41, 0.97) | .036 |
| Median zip-code income (per $1000) | 1.00 (0.99, 1.01) | .902 |
| BMI (per point) | 1.04 (1.00, 1.07) | .042 |
| Tobacco use | 0.47 (0.21, 1.05) | .066 |
| COPD | 1.28 (0.75, 2.20) | .366 |
| Peripheral vascular disease | 0.89 (0.48, 1.65) | .707 |
| Diabetes | 1.61 (0.53, 4.89) | .398 |
| Hypertension | 1.27 (0.87, 1.86) | .222 |
| Chronic kidney disease | 2.00 (1.34, 2.97) | .001 |
| Coronary artery disease | 1.17 (0.79, 1.75) | .439 |
Abbreviations: CI, confidence interval; COPD, chronic obstructive pulmonary disease; LVAD, left ventricular assist device.
4 |. DISCUSSION
In this single-center, retrospective, observational cohort study of 702 patients admitted with advanced heart failure, we found that Black patients did not have a statistically significant difference in rates of durable LVAD implantation compared with White patients. In a multivariable analysis, female gender, Medicaid insurance, increased age, and being a race other than White were independently associated with decreased likelihood of durable LVAD implantation in this cohort.
Our finding that Black patients in the study did not have a statistically significant difference in durable LVAD implantation compared to their White counterparts is in contrast to the body of research that has shown that Black patients are less likely to undergo procedures related to treating cardiac disease including coronary artery bypass grafting, percutaneous coronary intervention (PCI), internal cardiac defibrillators (ICDs) and heart transplantation. A retrospective cohort study of 40,526 patients which used data from the Nationwide Inpatient Sample of patients admitted with cardiogenic shock and congestive heart failure between 2002 and 2003 found that Black patients were less likely to receive LVAD (extracorporeal and intracorporeal) after adjustment for age, comorbidities, and insurance despite being more likely to be admitted to an academic center.10 A recent larger retrospective cohort study using State Inpatient Databases of 925,770 patients admitted with heart failure or cardiogenic shock over a 3-year period found that Black race and female gender were associated with a decreased likelihood of durable LVAD implantation after adjusting for age, comorbidities, type of insurance and median income.1
One explanation of our results is that it is unique to the geographical location or the academic center, and may reflect specific processes of the institution which help to eliminate structural discrimination, improve effective outreach, and trust building in the Black community. The institution in which the study was performed has a high surrounding Black population (34%) and a large proportion of Black patients are referred to the heart failure program. This may have influenced the program’s practices over time resulting in the ability to effectively manage potential barriers to advanced heart failure therapy more commonly present in this patient population. Our study is also different than prior studies in that it is single center and included more detailed data elements including employment status and zip-code income data. It is also possible that the lack of a statistically significant difference between the groups in this modest sample size is a Type II error. Studies on larger samples are needed to draw definitive conclusions. It should also be considered that the absence of a statistically significant difference in utilization of durable LVAD therapy in Black patients with advanced heart failure compared to White patients may still reflect underutilization in Black patients who have a higher prevalence of and mortality risk from heart failure.
The question of whether race and racial bias, in the absence of socioeconomic factors, influences decision-making by clinicians and patients in advanced heart failure therapy is an important one to answer. However, it is just as important to identify institutional and structural barriers that result in inequitable health care. Our study revealed that patients with Medicaid were less likely to receive a durable LVAD compared to patients with Medicare or private insurance. This finding is consistent with other research that has been performed. Two prior retrospective cohort studies have also found Medicaid insurance to be associated with a decreased likelihood of LVAD implantation.1,11 One of these studies was focused on one large academic hospital in a state where Medicaid provided limited reimbursement for LVAD surgery, the second study was multi-institutional across 15 states where Medicaid reimbursement for LVAD surgery was variable. Our finding of decreased durable LVAD implantation associated with Medicaid insurance in a state where Medicaid provides reimbursement for this procedure within an academic institution that serves a diverse population is significant. Unlike many other states, North Carolina Medicaid covers all expenses associated with durable LVAD implantation. In our study, Black, American Indian/Alaska Native, Asian, Multiracial, Native Hawaiian/Pacific Islander, other, and not reported/declined patients were more likely to be insured by Medicaid. The negative association of Medicaid insurance on the likelihood of durable LVAD implantation, despite its coverage of LVAD expenses, is suggestive of Medicaid being a marker of another unmeasured factor which influenced decision making. Determination of the contributing factors to this finding will be critical to addressing this disparity. Larger, prospective studies that obtain more granular information are needed.
In a secondary multivariable analysis, we found that White patients in the study were statistically more likely to receive a durable LVAD when compared to patients of all other races in the study. This difference was despite controlling for employment, comorbidities, insurance type, age, and gender. The reasons for this finding are unclear and are likely multifactorial and complex. Social determinants of health not fully captured in the study such as educational level and social support, physician counseling and referral practices, physician–patient relationships, structural, and provider bias as well as patient preference are potential contributing factors. Further study with more extensive social determinants of health data to determine whether this difference is patient-driven, physician-driven, system driven, or a combination of these is certainly warranted.
Sex was also found to have a statistically significant effect on the likelihood of durable LVAD implantation in our study population. Women were less likely to receive durable LVAD implantation after adjusting for type of insurance, employment status, income, race/ethnicity, age, and comorbidities. This finding is consistent with prior studies which have shown women to be less likely to receive advanced heart failure therapies.10,12,13 One potential cause for sex differences in durable LVAD implantation rates was due to size incompatibility with the larger, older pulsatile devices which required a BSA of >1.5 m2. Additionally, the Heartmate II LVAD was restricted to patients with a BSA > 1.2 m2 during the initial investigational period. However, the Heartmate II LVAD was FDA approved for DT in 2010 and this BSA requirement was no longer used in our institution after that time. Therefore this should not have contributed to the gender difference in implantation of durable LVAD in our study. Women have been shown to have an increased risk of stroke after durable LVAD implantation as compared to men in older-generation devices, including the Heartmate II device.13 Newer generation devices such as the Heartware and Heartmate 3 have not been shown to be associated with an increased stroke risk in women, but few women were included in these analyses. It is possible that clinicians are more reluctant to implant durable LVADs in women due to concern for an increased risk of stroke in this patient population. Gender bias is another potential cause for underutilization of advanced heart failure therapies in women. ICD implantation has been found to occur at lower rates in women than men with one large study revealing that women were less likely than men to receive ICD counseling.13 A recent study in 2020 that evaluated the influence of patient gender and race on allocation of advanced heart failure therapy, found that children were perceived as liabilities for women, and spouses were less likely to be perceived as adequate social support for women as compared to men.14 This perception was found to be most pronounced in Black women. Black women, who suffer the highest prevalence of heart failure among all racial/ethnic groups and sexes, and for whom heart failure with reduced ejection fraction is the most common cause for heart failure admission, may be at a particular disadvantage due to the intersectionality of race and gender, putting them at considerable risk for decreased utilization of this life-saving therapy. The implementation of implicit bias training with bias reduction strategies has been shown to reduce implicit bias and associated behavior among healthcare professionals.15,16 Consideration of implementation of such practices for advanced heart failure teams is warranted, however, their impact on advanced heart failure team decision-making has not been studied. This is an important area that merits future study.
5 |. LIMITATIONS
Our study had several limitations. It is a retrospective, observational, single-center study. This limited our sample size and power which may have limited our ability to detect a statistically significant difference in implantation rates between Black and White patients. Additionally, our findings may not reflect the national experience as it was limited to one institution. Another major limitation of our study is that it did not account for patients’ level of social support or severity of illness, both of which can influence clinician and patient decision-making. It is possible that the patient groups that were found to have a lower likelihood of LVAD implantation were more likely to have less social support or a higher severity of illness, making them less likely to be considered for this surgical intervention. The identification of decreased utilization of ECMO in Black patients in the study was notable, however, the reasons for this finding are unclear. The decision regarding offer or decline of ECMO utilization is complex and not clearly outlined in the electronic health record. Potential contributing factors could be a higher incidence of relative contraindications, ineligibility for advanced therapy, patient/family refusal, or provider referral practices. These potential factors could be included in a future prospective study for further exploration.
6 |. CONCLUSION
In this single-center study, Black patients in our study had similar rates of durable LVAD implantation compared with White patients. Larger prospective studies are needed to substantiate our findings related to healthcare disparities and the importance of ensuring equitable access for advanced heart failure therapies in partnership with the Centers for Medicare and Medicaid.
ACKNOWLEDGMENTS
Participating Investigators: Served as scientific advisor—Annemarie Thompson MD. Writing assistance—Monika Safford MD, Lisa Q. Rong MD, MSCE, FASE, FACC. Data Acquisition—Joseph Kelly, Alina Nicoara MD, Jonathan Piccini MD. Provided and cared for study patients- Jacob N. Schroder MD, Carmelo A. Milano MD.
Footnotes
CONFLICT OF INTEREST
The authors declare no conflict of interest.
REFERENCES
- 1.Wang X, Luke AA, Vader JM, Maddox TM, Joynt Maddox KE. Disparities and impact of medicaid expansion on left ventricular assist device implantation and outcomes. Circ Cardiovasc Qual Outcomes. 2020;13(6):e006284. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sharma A, Colvin-Adams M, Yancy CW. Heart failure in African Americans: disparities can be overcome. Cleve Clin J Med. 2014; 81(5):301–311. [DOI] [PubMed] [Google Scholar]
- 3.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] [PMC free article] [PubMed] [Google Scholar]
- 4.Peterson E, Yancy CW. Eliminating racial and ethnic disparities in cardiac care. N Engl J Med. 2009;360(12):1172–1174. [DOI] [PubMed] [Google Scholar]
- 5.Pinney SP. Disparities in heart failure care: now is the time to focus on health care delivery. J Am Coll Cardiol. 2014;64(8):808–810. [DOI] [PubMed] [Google Scholar]
- 6.Kressin NR, Petersen LA. Racial differences in the use of invasive cardiovascular procedures: review of the literature and prescription for future research. Ann Intern Med. 2001;135(5):352–366. [DOI] [PubMed] [Google Scholar]
- 7.Rose EA, Gelijns AC, Moskowitz AJ, et al. Long-term use of a left ventricular assist device for end-stage heart failure. N Engl J Med. 2001;345(20):1435–1443. [DOI] [PubMed] [Google Scholar]
- 8.Kamdar F, John R, Eckman P, Colvin-Adams M, Shumway SJ, Liao K. Postcardiac transplant survival in the current era in patients receiving continuous-flow left ventricular assist devices. J Thorac Cardiovasc Surg. 2013;145(2):575–581. [DOI] [PubMed] [Google Scholar]
- 9.Estep JD, Starling RC, Horstmanshof DA, et al. Risk assessment and comparative effectiveness of left ventricular assist device and medical management in ambulatory heart failure patients: results from the ROADMAP study. J Am Coll Cardiol. 2015;66(16):1747–1761. [DOI] [PubMed] [Google Scholar]
- 10.Joyce DL, Conte JV, Russell SD, Joyce LD, Chang DC. Disparities in access to left ventricular assist device therapy. J Surg Res. 2009; 152(1):111–117. [DOI] [PubMed] [Google Scholar]
- 11.Hutcheson SS, Phillips V, Patzer R, Smith A, Vega JD, Morris AA. Impact of insurance type on eligibility for advanced heart failure therapies and survival. Clin Transplant. 2018;32(8):e13328. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.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] [PubMed] [Google Scholar]
- 13.Hsich EM. Sex differences in advanced heart failure therapies. Circulation. 2019;139(8):1080–1093. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Breathett K, Yee E, Pool N, et al. Association of gender and race with allocation of advanced heart failure therapies. JAMA Netw Open. 2020;3(7):e2011044. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.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] [PMC free article] [PubMed] [Google Scholar]
- 16.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] [PMC free article] [PubMed] [Google Scholar]