- Email: [email protected]
Influence of race in heart failure and cardiac transplantation: Mortality differences are eliminated by specialized, comprehensive care
Journal of Cardiac Failure Vol. 9 No. 2 2003
Clinical Investigations
Influence of Race in Heart Failure and Cardiac Transplantation: Mortality Differences Are Eliminated by Specialized, Comprehensive Care SALPY V. PAMBOUKIAN, MD,* MARIA ROSA COSTANZO, MD,† PETER MEYER, PhD,* LINDA BARTLETT, RN,* MARY MCLEOD, RN,* ALAIN HEROUX, MD* Chicago, Illinois; Naperville, Illinois
ABSTRACT Background: Differences in mortality are thought to exist between African Americans and Caucasians with heart failure. These differences may be due to a variety of factors, including differences in disease process, socioeconomic status, and access to health care. Additionally, little data exist on racial differences between these two groups after cardiac transplantation. This study examines a single center, urban experience in treating African Americans and Caucasians with heart failure and after cardiac transplantation. We hypothesize that treatment in a specialized, comprehensive heart failure/cardiac transplantation program results in similar survival between African Americans and Caucasians. Methods: We retrospectively reviewed the Rush Heart Failure and Cardiac Transplant Database from July 1994 to August 2000. Variables analyzed in the cardiomyopathy patients included survival (until death, placement of left ventricular assist device or cardiac transplantation), number of hospitalizations per year, length of stay per year, and utilization of outpatient resources. Follow-up period was from initial visit to death, transplantation, or implantation of left ventricular assist device. In those who underwent cardiac transplantation, we examined rejection rates (cellular and humoral), rejection burden, hospitalization data, and 5-year survival. A subgroup bridged to cardiac transplantation with a left ventricular device was also analyzed. Results: Seven hundred thirty-four cardiomyopathy patients were identified: 203 were African Americans and 531 were Caucasians. The etiology of cardiomyopathy was more commonly ischemic in Caucasians as compared to non-ischemic in African Americans (P < .01). African Americans had more admissions to the hospital per year compared with Caucasians, 1.2 ⫾ 2.1 versus .5 ⫾ 1.1 (P < .01) with longer length of stay per year, 1.4 ⫾ 25.2 days versus 4.4 ⫾ 14.3 days (P < .01). Utilization of outpatient resources was significantly higher in African Americans compared with Caucasians with more use of
From the *Rush Presbyterian St. Luke’s Medical Center, Chicago, Illinois, and †Edward Hospital, Naperville, Illinois. Manuscript received October 30, 2002; revised manuscript received December 20, 2002; revised manuscript accepted December 23, 2002. Reprint requests: Dr. Salpy V. Pamboukian, Rush Presbyterian St. Luke’s Medical Center, 1725 W. Harrison St., Suite 439 PB, Chicago IL 60612. © 2003 Elsevier Inc. All rights reserved. 1071-9164/03/0902-0002$30.00/0 doi:10.1054/jcaf.2003.11
80
Influence of Race in Heart Failure and Cardiac Transplantation O Pamboukian et al
81
continuous inotropes (13% versus 6%, P < .01), intermittent inotropes (11% versus 5%, P < .01), and home nursing after hospital discharge (52% versus 32% of hospital discharges, P < .01). Survival by Kaplan-Meier analysis was comparable between the two groups (mean survival 1,470 ⫾ 72 days in African Americans versus 1521 ⫾ 46 days in Caucasians, log rank test [P ⫽ .6]). During this time, 30 African Americans and 73 Caucasians underwent cardiac transplantation. Fifty-three were bridged to transplantation with a left ventricular assist device (20 African Americans, 33 Caucasians). There were no differences in 5-year survival by Kaplan-Meier analysis despite higher peak preoperative panel reactive antibody levels in African Americans versus Caucasians (12% ⫾ 30% compared with 5% ⫾ 15%, P ⫽ .04), more overall treated rejection episodes per year in the African Americans (P < .01), as well as more posttransplant hospitalizations (2.2 ⫾ 1.2 times per year as compared with 1.7 ⫾ 2.1 times per year, P ⫽ .04). Conclusion: Delivery of care to heart failure patients in a comprehensive, specialized program results in similar survival regardless of race despite higher utilization of inpatient and outpatient resources. The finding that, after cardiac transplantation, African Americans do not have higher mortality rates, despite having higher rates of rejection overall and more hospitalizations, further supports the hypothesis that optimal care can improve outcomes despite unfavorable baseline clinical characteristics. Key Words: Cardiomyopathy, heart failure, congestive, cardiac transplantation, African Americans/blacks, Caucasians/whites.
Studies evaluating differences in mortality between African American and Caucasians patients with heart failure have produced conflicting results. A retrospective analysis of the Studies of Left Ventricular Dysfunction (SOLVD) trials demonstrated increased overall all-cause mortality with higher risk for death resulting from pump failure in blacks, even after adjusting for differences in education and socioeconomic status.1 In contrast, an examination of more than 90,000 individuals hospitalized in California between 1990 and 1993 demonstrated African Americans had lower mortality rates compared with whites and Asians in the 12 months after index hospitalization, despite higher rehospitalization rates.2 In a national cohort of 170,239 Medicare patients followed for 6 years after the index hospitalization for heart failure, white men had a 10% greater risk of mortality than did black men (P ⫽ .05).3 Other studies have also demonstrated either lower mortality in African Americans or no differences in survival between races.4,5 The influence of race on survival after cardiac transplantation is also an area of considerable debate. Differences in survival after transplantation may be related to donor-recipient human leukocyte antigen (HLA) mismatching, panel reactive antibody titers, and other immunologic factors. In addition, socioeconomic factors and medical noncompliance may also have roles to play. This study examined the experience at a single center, urban teaching hospital where patients received care in a comprehensive multidisciplinary heart failure and cardiac transplantation program. The hypothesis of the study was that the care provided by a specialized heart
failure and cardiac transplantation program eliminates differences in mortality between Caucasian and African American patients.
Methods Patient Population All patients referred to the Rush Heart Failure and Cardiac Transplant Program were eligible. Race was self-reported. All patients had some form of health insurance. Data Collection The Rush Heart Failure and Cardiac Transplant Database was reviewed from July 1994 to August 2000. Data collected and analyzed in the heart failure patients included etiology of heart disease, number of hospital admissions per year, length of hospital stay per year, utilization of oral heart failure medication, utilization of outpatient intermittent or continuous inotropes, use of home nursing after hospital discharge and survival. Endpoints were death, placement of left ventricular assistive device (LVAD), or cardiac transplantation. Data collected and analyzed in the cardiac transplant patients included donor characteristics, ischemic time, pump time, peak preoperative panel reactive antibody levels (PRA) by cytotoxic method, total treated rejections, treated cellular rejections, treated humoral rejections, rejection burden, hospitalizations per year, hospital
82 Journal of Cardiac Failure Vol. 9 No. 2 April 2003 length of stay per year, infection rates, and 5-year survival. In general, a cellular rejection episode was treated if it were International Society of Heart and Lung Transplantation (ISHLT) grade ⱖ2 in the first 12 months posttransplantation or ISHLT grade ⱖ3 thereafter at the discretion of the transplant cardiologist. Humoral rejection was defined clinically as allograft dysfunction on echocardiography or hemodynamic compromise on right heart catheterization without evidence of cellular rejection on endomyocardial biopsy and treated by one or more of pulse steroids (oral or intravenous), cyclophosphamide, plasmapheresis, intravenous gamma globulin (IVIG), or total lymphoid irradiation. To account for the number of biopsies and the severity of rejection over time, rejection burden was calculated by assigning a number to the ISHLT grade and dividing by the length of follow-up (ISHLT grade 1A and 1B ⫽ 1, grade 2 ⫽ 2, grade 3A and 3B ⫽ 3, grade 4 ⫽ 4). With regard to PRA, the patients were considered sensitized if the PRA was ⱖ10%. Elevated PRA pretransplantation may have been treated with IVIG, cyclophosphamide, mycophenolate mofetil, or plasmapheresis at the discretion of the transplant team. All patients with PRA ⱖ10% had a prospective crossmatch before cardiac transplantation. The number of HLA class I (A and B loci) and class II (DR locus) matches between donors and recipients were compared for African Americans and Caucasians. The number of HLA class I matches was calculated by summing the number of matches for both the A and B loci. Therefore a donor-recipient pair could have 0, 1, 2, 3, or 4 alleles matching at the A and B loci. Number or HLA type II matches were calculated by summing the number of matches for the DR locus. Therefore a donor-recipient pair could have 0, 1, or 2 alleles matching at the DR locus. Because only 2 patients had 3 or more HLA type I matches and only 2 patients had 2 HLA type II matches, the analysis was also performed comparing no matches vs. 1 or more matches to increase the power of the comparison. HLA compatibility was compared with the number of episodes of treated rejections per year, humoral rejections per year, cellular rejections per year, and overall survival. Univariate comparisons were performed using MannWhitney tests for continuous or ordinal outcomes and using Pearson’s chi-square with continuity correction or Fisher’s exact test for categorical variables, as appropriate. Kaplan-Meier survival curves were used for graphiTable 1. Etiology of Cardiomyopathy Etiology of Cardiomyopathy
African Americans
Caucasians
Ischemic Idiopathic/nonischemic Peripartum
36 (18%) 140 (69%) 9 (4%)
273 (51%) 200 (38%) 6 (1%)
P Value <.0001 <.0001 .005
cal comparisons of survival, with log-rank tests used to compare survival curves numerically.
Results We identified 203 African Americans and 531 Caucasians with heart failure. During this period, 30 African Americans and 73 Caucasians underwent cardiac transplantation. Of these, 20 African Americans and 33 Caucasians were bridged to heart transplantation with the Heart Mate pneumatic or vented electric LVAD (Thoratec, Pleasanton, CA). Heart Failure The etiology of heart failure for African Americans and Caucasians is summarized in Table 1. Caucasians had significantly more ischemic cardiomyopathy versus African Americans who had more idiopathic/nonischemic cardiomyopathy. Significantly more African American women had peripartum cardiomyopathy. Data regarding number of pregnancies were not collected. Upon initial presentation to the heart failure program, left ventricular ejection fraction (LVEF), as measured by echocardiography, was significantly lower in African Americans than in Caucasians (23% ⫾ 11% versus 27% ⫾ 12%, respectively, P < .01). At the end of the follow-up period, this difference was no longer observed, with improvement of LVEF to 32% ⫾ 18% in African Americans versus 33 ⫾ 14% in Caucasians (P ⫽ NS). Use of angiotensin-converting enzyme inhibitors (ACE) was 94% in African Americans versus 87% in Caucasians (P ⫽ .02). Beta-blocker use was similar between the two groups (52% in African Americans versus 54% in Caucasians, P ⫽ NS). African Americans had more admissions to the hospital per year compared with Caucasians, 1.2 ⫾ 2.1 versus .5 ⫾ 1.1 (P < .01) with longer length of stay per year, 10.4 ⫾ 25.2 days versus 4.4 ⫾ 14.3 days (P < .01). Utilization of outpatient resources in the form of inotropes and home nursing care after hospital discharge was also greater among African Americans. In the African American group, 52% of hospital discharges were followed up with nursing services in the home versus 32% in the Caucasian group (P < .01). Intermittent outpatient inotropes were used in 11% of African Americans versus 5% of Caucasians (P < .01). Continuous outpatient inotropes were used in 13% of African Americans versus 6% of Caucasians (P < .01). There was no difference in the number of LVADs placed in the two groups. Survival until death, transplantation, or LVAD placement was comparable with mean survival of 1,470 ⫾ 72 days in the African American group compared with 1,521 ⫾ 46 days in the Caucasian group (P ⫽ .6) (Fig. 1). When the data
Influence of Race in Heart Failure and Cardiac Transplantation O Pamboukian et al
Fig. 1. Survival to left ventricular assist device or transplantation.
were analyzed according to gender, there were no differences in survival between males and females, either within or between racial groups. Data regarding type of health care coverage are summarized in Table 2. All patients had some form of health insurance. African Americans were insured significantly more often by Medicare/Medicaid compared with Caucasians who were insured by preferred provider organization, point of service, health maintenance organization or indemnity plans.
Cardiac Transplantation Thirty African Americans (30/203 ⫽ 14%) underwent cardiac transplantation, with 20 patients preoperatively supported by a LVAD. The etiology of cardiomyopathy was idiopathic or nonischemic in 77%. Seventy-three Caucasians (73/531 ⫽ 14%) underwent cardiac transplantation, with 33 patients bridged by LVAD. The etiology of cardiomyopathy was ischemic in 67%. African Americans were significantly younger at the time of transplantation compared with Caucasians, 46 years of age versus 56 years of age (P < .01), and there were more women among the African Americans as compared with Caucasians, 30% versus 8% (P ⫽ .01).
Hemodynamic data at time of LVAD implantation are presented in Table 3. Some hemodynamic indices were significantly worse in African Americans at the time of LVAD placement, including pulmonary vascular resistance, which was higher, and cardiac output, which was lower. There was no difference between groups with regard to use of inotropes for right ventricular support after LVAD implantation. There was no difference in LVAD-related infection rates between the 2 groups. Mean number of days to transplantation bridged with LVAD was 106 ⫾ 103 days in the African American group versus 72 ⫾ 63 days in the Caucasian group (P ⫽ NS). One patient in each group died with LVAD before reaching cardiac transplantation. African Americans bridged with LVAD had higher pretransplant peak PRA compared with Caucasians, 22% ⫾ 34% versus 9% ⫾ 23% respectively (P ⫽ .03). African Americans received more red cell transfusions that were not leukocytereduced compared with Caucasians, 3 ⫾ 5 units compared with .4 ⫾ .7 units (P < .01). There was no difference in number of leukocyte-reduced red cell or platelet transfusions. At the time of cardiac transplantation, there were no differences in donor characteristics, ischemic time, or pump time between the two groups. Donor/recipient race match was 23% for African Americans compared with 73% for Caucasians (P < .0001). Peak preoperative PRA was higher in African Americans versus Caucasians, 12% ⫾ 30% compared with 5% ⫾ 15% (P ⫽ .04). All those with PRA greater than 10% had a prospective cross-match before transplantation. In some cases, pretreatment with IVIG, cyclophosphamide, mycophenylate, or plasmapheresis was undertaken at the discretion of the physician. Post-transplantation immunosuppression consisted of triple therapy with cyclosporine, prednisone, and azathioprine or mycophenylate mofetil in all patients. Subsequent changes in this regimen were made at the discretion of the attending cardiologist. African Americans had more episodes of rejection overall (P < .01), with more cellular (P ⫽ .03) as well as humoral rejections (P ⫽ .02). Rejection burden was similar between groups (P ⫽ NS). None of recipient/donor race mismatch, bridging with LVAD, or cytomegalovirus donor/recipient mismatch was associated with increased rejection after transplantation in
Table 2. Summary of Insurance Coverage Race African American Caucasian Totals
Medicare/Medicaid No. Patients (%)
PPO/POS No. Patients (%)
HMO No. Patients (%)
Indemnity No. Patients (%)
141 (72) 263 (51) 404
32 (16) 186 (36) 218
19 (10) 33 (6) 52
4 (2) 31 (6) 35
P < .001. *There were 5 self-pay patients who were excluded from the analysis.
83
Totals* 196 513 709
84 Journal of Cardiac Failure Vol. 9 No. 2 April 2003 Table 3. Hemodynamics at Time of LAVD Implantation Hemodynamic Measurements Right atrium (mm Hg) Pulmonary artery systolic (mm Hg) Pulmonary artery diastolic (mm Hg) Pulmonary artery mean (mm Hg) Pulmonary capillary wedge (mm Hg) Pulmonary vascular resistance (Wood units) Cardiac output (L/min) Cardiac index (L•min•m2)
African Americans 17 56 32 41 31 3.5 3.5 1.8
Caucasians 13 54 29 38 27 2.5 4.4 2.2
P Value .07 NS NS NS NS .03 .03 .07
NS, not significant.
either group. African Americans were hospitalized more frequently than Caucasians, 2.2 ⫾ 1.2 times per year as compared with 1.7 ⫾ 2.1 times per year (P ⫽ .04). Length of stay per year was also longer in the African American group, 17.2 ⫾ 15.6 day per year versus 11.9 ⫾ 20.0 days per year (P ⫽ .04). There was no difference in infection rates posttransplantation. HLA matching was poor in both groups. Out of a potential of 4 HLA type I allele matches, 96% of African American and 80% of Caucasians had 0 or only 1 match. Out of a potential of 2 HLA type II allele matches, 96% of African Americans and 98% of Caucasians had 0 or 1 match. There was no relationship between number of HLA matches and survival, cellular rejections per year, humoral rejections per year, or treated rejections per year. There was no difference in mortality after cardiac transplantation between African American and Caucasians requiring a LVAD preoperatively. Five-year survival postcardiac transplantation did not differ between the groups (Fig. 2).
Fig. 2. Survival after cardiac transplantation.
Discussion The principal finding of this study was that mortality was the same for African Americans and Caucasians with advanced heart failure and after heart transplantation despite a greater severity of illness in African Americans than in Caucasians. Indicators of more severe illness in the African American heart failure patients included a lower ejection fraction at time of index visit as well as more intensive utilization of inpatient and outpatient resources. Posttransplantation, African Americans had more episodes of treated rejection and more hospitalizations compared with their Caucasian counterparts. Although it is not clear in the literature whether mortality rates are higher in African Americans with heart failure, other epidemiologic features are more consistently observed and in keeping with our study findings. African Americans tend to be younger,1,2 with lower incidence of ischemia-related left ventricular dysfunction,1 and less angiographically documented coronary artery disease.6 African Americans have higher rates of uncontrolled, long-standing hypertension,1,7 likely accounting for the observed higher rate of nonischemic cardiomyopathy. Hospitalization rates are higher for African Americans,2 with longer length of stay.8 These differences in hospitalization rates may be due to a greater prevalence of comorbidities in African Americans with heart failure. When adjustments are made for these comorbidities, some of the observed racial differences are no longer present.9 Although there may be differences in the natural history of heart failure between African Americans and Caucasians the influence of societal factors on disease outcomes cannot be ignored. For many conditions an inverse relationship between socioeconomic status and mortality has been identified and the decline in overall death rates in the United States is not equally distributed among all socioeconomic groups.10 A disparity exists between racial groups with regard to health care access and quality. In the SUPPORT (Study to Understand Prognoses and Preferences for Outcomes and Risks of
Influence of Race in Heart Failure and Cardiac Transplantation O Pamboukian et al
Treatment) study, blacks were sicker on admission, used fewer resources after adjustment for severity of illness, and were less likely to have a cardiologist as their attending physician as compared with white patients.11 Ayanian and colleagues observed that blacks hospitalized with congestive heart failure received lower overall quality of care than other patients by both global evaluation of quality of care as well as by disease-specific criteria.12 African Americans are more likely to be referred from emergency departments, be treated at large urban, teaching hospitals, and to be uninsured.4,8 Higher hospitalization rates in African Americans, at least in part, reflect lack of outpatient support and patient education. Philbin and colleagues found black patients tended to be referred less often to skilled nursing facilities or to receive postdischarge home care.8 In 1 study the most common cause of cardiac decompensation in blacks was noncompliance with the prescribed medical regimen and dietary restriction,13 factors that can be modified by close outpatient monitoring and patient education. In our study, African Americans belonged to a lower socioeconomic stratum, as can be surmised by the significantly higher proportion of this ethnic group to be enrolled in Medicare and Medicaid programs. Despite this difference, African Americans received more intensive outpatient support compared with Caucasians. It may be that greater utilization of outpatient resources in a group with higher hospitalization rates and longer length of stay contributed to the elimination of differences in survival between African Americans and Caucasians. Utilization of ACE inhibitors at any time during the study period was significantly higher in the African American group, perhaps because Caucasians had already demonstrated an intolerance or contraindication to ACE inhibitors before referral to our program. Also of note, African Americans had a proportionally greater improvement in LVEF over the course of the treatment period. This, along with the higher use of ACE inhibitors, may have positively contributed to survival in the African American group. Interestingly, exactly the same percentage of African Americans and Caucasians underwent cardiac transplantation, indicating a lack of bias in the access of patients to cardiac transplantation. African Americans are thought to do worse after cardiac transplantation because of poorer recipient-donor HLA matching.14 The impact of HLA matching on outcomes after cardiac transplantation is controversial. Some authors have found that extent of HLA matching at the A, B, and DR loci significantly impacts survival,15 whereas others have not found this to be the case.16 Park and colleagues demonstrated that 10-year actuarial survival after heart transplantation was significantly lower in African Americans than in Caucasians. However, when the analysis was restricted to poorly matched
85
allografts for Class I and Class II, there was no racial difference in survival.17 In our study, patients in both groups received poorly matched allografts. African Americans had higher panel reactive antibody titers, at least in part, because of higher rates of red cell transfusion. Although African Americans had higher rates of cellular and humoral rejection posttransplantation, overall survival was similar between the two racial groups. This further supports the hypothesis that aggressive surveillance and medical intervention may negate some of the effects of a greater risk of rejection in the African American patient population. The main limitation of this study is that results were obtained from a retrospective review of a clinical database. No data were available regarding the duration of heart failure before the index visit. Detailed information on socioeconomic and educational status was not available.
Conclusion Health outcomes are the result of complex interactions between biomedical, socioeconomic, demographic, and psychosocial factors.18 Anthropologists, geneticists, and evolutionary biologists have largely abandoned using race as a method of classification because of the heterogeneity within given populations.19 Recent research has suggested that at least 80% of American blacks have at least one white ancestor.19 On a “level playing field,” many of the differences in mortality attributed to race may be eliminated. We have demonstrated similar survival rates between African Americans and Caucasians with heart failure and after cardiac transplantation when treated in an intensive, specialized, multidisciplinary program. The results of our study suggest that the type of care has a greater impact than race on the outcomes of patients with severe heart failure and after heart transplantation.
References 1. Dries DL, Exner DV, Gersh BJ, Cooper HA, Carson PE, Domanski MJ. Racial differences in the outcome of left ventricular dysfunction. N Engl J Med 1999;340:609-16. 2. Alexander M, Grumbach K, Remy L, Rowell R, Massie B. Congestive heart failure hospitalizations and survival in California: Patterns according to race/ethnicity. Am Heart J 1999;137:919-27. 3. Croft J, Giles WH, Pollard RA, Keenan NL, Casper ML, Anda RF. Heart failure survival among older adults in the United States: poor prognosis for an emerging epidemic in the Medicare population. Arch Intern Med 1999;159: 505-10.
86 Journal of Cardiac Failure Vol. 9 No. 2 April 2003 4. Gordon HS, Harper DL, Rosenthal GE. Racial variation in predicted and observed in-hospital death. A regional analysis. JAMA 1996;276:1639-44. 5. Dunlap SH, Sueta CA, Schwartz TA, Islam T, Webster K, Adams Jr. KA. Survival rates are similar between African-American and other races with heart failure (abstract). JACC 2000;35(Suppl A):177A. 6. Philbin EF, Weil HFC, Francis CA, Marx HJ, Jenkins PL, Pearson TA, et al. Race-related differences among patients with left ventricular dysfunction: observations from a biracial angiographic cohort. J Cardiac Failure 2000;6:187-93. 7. Mathew J, Davidson S, Narra L, Hafeez T, Garg R. Etiology and characteristics of congestive heart failure in blacks. Am J Cardiol 1996;78:1447-9. 8. Philbin EF, DiSalvo TG. Influence of race and gender on care process, resource use, and hospital based outcomes in congestive heart failure. Am J Cardiol 1998;82:76-81. 9. Alexander M, Grumbach K, Selby J, Brown A, Washington E. Hospitalization for congestive heart failure. Explaining racial differences. JAMA 1995;274:1037-42. 10. Pappas G, Queen S, Hadden W, Fisher G. The increasing disparity in mortality between socioeconomic groups in the United States, 1960 and 1986. N Engl J Med 1993; 329:103-9. 11. Phillips RS, Hamel MB, Teno JM, Bellamy P, Broset DK, Califf RM, et al. Race, resource use and survival in seriously ill hospitalized adults. J Gen Intern Med 1996; 11:387-96.
12. Ayanian JZ, Weissman JS, Chasan-Taber S, Epstein AM. Quality of care by race and gender for congestive heart failure and pneumonia. Med Care 1999;37:1260-9. 13. Ghali JK, Kadakia S, Cooper R, Ferlinz J. Precipitating factors leading to decompensation of heart failure. Traits among urban blacks. Arch Intern Med 1988;148:2013-6. 14. Reed E, Cohen DJ, Barr ML, Ho E, Reemtsma K, Rose E, et al. Effect of recipient gender and race on heart and kidney allograft survival. Transplant Proc 1992:24: 2670-1. 15. Opelz G, Wujciak T. The influence of HLA compatibility on graft survival after heart transplantation. N Engl J Med 1994;330:816-9. 16. Kerman RH, Kimball P, Scheinen S, Radovancevic B, Van Buren CT, Kahan BD, et al. The relationship among donor-recipient HLA mismatches, rejection and death from coronary artery disease in cardiac transplant recipients. Transplantation 1994;57:884-8. 17. Park MH, Tolman DE, Kimball PM. Disproportionate HLA matching may contribute to racial disparity in patient survival following cardiac transplantation. Clin Transplantation 1996;10:625-8. 18. Williams DR. Socioeconomic differentials in health: a review and redirection. Social Psychol Q 1990;53:81-99. 19. Stolley PD. Racial differences in the outcome of left ventricular dysfunction (letter). N Engl J Med 1999;341: 287.
Clinical Investigations
Influence of Race in Heart Failure and Cardiac Transplantation: Mortality Differences Are Eliminated by Specialized, Comprehensive Care SALPY V. PAMBOUKIAN, MD,* MARIA ROSA COSTANZO, MD,† PETER MEYER, PhD,* LINDA BARTLETT, RN,* MARY MCLEOD, RN,* ALAIN HEROUX, MD* Chicago, Illinois; Naperville, Illinois
ABSTRACT Background: Differences in mortality are thought to exist between African Americans and Caucasians with heart failure. These differences may be due to a variety of factors, including differences in disease process, socioeconomic status, and access to health care. Additionally, little data exist on racial differences between these two groups after cardiac transplantation. This study examines a single center, urban experience in treating African Americans and Caucasians with heart failure and after cardiac transplantation. We hypothesize that treatment in a specialized, comprehensive heart failure/cardiac transplantation program results in similar survival between African Americans and Caucasians. Methods: We retrospectively reviewed the Rush Heart Failure and Cardiac Transplant Database from July 1994 to August 2000. Variables analyzed in the cardiomyopathy patients included survival (until death, placement of left ventricular assist device or cardiac transplantation), number of hospitalizations per year, length of stay per year, and utilization of outpatient resources. Follow-up period was from initial visit to death, transplantation, or implantation of left ventricular assist device. In those who underwent cardiac transplantation, we examined rejection rates (cellular and humoral), rejection burden, hospitalization data, and 5-year survival. A subgroup bridged to cardiac transplantation with a left ventricular device was also analyzed. Results: Seven hundred thirty-four cardiomyopathy patients were identified: 203 were African Americans and 531 were Caucasians. The etiology of cardiomyopathy was more commonly ischemic in Caucasians as compared to non-ischemic in African Americans (P < .01). African Americans had more admissions to the hospital per year compared with Caucasians, 1.2 ⫾ 2.1 versus .5 ⫾ 1.1 (P < .01) with longer length of stay per year, 1.4 ⫾ 25.2 days versus 4.4 ⫾ 14.3 days (P < .01). Utilization of outpatient resources was significantly higher in African Americans compared with Caucasians with more use of
From the *Rush Presbyterian St. Luke’s Medical Center, Chicago, Illinois, and †Edward Hospital, Naperville, Illinois. Manuscript received October 30, 2002; revised manuscript received December 20, 2002; revised manuscript accepted December 23, 2002. Reprint requests: Dr. Salpy V. Pamboukian, Rush Presbyterian St. Luke’s Medical Center, 1725 W. Harrison St., Suite 439 PB, Chicago IL 60612. © 2003 Elsevier Inc. All rights reserved. 1071-9164/03/0902-0002$30.00/0 doi:10.1054/jcaf.2003.11
80
Influence of Race in Heart Failure and Cardiac Transplantation O Pamboukian et al
81
continuous inotropes (13% versus 6%, P < .01), intermittent inotropes (11% versus 5%, P < .01), and home nursing after hospital discharge (52% versus 32% of hospital discharges, P < .01). Survival by Kaplan-Meier analysis was comparable between the two groups (mean survival 1,470 ⫾ 72 days in African Americans versus 1521 ⫾ 46 days in Caucasians, log rank test [P ⫽ .6]). During this time, 30 African Americans and 73 Caucasians underwent cardiac transplantation. Fifty-three were bridged to transplantation with a left ventricular assist device (20 African Americans, 33 Caucasians). There were no differences in 5-year survival by Kaplan-Meier analysis despite higher peak preoperative panel reactive antibody levels in African Americans versus Caucasians (12% ⫾ 30% compared with 5% ⫾ 15%, P ⫽ .04), more overall treated rejection episodes per year in the African Americans (P < .01), as well as more posttransplant hospitalizations (2.2 ⫾ 1.2 times per year as compared with 1.7 ⫾ 2.1 times per year, P ⫽ .04). Conclusion: Delivery of care to heart failure patients in a comprehensive, specialized program results in similar survival regardless of race despite higher utilization of inpatient and outpatient resources. The finding that, after cardiac transplantation, African Americans do not have higher mortality rates, despite having higher rates of rejection overall and more hospitalizations, further supports the hypothesis that optimal care can improve outcomes despite unfavorable baseline clinical characteristics. Key Words: Cardiomyopathy, heart failure, congestive, cardiac transplantation, African Americans/blacks, Caucasians/whites.
Studies evaluating differences in mortality between African American and Caucasians patients with heart failure have produced conflicting results. A retrospective analysis of the Studies of Left Ventricular Dysfunction (SOLVD) trials demonstrated increased overall all-cause mortality with higher risk for death resulting from pump failure in blacks, even after adjusting for differences in education and socioeconomic status.1 In contrast, an examination of more than 90,000 individuals hospitalized in California between 1990 and 1993 demonstrated African Americans had lower mortality rates compared with whites and Asians in the 12 months after index hospitalization, despite higher rehospitalization rates.2 In a national cohort of 170,239 Medicare patients followed for 6 years after the index hospitalization for heart failure, white men had a 10% greater risk of mortality than did black men (P ⫽ .05).3 Other studies have also demonstrated either lower mortality in African Americans or no differences in survival between races.4,5 The influence of race on survival after cardiac transplantation is also an area of considerable debate. Differences in survival after transplantation may be related to donor-recipient human leukocyte antigen (HLA) mismatching, panel reactive antibody titers, and other immunologic factors. In addition, socioeconomic factors and medical noncompliance may also have roles to play. This study examined the experience at a single center, urban teaching hospital where patients received care in a comprehensive multidisciplinary heart failure and cardiac transplantation program. The hypothesis of the study was that the care provided by a specialized heart
failure and cardiac transplantation program eliminates differences in mortality between Caucasian and African American patients.
Methods Patient Population All patients referred to the Rush Heart Failure and Cardiac Transplant Program were eligible. Race was self-reported. All patients had some form of health insurance. Data Collection The Rush Heart Failure and Cardiac Transplant Database was reviewed from July 1994 to August 2000. Data collected and analyzed in the heart failure patients included etiology of heart disease, number of hospital admissions per year, length of hospital stay per year, utilization of oral heart failure medication, utilization of outpatient intermittent or continuous inotropes, use of home nursing after hospital discharge and survival. Endpoints were death, placement of left ventricular assistive device (LVAD), or cardiac transplantation. Data collected and analyzed in the cardiac transplant patients included donor characteristics, ischemic time, pump time, peak preoperative panel reactive antibody levels (PRA) by cytotoxic method, total treated rejections, treated cellular rejections, treated humoral rejections, rejection burden, hospitalizations per year, hospital
82 Journal of Cardiac Failure Vol. 9 No. 2 April 2003 length of stay per year, infection rates, and 5-year survival. In general, a cellular rejection episode was treated if it were International Society of Heart and Lung Transplantation (ISHLT) grade ⱖ2 in the first 12 months posttransplantation or ISHLT grade ⱖ3 thereafter at the discretion of the transplant cardiologist. Humoral rejection was defined clinically as allograft dysfunction on echocardiography or hemodynamic compromise on right heart catheterization without evidence of cellular rejection on endomyocardial biopsy and treated by one or more of pulse steroids (oral or intravenous), cyclophosphamide, plasmapheresis, intravenous gamma globulin (IVIG), or total lymphoid irradiation. To account for the number of biopsies and the severity of rejection over time, rejection burden was calculated by assigning a number to the ISHLT grade and dividing by the length of follow-up (ISHLT grade 1A and 1B ⫽ 1, grade 2 ⫽ 2, grade 3A and 3B ⫽ 3, grade 4 ⫽ 4). With regard to PRA, the patients were considered sensitized if the PRA was ⱖ10%. Elevated PRA pretransplantation may have been treated with IVIG, cyclophosphamide, mycophenolate mofetil, or plasmapheresis at the discretion of the transplant team. All patients with PRA ⱖ10% had a prospective crossmatch before cardiac transplantation. The number of HLA class I (A and B loci) and class II (DR locus) matches between donors and recipients were compared for African Americans and Caucasians. The number of HLA class I matches was calculated by summing the number of matches for both the A and B loci. Therefore a donor-recipient pair could have 0, 1, 2, 3, or 4 alleles matching at the A and B loci. Number or HLA type II matches were calculated by summing the number of matches for the DR locus. Therefore a donor-recipient pair could have 0, 1, or 2 alleles matching at the DR locus. Because only 2 patients had 3 or more HLA type I matches and only 2 patients had 2 HLA type II matches, the analysis was also performed comparing no matches vs. 1 or more matches to increase the power of the comparison. HLA compatibility was compared with the number of episodes of treated rejections per year, humoral rejections per year, cellular rejections per year, and overall survival. Univariate comparisons were performed using MannWhitney tests for continuous or ordinal outcomes and using Pearson’s chi-square with continuity correction or Fisher’s exact test for categorical variables, as appropriate. Kaplan-Meier survival curves were used for graphiTable 1. Etiology of Cardiomyopathy Etiology of Cardiomyopathy
African Americans
Caucasians
Ischemic Idiopathic/nonischemic Peripartum
36 (18%) 140 (69%) 9 (4%)
273 (51%) 200 (38%) 6 (1%)
P Value <.0001 <.0001 .005
cal comparisons of survival, with log-rank tests used to compare survival curves numerically.
Results We identified 203 African Americans and 531 Caucasians with heart failure. During this period, 30 African Americans and 73 Caucasians underwent cardiac transplantation. Of these, 20 African Americans and 33 Caucasians were bridged to heart transplantation with the Heart Mate pneumatic or vented electric LVAD (Thoratec, Pleasanton, CA). Heart Failure The etiology of heart failure for African Americans and Caucasians is summarized in Table 1. Caucasians had significantly more ischemic cardiomyopathy versus African Americans who had more idiopathic/nonischemic cardiomyopathy. Significantly more African American women had peripartum cardiomyopathy. Data regarding number of pregnancies were not collected. Upon initial presentation to the heart failure program, left ventricular ejection fraction (LVEF), as measured by echocardiography, was significantly lower in African Americans than in Caucasians (23% ⫾ 11% versus 27% ⫾ 12%, respectively, P < .01). At the end of the follow-up period, this difference was no longer observed, with improvement of LVEF to 32% ⫾ 18% in African Americans versus 33 ⫾ 14% in Caucasians (P ⫽ NS). Use of angiotensin-converting enzyme inhibitors (ACE) was 94% in African Americans versus 87% in Caucasians (P ⫽ .02). Beta-blocker use was similar between the two groups (52% in African Americans versus 54% in Caucasians, P ⫽ NS). African Americans had more admissions to the hospital per year compared with Caucasians, 1.2 ⫾ 2.1 versus .5 ⫾ 1.1 (P < .01) with longer length of stay per year, 10.4 ⫾ 25.2 days versus 4.4 ⫾ 14.3 days (P < .01). Utilization of outpatient resources in the form of inotropes and home nursing care after hospital discharge was also greater among African Americans. In the African American group, 52% of hospital discharges were followed up with nursing services in the home versus 32% in the Caucasian group (P < .01). Intermittent outpatient inotropes were used in 11% of African Americans versus 5% of Caucasians (P < .01). Continuous outpatient inotropes were used in 13% of African Americans versus 6% of Caucasians (P < .01). There was no difference in the number of LVADs placed in the two groups. Survival until death, transplantation, or LVAD placement was comparable with mean survival of 1,470 ⫾ 72 days in the African American group compared with 1,521 ⫾ 46 days in the Caucasian group (P ⫽ .6) (Fig. 1). When the data
Influence of Race in Heart Failure and Cardiac Transplantation O Pamboukian et al
Fig. 1. Survival to left ventricular assist device or transplantation.
were analyzed according to gender, there were no differences in survival between males and females, either within or between racial groups. Data regarding type of health care coverage are summarized in Table 2. All patients had some form of health insurance. African Americans were insured significantly more often by Medicare/Medicaid compared with Caucasians who were insured by preferred provider organization, point of service, health maintenance organization or indemnity plans.
Cardiac Transplantation Thirty African Americans (30/203 ⫽ 14%) underwent cardiac transplantation, with 20 patients preoperatively supported by a LVAD. The etiology of cardiomyopathy was idiopathic or nonischemic in 77%. Seventy-three Caucasians (73/531 ⫽ 14%) underwent cardiac transplantation, with 33 patients bridged by LVAD. The etiology of cardiomyopathy was ischemic in 67%. African Americans were significantly younger at the time of transplantation compared with Caucasians, 46 years of age versus 56 years of age (P < .01), and there were more women among the African Americans as compared with Caucasians, 30% versus 8% (P ⫽ .01).
Hemodynamic data at time of LVAD implantation are presented in Table 3. Some hemodynamic indices were significantly worse in African Americans at the time of LVAD placement, including pulmonary vascular resistance, which was higher, and cardiac output, which was lower. There was no difference between groups with regard to use of inotropes for right ventricular support after LVAD implantation. There was no difference in LVAD-related infection rates between the 2 groups. Mean number of days to transplantation bridged with LVAD was 106 ⫾ 103 days in the African American group versus 72 ⫾ 63 days in the Caucasian group (P ⫽ NS). One patient in each group died with LVAD before reaching cardiac transplantation. African Americans bridged with LVAD had higher pretransplant peak PRA compared with Caucasians, 22% ⫾ 34% versus 9% ⫾ 23% respectively (P ⫽ .03). African Americans received more red cell transfusions that were not leukocytereduced compared with Caucasians, 3 ⫾ 5 units compared with .4 ⫾ .7 units (P < .01). There was no difference in number of leukocyte-reduced red cell or platelet transfusions. At the time of cardiac transplantation, there were no differences in donor characteristics, ischemic time, or pump time between the two groups. Donor/recipient race match was 23% for African Americans compared with 73% for Caucasians (P < .0001). Peak preoperative PRA was higher in African Americans versus Caucasians, 12% ⫾ 30% compared with 5% ⫾ 15% (P ⫽ .04). All those with PRA greater than 10% had a prospective cross-match before transplantation. In some cases, pretreatment with IVIG, cyclophosphamide, mycophenylate, or plasmapheresis was undertaken at the discretion of the physician. Post-transplantation immunosuppression consisted of triple therapy with cyclosporine, prednisone, and azathioprine or mycophenylate mofetil in all patients. Subsequent changes in this regimen were made at the discretion of the attending cardiologist. African Americans had more episodes of rejection overall (P < .01), with more cellular (P ⫽ .03) as well as humoral rejections (P ⫽ .02). Rejection burden was similar between groups (P ⫽ NS). None of recipient/donor race mismatch, bridging with LVAD, or cytomegalovirus donor/recipient mismatch was associated with increased rejection after transplantation in
Table 2. Summary of Insurance Coverage Race African American Caucasian Totals
Medicare/Medicaid No. Patients (%)
PPO/POS No. Patients (%)
HMO No. Patients (%)
Indemnity No. Patients (%)
141 (72) 263 (51) 404
32 (16) 186 (36) 218
19 (10) 33 (6) 52
4 (2) 31 (6) 35
P < .001. *There were 5 self-pay patients who were excluded from the analysis.
83
Totals* 196 513 709
84 Journal of Cardiac Failure Vol. 9 No. 2 April 2003 Table 3. Hemodynamics at Time of LAVD Implantation Hemodynamic Measurements Right atrium (mm Hg) Pulmonary artery systolic (mm Hg) Pulmonary artery diastolic (mm Hg) Pulmonary artery mean (mm Hg) Pulmonary capillary wedge (mm Hg) Pulmonary vascular resistance (Wood units) Cardiac output (L/min) Cardiac index (L•min•m2)
African Americans 17 56 32 41 31 3.5 3.5 1.8
Caucasians 13 54 29 38 27 2.5 4.4 2.2
P Value .07 NS NS NS NS .03 .03 .07
NS, not significant.
either group. African Americans were hospitalized more frequently than Caucasians, 2.2 ⫾ 1.2 times per year as compared with 1.7 ⫾ 2.1 times per year (P ⫽ .04). Length of stay per year was also longer in the African American group, 17.2 ⫾ 15.6 day per year versus 11.9 ⫾ 20.0 days per year (P ⫽ .04). There was no difference in infection rates posttransplantation. HLA matching was poor in both groups. Out of a potential of 4 HLA type I allele matches, 96% of African American and 80% of Caucasians had 0 or only 1 match. Out of a potential of 2 HLA type II allele matches, 96% of African Americans and 98% of Caucasians had 0 or 1 match. There was no relationship between number of HLA matches and survival, cellular rejections per year, humoral rejections per year, or treated rejections per year. There was no difference in mortality after cardiac transplantation between African American and Caucasians requiring a LVAD preoperatively. Five-year survival postcardiac transplantation did not differ between the groups (Fig. 2).
Fig. 2. Survival after cardiac transplantation.
Discussion The principal finding of this study was that mortality was the same for African Americans and Caucasians with advanced heart failure and after heart transplantation despite a greater severity of illness in African Americans than in Caucasians. Indicators of more severe illness in the African American heart failure patients included a lower ejection fraction at time of index visit as well as more intensive utilization of inpatient and outpatient resources. Posttransplantation, African Americans had more episodes of treated rejection and more hospitalizations compared with their Caucasian counterparts. Although it is not clear in the literature whether mortality rates are higher in African Americans with heart failure, other epidemiologic features are more consistently observed and in keeping with our study findings. African Americans tend to be younger,1,2 with lower incidence of ischemia-related left ventricular dysfunction,1 and less angiographically documented coronary artery disease.6 African Americans have higher rates of uncontrolled, long-standing hypertension,1,7 likely accounting for the observed higher rate of nonischemic cardiomyopathy. Hospitalization rates are higher for African Americans,2 with longer length of stay.8 These differences in hospitalization rates may be due to a greater prevalence of comorbidities in African Americans with heart failure. When adjustments are made for these comorbidities, some of the observed racial differences are no longer present.9 Although there may be differences in the natural history of heart failure between African Americans and Caucasians the influence of societal factors on disease outcomes cannot be ignored. For many conditions an inverse relationship between socioeconomic status and mortality has been identified and the decline in overall death rates in the United States is not equally distributed among all socioeconomic groups.10 A disparity exists between racial groups with regard to health care access and quality. In the SUPPORT (Study to Understand Prognoses and Preferences for Outcomes and Risks of
Influence of Race in Heart Failure and Cardiac Transplantation O Pamboukian et al
Treatment) study, blacks were sicker on admission, used fewer resources after adjustment for severity of illness, and were less likely to have a cardiologist as their attending physician as compared with white patients.11 Ayanian and colleagues observed that blacks hospitalized with congestive heart failure received lower overall quality of care than other patients by both global evaluation of quality of care as well as by disease-specific criteria.12 African Americans are more likely to be referred from emergency departments, be treated at large urban, teaching hospitals, and to be uninsured.4,8 Higher hospitalization rates in African Americans, at least in part, reflect lack of outpatient support and patient education. Philbin and colleagues found black patients tended to be referred less often to skilled nursing facilities or to receive postdischarge home care.8 In 1 study the most common cause of cardiac decompensation in blacks was noncompliance with the prescribed medical regimen and dietary restriction,13 factors that can be modified by close outpatient monitoring and patient education. In our study, African Americans belonged to a lower socioeconomic stratum, as can be surmised by the significantly higher proportion of this ethnic group to be enrolled in Medicare and Medicaid programs. Despite this difference, African Americans received more intensive outpatient support compared with Caucasians. It may be that greater utilization of outpatient resources in a group with higher hospitalization rates and longer length of stay contributed to the elimination of differences in survival between African Americans and Caucasians. Utilization of ACE inhibitors at any time during the study period was significantly higher in the African American group, perhaps because Caucasians had already demonstrated an intolerance or contraindication to ACE inhibitors before referral to our program. Also of note, African Americans had a proportionally greater improvement in LVEF over the course of the treatment period. This, along with the higher use of ACE inhibitors, may have positively contributed to survival in the African American group. Interestingly, exactly the same percentage of African Americans and Caucasians underwent cardiac transplantation, indicating a lack of bias in the access of patients to cardiac transplantation. African Americans are thought to do worse after cardiac transplantation because of poorer recipient-donor HLA matching.14 The impact of HLA matching on outcomes after cardiac transplantation is controversial. Some authors have found that extent of HLA matching at the A, B, and DR loci significantly impacts survival,15 whereas others have not found this to be the case.16 Park and colleagues demonstrated that 10-year actuarial survival after heart transplantation was significantly lower in African Americans than in Caucasians. However, when the analysis was restricted to poorly matched
85
allografts for Class I and Class II, there was no racial difference in survival.17 In our study, patients in both groups received poorly matched allografts. African Americans had higher panel reactive antibody titers, at least in part, because of higher rates of red cell transfusion. Although African Americans had higher rates of cellular and humoral rejection posttransplantation, overall survival was similar between the two racial groups. This further supports the hypothesis that aggressive surveillance and medical intervention may negate some of the effects of a greater risk of rejection in the African American patient population. The main limitation of this study is that results were obtained from a retrospective review of a clinical database. No data were available regarding the duration of heart failure before the index visit. Detailed information on socioeconomic and educational status was not available.
Conclusion Health outcomes are the result of complex interactions between biomedical, socioeconomic, demographic, and psychosocial factors.18 Anthropologists, geneticists, and evolutionary biologists have largely abandoned using race as a method of classification because of the heterogeneity within given populations.19 Recent research has suggested that at least 80% of American blacks have at least one white ancestor.19 On a “level playing field,” many of the differences in mortality attributed to race may be eliminated. We have demonstrated similar survival rates between African Americans and Caucasians with heart failure and after cardiac transplantation when treated in an intensive, specialized, multidisciplinary program. The results of our study suggest that the type of care has a greater impact than race on the outcomes of patients with severe heart failure and after heart transplantation.
References 1. Dries DL, Exner DV, Gersh BJ, Cooper HA, Carson PE, Domanski MJ. Racial differences in the outcome of left ventricular dysfunction. N Engl J Med 1999;340:609-16. 2. Alexander M, Grumbach K, Remy L, Rowell R, Massie B. Congestive heart failure hospitalizations and survival in California: Patterns according to race/ethnicity. Am Heart J 1999;137:919-27. 3. Croft J, Giles WH, Pollard RA, Keenan NL, Casper ML, Anda RF. Heart failure survival among older adults in the United States: poor prognosis for an emerging epidemic in the Medicare population. Arch Intern Med 1999;159: 505-10.
86 Journal of Cardiac Failure Vol. 9 No. 2 April 2003 4. Gordon HS, Harper DL, Rosenthal GE. Racial variation in predicted and observed in-hospital death. A regional analysis. JAMA 1996;276:1639-44. 5. Dunlap SH, Sueta CA, Schwartz TA, Islam T, Webster K, Adams Jr. KA. Survival rates are similar between African-American and other races with heart failure (abstract). JACC 2000;35(Suppl A):177A. 6. Philbin EF, Weil HFC, Francis CA, Marx HJ, Jenkins PL, Pearson TA, et al. Race-related differences among patients with left ventricular dysfunction: observations from a biracial angiographic cohort. J Cardiac Failure 2000;6:187-93. 7. Mathew J, Davidson S, Narra L, Hafeez T, Garg R. Etiology and characteristics of congestive heart failure in blacks. Am J Cardiol 1996;78:1447-9. 8. Philbin EF, DiSalvo TG. Influence of race and gender on care process, resource use, and hospital based outcomes in congestive heart failure. Am J Cardiol 1998;82:76-81. 9. Alexander M, Grumbach K, Selby J, Brown A, Washington E. Hospitalization for congestive heart failure. Explaining racial differences. JAMA 1995;274:1037-42. 10. Pappas G, Queen S, Hadden W, Fisher G. The increasing disparity in mortality between socioeconomic groups in the United States, 1960 and 1986. N Engl J Med 1993; 329:103-9. 11. Phillips RS, Hamel MB, Teno JM, Bellamy P, Broset DK, Califf RM, et al. Race, resource use and survival in seriously ill hospitalized adults. J Gen Intern Med 1996; 11:387-96.
12. Ayanian JZ, Weissman JS, Chasan-Taber S, Epstein AM. Quality of care by race and gender for congestive heart failure and pneumonia. Med Care 1999;37:1260-9. 13. Ghali JK, Kadakia S, Cooper R, Ferlinz J. Precipitating factors leading to decompensation of heart failure. Traits among urban blacks. Arch Intern Med 1988;148:2013-6. 14. Reed E, Cohen DJ, Barr ML, Ho E, Reemtsma K, Rose E, et al. Effect of recipient gender and race on heart and kidney allograft survival. Transplant Proc 1992:24: 2670-1. 15. Opelz G, Wujciak T. The influence of HLA compatibility on graft survival after heart transplantation. N Engl J Med 1994;330:816-9. 16. Kerman RH, Kimball P, Scheinen S, Radovancevic B, Van Buren CT, Kahan BD, et al. The relationship among donor-recipient HLA mismatches, rejection and death from coronary artery disease in cardiac transplant recipients. Transplantation 1994;57:884-8. 17. Park MH, Tolman DE, Kimball PM. Disproportionate HLA matching may contribute to racial disparity in patient survival following cardiac transplantation. Clin Transplantation 1996;10:625-8. 18. Williams DR. Socioeconomic differentials in health: a review and redirection. Social Psychol Q 1990;53:81-99. 19. Stolley PD. Racial differences in the outcome of left ventricular dysfunction (letter). N Engl J Med 1999;341: 287.