- Email: [email protected]
Survival Among Mexican-Americans, Non-Hispanic Whites, and African-Americans With End-Stage Renal Disease: The Emergence of a Minority Pattern of Increased Incidence and Prolonged Survival
Survival Among Mexican-Americans, Non-Hispanic Whites, and African-Americans With End-Stage Renal Disease: The Emergence of a Minority Pattern of Increased Incidence and Prolonged Survival Jacqueline A. Pugh, MD, Michael R. Tuley, PhD, and Srabashi Basu, PhD • We undertook this study to determine whether there is a significant difference in survival on treatment for endstage renal disease between Mexican-Americans, non-Hispanic whites, and African-Americans. A database covering the years 1975 to 1986 was obtained from the Texas Kidney Health Program. Eight-eight percent to 90% of patients starting renal replacement therapy in Texas were included in this database. The patients were followed until death, for 3 years after successful transplantation, or until they were lost to fOllow-up. Life table analysis as well as ageadjusted analysis using the Cox proportional hazards model were performed comparing ethnic/racial groups, disease etiology, and treatment type. In life-table analyses, African-Americans and Mexican-Americans had a survival advantage in most age, disease, and treatment groups. With age adjustment, this survival advantage remained for all etiologies combined, for diabetes and hypertension cases, and for patients receiving hemodialysis in a center. Multivariate analysis revealed a persistent survival advantage for Mexican-Americans independent of traditional predictor variables, such as age, disease etiology, treatment type, or size of the center in which they received treatment. In this same analysis, African-Americans showed an advantage in the older age groups. Both African-Americans and MexicanAmericans on renal replacement therapy have an increased survival advantage compared with non-Hispanic whites. Given the additional burden of increased incidence of end-stage renal disease in these groups, the cost of renal replacement therapy for these minorities is disproportionately high. Further study should be aimed at elucidation of the mechanisms by which minorities achieve their survival advantage. © 1994 by the National Kidney Foundation, Inc. INDEX WORDS: Survival; end-stage renal disease; Hispanic; Mexican-American; minorities.
T
HE COST OF treatment of end-stage renal disease (ESRD) is enormous in both actual dollars (7.5 billion dollars per year!), and human suffering. The enrollment in the Medicare EndStage Renal Disease Program has grown steadily since its inception in 1973.2,3 Minorities are disproportionately represented among the individuals requiring renal replacement therapy.2,4 African-Americans have a threefold to fourfold higher incidence of treatment for ESRD than whites. 2,3 Fifty-six percent of new cases ofESRD in African-Americans are secondary to longstanding chronic diseases, such as hypertension and diabetes. 3 Other minorities, including Mexican-Americans,4 Native Americans,3 and the "all other" category in the Medicare data,2 also have higher rates ofESRD treatment than whites. Despite a higher incidence of ESRD, AfricanAmerican, Asian, and Native American dialysis patients also have been shown to have better 1year survival than whites. 5 African-Americans also have better 5-year survival rates. 2 The life expectancy for all ESRD patients is still significantly below the general population. 3 Numerous factors have been found to affect the survival of patients receiving renal replacement therapy, including age at onset oftreatment,5,6 type of treatment, 7 comorbid disease prior to treatment,6,8 etiology of renal disease,3,5,6 size of the center in
which treatment is given/ dialyzer reuse,7 voluntary withdrawal,9 depression,IO,!! and country in which the person is dialyzed,!2 The reasons African-Americans, Asians, and Native Americans have better survival are not intuitively obvious. Voluntary withdrawal from dialysis is lower among minorities,5,9 but the percentage of deaths explained by this mechanism is relatively small. African-Americans are less likely to undergo renal transplantation 2 and therefore healthier patients remain on dialysis. However, the survival advantage for African-Americans is From the Department ofMedicine, the Division of General Internal Medicine, University of Texas Health Science Center at San Antonio and the Audie L. Murphy Memorial Veterans Hospital, San Antonio, TX Received January 25, 1994; accepted in revised form February 15, 1994, Supported by grants from the RGK Foundation, National Institute of Diabetes and Digestive and Kidney Diseases (DK38392); the Geriatric Research, Education, and Clinical Center at the Audie L. Murphy Memorial Veterans Hospital; and the Mexican American Medical Treatment Ef fectiveness Center, a Minority Medical Treatment Effectiveness Program Center, Agency for Health Care Policy and Research (lU01HS0739701). Address reprint requests to Jacqueline A. Pugh, MD, Ambulatory Care llC6, Audie L. Murphy Memorial Veterans Hospital, 7400 Merton Minter Blvd, San Antonio, TX 78284. © 1994 by the National Kidney Foundation, Inc. 0272-6386/94/2306-0006$3.00/0
American Journal of Kidney Diseases, Vol 23, No 6 (June), 1994: pp 803-807
803
PUGH, TULEY, AND BASU
804
Table 1. Reason for Censorship From the Database Non-Hispanic Whites ('¥oj
Mexican-Americans ('¥oj
African-Americans ('¥oj
Death Relocation Withdrawal Lost to follow-up Recovered function Currently treated
2,672 (49.6) 131 (2.4) 6 (0.1) 672 (12.5) 94 (1.7) 1,814 (33.7)
1,529 (48.3) 32 (1.0) 7 (0.2) 310 (9.8) 31 (1) 1,254 (39.7)
1,407 (41.1) 36 (1.0) 0(0) 437 (12.8) 38 (1.1) 1,508 (44.0)
Total
5,389 (100)
3,163 (100)
3,426 (100)
even higher in the older age groups, in which transplantation is unlikely to be offered to either race. Lower income has been shown to predict higher coronary disease mortality l3 and, since the majority of ESRD patients die from coronary disease, the minority survival advantage is opposite of what would be predicted based on income. Another possible explanation is that sicker minority patients are not being offered renal replacement therapy. While there are no population-based data to refute this at present, the high rates of ESRD treatment among minorities and the continuing growth in the incidence of treatment make this an unlikely explanation for the minority survival advantage. Mexican-Americans, at 14 million, are the second largest minority group in the United States and their numbers are growing five times faster than the rest of the population. l4 • l5 Mexican-Americans are known to have a threefold higher incidence of ESRD compared with nonHispanic whites and a sixfold higher incidence of diabetic ESRD.4 Given that Mexican-Americans are similar to African-Americans with regard to increased incidence of ESRD, we posed the question, do Mexican-Americans receiving treatment for ESRD have a prolonged survival when compared with non-Hispanic whites? MATERIALS AND METHODS The Texas Kidney Health Program, a division of the Texas Department of Health, provided a database. The population included in this database has been described previously.' Briefly, all legal residents of Texas with ESRD may apply for enrollment in the Texas Kidney Health Program, which provides coverage for drugs and transportation as well as for expenses incurred prior to Medicare eligibility. Eighty-eight percent to 90% of all patients starting dialysis in Texas are registered with the Texas Kidney Health Program.' These analyses were restricted to patients enrolled between January I, 1975, and September 30, 1985. Patients younger than 20 years at the start of dialysis were excluded. The analyses included 11,978 subjects.
Total ('¥oj
5,608 (46.8) 199(1.7) 13 (0.1) 1,419 (11.9) 163(1.4) 4,576 (38.2) 11,978 (100)
Simple life-table analyses were performed using BMDP IL to test the proportionality of the independent variables. 16 The Cox proportional hazard model was used to age adjust the data in otherwise univariate analyses and to build a more comprehensive multivariate model using BMDP 2L. 16 Home hemodialysis patients were excluded from the multivariate analysis both because of small numbers and because their survival curve was not proportional to the other three treatments. Patients were censored at the time oflast contact with the Texas Kidney Health Program. Variables available for analysis included age at the start of dialysis, etiology of renal disease, race/ethnicity, sex, size of the dialysis treatment center, first treatment, current treatment, and one interim treatment if different. The above-listed variables as well as interaction variables between ethnicity/race and age, treatment, and disease etiology were tested. Current treatment was used unless the patient underwent transplantation. For those who had received a transplant, "transplantation" was assigned as their treatment and the date of the first transplant was taken as the date of entry. The structure of the database did not allow assignment of person-years on a given treatment because months or dates on each treatment were not included in the database. However, for those alive at the time the database was obtained, 80.4% were receiving treatment that was the same as their entry treatment. Not included were any measure of comorbid illness, dialyzer reuse, psychological assessment, or socioeconomic status.
RESULTS
Table 1 shows the percentage and number of cases in the database censored because of death, geographic relocation out of the state, voluntary withdrawal from dialysis, loss to follow-up, or recovery of renal function no longer requiring dialysis; it also included those patients currently receiving treatment. Overall, 11.9% of patients were lost to follow-up, and a lower percentage of Mexican-Americans were lost to follow-up than either non-Hispanic whites or African-Americans. The percentage of patients with voluntary withdrawal did not differ between groups, but was low in comparison to the Medicare data 5,9 and is not likely to have been accurately reported. Individuals who voluntarily withdrew from treatment were censored at the time of withdrawal
805
PROLONGED ESRD SURVIVAL FOR MINORITIES Table 2. Distribution of Patients by Age, Disease Etiology or Treatment, and Ethnic Group at the Start of End-Stage Renal Disease Treatment
Table 3. Cumulative Survival Rate for Each Ethnic Group at 1 and 5 Years, Stratified by Age" 1 Yr
Age group (yrs) 20-29 30-39 40-49 50-59 60-69 70-79 >80 All ages Oisease HTN GMN OM OTH All diseases Treatment CTR CAPO HOME TRANS All treatments
Non-Hispanic Whites
MexicanAmericans
AfricanAmericans
Total
555 699 791 1,121 1,284 789 147 5,386
232 301 447 910 808 398 66 3,162
323 453 595 870 760 356 65 3,422
1,110 1,453 1,833 2,901 2,852 1,543 278 11,970*
988 992 1,186 2,220 5,386
455 465 1,389 853 3,162
1,508 385 687 842 3,422
2,951 1,842 3,262 3,915 11,970*
4,106 462 163 649 5,380
2,734 153 11 264 3,162
3,021 115 24 261 3,421
9,861 730 198 1,174 11,963*
Abbreviations: HTN, hypertension; GMN, glomerulonephritis; OM, diabetes mellitus; OTH, other; CTR, center hemodialysis; CAPO, continuous ambulatory peritoneal dialysis; HOME, home hemodialysis; TRANS, transplant. * Each category had patients who had missing data for a variable: eight for age, eight for disease, and 15 for treatment.
because their final outcome was not available in the Texas Kidney Health Program database. Table 2 displays the distribution of patients at onset ofESRD treatment (time 0) by age, disease etiology or treatment, and ethnicity_ Table 3 displays age-specific 1- and 5-year survival for each ethnic group_ While MexicanAmericans display a small to moderate survival advantage over non-Hispanic whites in all but three age groups (I-year 70- to 79-year age group, 5-year 50- to 59-year age group, and 5-year 60to 69-year age group), African-Americans show no advantage in the 20- to 29-year age group and then progressively better survival in all other age groups. Table 4 shows the percentage of survival at 1 and 5 years for each ethnic group, stratified by disease etiology and treatment type_ Among disease etiologies, there is a consistent pattern of improved survival of African-Americans over non-Hispanic whites at both 1 and 5 years. Except for diabetic survival at 5 years, the survival of Mexican-Americans falls between that of AfricanAmericans and non-Hispanic whites, sometimes
Age (yr) 20-29 30-39 40-49 50-59 60-69 70-79 >80 All ages combined
5 Yr
NHW
MA
AA
Total
NHW
MA
AA
Total
91 85 84 81 73 66 61
96 91 87 83 75 65 63
90 89 90 86 82 81 69
92 88 87 83 76 69 63
73 52 47 42 26 16 7
83 69 54 39 23 18 17
61 68 61 46 38 33 16
72 60 53 42 28 21 10
78
81
86
38
40
49
Abbreviations: NHW, non-Hispanic whites; MA, Mexican-Americans; AA, African-Americans. * This proportion takes into account censored patients throughout the 5 years.
closer to non-Hispanic whites and sometimes closer to African-Americans. For treatment type, treatment received in a hemodialysis center presents the same pattern as among disease etiologies, with African-Americans having the best survival followed by Mexican-Americans and then non-Hispanic whites. For continuous ambulatory peritoneal dialysis the pattern is also similar, except that the 5-year survival for African-Americans is the same as that for MexicanTable 4. Cumulative Survival Rate for Each Ethnic Group at 1 and 5 Years, Stratified by Disease Etiology and Treatment Type" 1 Yr
Disease HTN GMN OM OTH All diagnoses Treatment CTR CAPO HOME TRANS All treatment
5Yr
NHW
MA
AA
Total
NHW
MA
AA
Total
75 88 74 78
82 88 76 83
87 91 82 84
82 88 77 80
31 50 25 42
44 60 24 49
50 61 35 51
43 55 26 46
78
81
86
38
40
49
76 81 91 91
79 85 68 96
85 93 95 90
32 52 60 81
35 62 41 88
47 62 59 83
78
81
86
38
40
49
79 84 90 92
37 55 59 83
Abbreviations: NHW, non-Hispanic whites; MA, Mexican-American; AA, African-American; HTN, hypertension; GMN, glomerulonephritis; OM, diabetes mellitus; OTH, other; CTR, center hemodialysis; CAPO, continuous ambulatory peritoneal dialysis; HOME, home hemodialysis; TRANS, transplant. * This proportion takes into account censored patients throughout the 5 years.
PUGH, TULEY, AND BASU
806
Table 5. Hazard Ratio for Death Compared With Non-Hispanic Whites Stratified by Disease Type, and Treatment Type, Age-Adjusted
Disease type All etiologies (n = 11,958) Diabetes (n = 3,260) Hypertension (n = 2,951) Glomerulonephritis (n = 1,840) Other (n = 3,907) Treatment type Center hemodialysis (n = 9,848) Home hemodialysis (n = 198) Transplantation (n = 1,168) CAPO (n = 730)
African-American HR(95%CI)
Mexican-American HR (95% CI)
0.71 (0.67,0.76)
0.93 (0.88, 0.99)
0.55 (0.48, 0.63)
0.77 (0.69, 0.86)
0.65 (0.58, 0.74)
0.82 (0.69, 0.96)
0.86 (0.70, 1.06) 0.85 (0.75, 0.95)
0.87 (0.72, 1.05) 0.90 (0.80, 1.01)
0.66 (0.61, 0.70)
0.89 (0.84, 0.95)
1.11 (0.53, 2.33)
2.33 (1.05, 5.15)
0.98 (0.65, 1.47) 0.80 (0.54, 1.19)
0.62 (0.39, 1.00) 0.81 (0.56, 1.16)
Abbreviations: HR, hazard ratio; CI, confidence interval; CAPO, continuous ambulatory peritoneal dialysis.
Americans. For home hemodialysis, AfricanAmericans maintain their survival advantage but Mexican-Americans have a much worse survival than either non-Hispanic whites or AfricanAmericans. For transplantation, MexicanAmericans have the best survival while AfricanAmericans are slightly worse at 1 year and slightly better at 5 years than non-Hispanic whites. These data are not age adjusted and there are known age differences in the ethnic groups, especially within the category of diabetic nephropathy. Since age and diabetic nephropathy are two of the strongest predictors of survival with ESRD, we performed age-adjusted analyses. Table 5 displays the hazard ratio for death compared with non-Hispanic whites stratified by disease or treatment type and age adjusted. African-Americans have a survival advantage overall, especially for hypertensive and diabetic disease. (The lower the hazard ratio, the better the survival in comparison to non-Hispanic whites. A hazard ratio greater than one would indicate a survival disadvantage.) Mexican-Americans have an overall advantage as well as better survival among both the hypertensive and diabetic categories. Both African-Americans and Mexican-Americans have a better survival than nonHispanic whites receiving hemodialysis in a center, which is the largest treatment group.
Mexican-Americans have somewhat worse survival on home hemodialysis. African-Americans and Mexican-Americans have survival similar to that of non-Hispanic whites for both transplantation and continuous ambulatory peritoneal dialysis, although Mexican-Americans just missed achieving an advantage with transplantation. To assess the relative contribution of ethnicity/ race as compared with traditional predictors of survival, such as age and disease type, we performed a multivariate analysis, which is displayed in Table 6. The hazard ratios displayed are in comparison to non-Hispanic whites with diabetic nephropathy, the group with the worst survival. Age and disease, as expected, were strong predictors of survival. Mexican-Americans were found to have a survival advantage. AfricanAmericans also have an advantage in the older age groups (African-American/age interaction). For the interaction between other diseases and race, African-Americans and Mexican-Americans have worse survival. Center size, transplantation, and continuous ambulatory peritoneal dialysis all showed a survival advantage. Table 6. Multivariate Model for Survival Comparing All Other Patients With Diabetic Non-Hispanic Whites (n = 11,556) Predictor Variable
HR
Probability Value
Age (1 O-yr interval) Mexican-American African-Americant Disease Other Glomerulonephritis Hypertension Center size Transplantation CAPO MA: age AA: Age MA: diseases, other AA: diseases, other
1.30' 0.47 0.90
0.000 0.000 0.501
0.57 0.53 0.66 0.93 0.37 0.68 1.07* 0.93' 1.19 1.35
0.000 0.000 0.000 0.010 0.000 0.000 0.012 0.000 0.021 0.000
NOTE. Required P = 0.10 to enter the model for the sample. Sex showed no univariate association and was not included in the model. Abbreviations: AA, African-American; MA, MexicanAmerican; HR, hazard ratio; CAPO, continuous ambulatory peritoneal dialySiS. , Hazard ratio calculated relative to each increasing age category. t African-Americans were kept in the model to preserve the structure of the variable race.
807
PROLONGED ESRD SURVIVAL FOR MINORITIES
DISCUSSION
The data reported here support a minority pattern of ESRD consisting of increased incidence of disease 2-4 and better survival on treatment. But what is its significance? First, for planning for the future, accurate projection of numbers of patients in need of renal replacement therapy is crucial for predicting the cost of the Medicare program. The cost of maintaining minorities on renal replacement therapy will continue to increase given the double impact of both increased incidence and increased survival. For Mexican-Americans, this will be compounded further by the rapid increase in the population itself. Second, the factors contributing to the increased survival of minorities need to be studied further. Can we learn from minorities how to improve survival for non-Hispanic whites? What is it about being African-American or MexicanAmerican that prolongs survival on renal replacement therapy? Eggers5 has pointed out that both increased voluntary withdrawal by nonHispanic whites, and possibly increased severity of illness among non-Hispanic whites, may contribute. The difference in voluntary withdrawal accounts for 23% of the difference in I-year survival rates between whites and African-Americans. The difference in severity of illness has yet to be documented, although African-Americans are transplanted less leaving healthier individuals receiving dialysis. One could also postulate other influences, such as social support, religion, extended family, or other cultural differences. Given the high mortality of the ESRD population in comparison to their age-matched peers, an effort at looking in greater depth at predictors of survival is warranted and has been supported by the National Institute of Diabetes, Digestive, and Kidney Diseases. The unexpectedly strong influence of minority status on survival could be used as a starting point for further study. Finally, prevention efforts for ESRD may need to be intensified in minority groups. Hypertensive and diabetic ESRD are predominant among minorities (not just African-American and MexicanAmerican, but also oriental and Native American). The renal consequences of both of these
diseases are potentially preventable. However, prevention is likely to require available, affordable long-term outpatient medical care. 17 Without universal health care coverage, these goals may never be achieved. REFERENCES I. US Renal Data System Annual Report: National Institutes of Health. Bethesda, MD, 1993 2. Eggers PW, Connerton R, McMullan M: The Medicare experience with end-stage renal disease: Trends in incidence, prevalence, and survival. Health Care Finane Rev 5:69-88, 1984 3. US Renal Data System Annual Report: National Institutes of Health. Bethesda, MD, 1990 4. Pugh lA, Stern MP, Haffner SM, Eifler CW, Zapata M: Excess incidence of treatment of end-stage renal disease in Mexican Americans. Am I Epidemiol 127:135-144, 1988 5. Eggers PW: Mortality rates among dialysis patients in Medicare's end-stage renal disease program. Am I Kidney Dis 15:414-421, 1990 6. Hutchinson TA, Thomas DC, MacGibbon B: Predicting survival in adults with end-stage renal disease: An age equivalence index. Ann Intern Med 96:417-423,1982 7. Held PI, Pauly MV, Diamond L: Survival analysis of patients undergoing dialysis. lAMA 257:645-663, 1987 8. Umen AJ, Le CT: Prognostic factors, models and related statistical problems in the survival of end-stage renal disease patients on hemodialysis. Stat Med 5:637-652, 1986 9. Port FK, Wolfe RA, Hawthorne VM, Ferguson CW: Discontinuation of dialysis therapy as a cause of death. Am I NephroI9:145-149, 1989 10. Shulman R, Price JDE, Spinelli J: Biopsychosocial aspects oflong-term survival on end-stage renal failure therapy. Psychol Med 19:945-954, 1989 II. Burton HJ, Kline SA, Lindsay RM, Heidenheim AP: The relationship of depression to survival in chronic renal failure. Psychosom Med 48:262-269, 1986 12. Held PJ, Brunner F, Odaka M, Garcia JR, Port FK, Gaylin DS: Five-year survival for end-stage renal disease patients in the United States, Europe, and Japan, 1982 to 1987. Am J Kidney Dis 15:451-457, 1990 13. Williams RB, Barefoot JC, Califf RM, Haney TL, Saunders WB, Pryor DB, Hlatky MA, Siegler Ie, Mark DB: Prognostic importance of social and economic resources among medically treated patients with angiographically documented coronary artery disease. JAMA 267:520-524, 1992 14. Associated Press wire report: Hispanics in US growing by one-third. San Antonio Light 232:1, 1988 (Sept 7) 15. Owen JR: Report paints a grim portrait of US Hispanics. San Antonio Light 19: I, 1992 (Feb 7) 16. BMDP: BMDP Statistical Software Manual, vol 2. Berkeley, CA, University of California Press, 1988 17. Pugh JA, Tuley MR, Stern MP, Hazuda H: The influence of outpatient insurance coverage on the microvascular complications of non-insulin-dependent diabetes in MexicanAmericans. J Diabet Complications 6:236-241, 1992
T
HE COST OF treatment of end-stage renal disease (ESRD) is enormous in both actual dollars (7.5 billion dollars per year!), and human suffering. The enrollment in the Medicare EndStage Renal Disease Program has grown steadily since its inception in 1973.2,3 Minorities are disproportionately represented among the individuals requiring renal replacement therapy.2,4 African-Americans have a threefold to fourfold higher incidence of treatment for ESRD than whites. 2,3 Fifty-six percent of new cases ofESRD in African-Americans are secondary to longstanding chronic diseases, such as hypertension and diabetes. 3 Other minorities, including Mexican-Americans,4 Native Americans,3 and the "all other" category in the Medicare data,2 also have higher rates ofESRD treatment than whites. Despite a higher incidence of ESRD, AfricanAmerican, Asian, and Native American dialysis patients also have been shown to have better 1year survival than whites. 5 African-Americans also have better 5-year survival rates. 2 The life expectancy for all ESRD patients is still significantly below the general population. 3 Numerous factors have been found to affect the survival of patients receiving renal replacement therapy, including age at onset oftreatment,5,6 type of treatment, 7 comorbid disease prior to treatment,6,8 etiology of renal disease,3,5,6 size of the center in
which treatment is given/ dialyzer reuse,7 voluntary withdrawal,9 depression,IO,!! and country in which the person is dialyzed,!2 The reasons African-Americans, Asians, and Native Americans have better survival are not intuitively obvious. Voluntary withdrawal from dialysis is lower among minorities,5,9 but the percentage of deaths explained by this mechanism is relatively small. African-Americans are less likely to undergo renal transplantation 2 and therefore healthier patients remain on dialysis. However, the survival advantage for African-Americans is From the Department ofMedicine, the Division of General Internal Medicine, University of Texas Health Science Center at San Antonio and the Audie L. Murphy Memorial Veterans Hospital, San Antonio, TX Received January 25, 1994; accepted in revised form February 15, 1994, Supported by grants from the RGK Foundation, National Institute of Diabetes and Digestive and Kidney Diseases (DK38392); the Geriatric Research, Education, and Clinical Center at the Audie L. Murphy Memorial Veterans Hospital; and the Mexican American Medical Treatment Ef fectiveness Center, a Minority Medical Treatment Effectiveness Program Center, Agency for Health Care Policy and Research (lU01HS0739701). Address reprint requests to Jacqueline A. Pugh, MD, Ambulatory Care llC6, Audie L. Murphy Memorial Veterans Hospital, 7400 Merton Minter Blvd, San Antonio, TX 78284. © 1994 by the National Kidney Foundation, Inc. 0272-6386/94/2306-0006$3.00/0
American Journal of Kidney Diseases, Vol 23, No 6 (June), 1994: pp 803-807
803
PUGH, TULEY, AND BASU
804
Table 1. Reason for Censorship From the Database Non-Hispanic Whites ('¥oj
Mexican-Americans ('¥oj
African-Americans ('¥oj
Death Relocation Withdrawal Lost to follow-up Recovered function Currently treated
2,672 (49.6) 131 (2.4) 6 (0.1) 672 (12.5) 94 (1.7) 1,814 (33.7)
1,529 (48.3) 32 (1.0) 7 (0.2) 310 (9.8) 31 (1) 1,254 (39.7)
1,407 (41.1) 36 (1.0) 0(0) 437 (12.8) 38 (1.1) 1,508 (44.0)
Total
5,389 (100)
3,163 (100)
3,426 (100)
even higher in the older age groups, in which transplantation is unlikely to be offered to either race. Lower income has been shown to predict higher coronary disease mortality l3 and, since the majority of ESRD patients die from coronary disease, the minority survival advantage is opposite of what would be predicted based on income. Another possible explanation is that sicker minority patients are not being offered renal replacement therapy. While there are no population-based data to refute this at present, the high rates of ESRD treatment among minorities and the continuing growth in the incidence of treatment make this an unlikely explanation for the minority survival advantage. Mexican-Americans, at 14 million, are the second largest minority group in the United States and their numbers are growing five times faster than the rest of the population. l4 • l5 Mexican-Americans are known to have a threefold higher incidence of ESRD compared with nonHispanic whites and a sixfold higher incidence of diabetic ESRD.4 Given that Mexican-Americans are similar to African-Americans with regard to increased incidence of ESRD, we posed the question, do Mexican-Americans receiving treatment for ESRD have a prolonged survival when compared with non-Hispanic whites? MATERIALS AND METHODS The Texas Kidney Health Program, a division of the Texas Department of Health, provided a database. The population included in this database has been described previously.' Briefly, all legal residents of Texas with ESRD may apply for enrollment in the Texas Kidney Health Program, which provides coverage for drugs and transportation as well as for expenses incurred prior to Medicare eligibility. Eighty-eight percent to 90% of all patients starting dialysis in Texas are registered with the Texas Kidney Health Program.' These analyses were restricted to patients enrolled between January I, 1975, and September 30, 1985. Patients younger than 20 years at the start of dialysis were excluded. The analyses included 11,978 subjects.
Total ('¥oj
5,608 (46.8) 199(1.7) 13 (0.1) 1,419 (11.9) 163(1.4) 4,576 (38.2) 11,978 (100)
Simple life-table analyses were performed using BMDP IL to test the proportionality of the independent variables. 16 The Cox proportional hazard model was used to age adjust the data in otherwise univariate analyses and to build a more comprehensive multivariate model using BMDP 2L. 16 Home hemodialysis patients were excluded from the multivariate analysis both because of small numbers and because their survival curve was not proportional to the other three treatments. Patients were censored at the time oflast contact with the Texas Kidney Health Program. Variables available for analysis included age at the start of dialysis, etiology of renal disease, race/ethnicity, sex, size of the dialysis treatment center, first treatment, current treatment, and one interim treatment if different. The above-listed variables as well as interaction variables between ethnicity/race and age, treatment, and disease etiology were tested. Current treatment was used unless the patient underwent transplantation. For those who had received a transplant, "transplantation" was assigned as their treatment and the date of the first transplant was taken as the date of entry. The structure of the database did not allow assignment of person-years on a given treatment because months or dates on each treatment were not included in the database. However, for those alive at the time the database was obtained, 80.4% were receiving treatment that was the same as their entry treatment. Not included were any measure of comorbid illness, dialyzer reuse, psychological assessment, or socioeconomic status.
RESULTS
Table 1 shows the percentage and number of cases in the database censored because of death, geographic relocation out of the state, voluntary withdrawal from dialysis, loss to follow-up, or recovery of renal function no longer requiring dialysis; it also included those patients currently receiving treatment. Overall, 11.9% of patients were lost to follow-up, and a lower percentage of Mexican-Americans were lost to follow-up than either non-Hispanic whites or African-Americans. The percentage of patients with voluntary withdrawal did not differ between groups, but was low in comparison to the Medicare data 5,9 and is not likely to have been accurately reported. Individuals who voluntarily withdrew from treatment were censored at the time of withdrawal
805
PROLONGED ESRD SURVIVAL FOR MINORITIES Table 2. Distribution of Patients by Age, Disease Etiology or Treatment, and Ethnic Group at the Start of End-Stage Renal Disease Treatment
Table 3. Cumulative Survival Rate for Each Ethnic Group at 1 and 5 Years, Stratified by Age" 1 Yr
Age group (yrs) 20-29 30-39 40-49 50-59 60-69 70-79 >80 All ages Oisease HTN GMN OM OTH All diseases Treatment CTR CAPO HOME TRANS All treatments
Non-Hispanic Whites
MexicanAmericans
AfricanAmericans
Total
555 699 791 1,121 1,284 789 147 5,386
232 301 447 910 808 398 66 3,162
323 453 595 870 760 356 65 3,422
1,110 1,453 1,833 2,901 2,852 1,543 278 11,970*
988 992 1,186 2,220 5,386
455 465 1,389 853 3,162
1,508 385 687 842 3,422
2,951 1,842 3,262 3,915 11,970*
4,106 462 163 649 5,380
2,734 153 11 264 3,162
3,021 115 24 261 3,421
9,861 730 198 1,174 11,963*
Abbreviations: HTN, hypertension; GMN, glomerulonephritis; OM, diabetes mellitus; OTH, other; CTR, center hemodialysis; CAPO, continuous ambulatory peritoneal dialysis; HOME, home hemodialysis; TRANS, transplant. * Each category had patients who had missing data for a variable: eight for age, eight for disease, and 15 for treatment.
because their final outcome was not available in the Texas Kidney Health Program database. Table 2 displays the distribution of patients at onset ofESRD treatment (time 0) by age, disease etiology or treatment, and ethnicity_ Table 3 displays age-specific 1- and 5-year survival for each ethnic group_ While MexicanAmericans display a small to moderate survival advantage over non-Hispanic whites in all but three age groups (I-year 70- to 79-year age group, 5-year 50- to 59-year age group, and 5-year 60to 69-year age group), African-Americans show no advantage in the 20- to 29-year age group and then progressively better survival in all other age groups. Table 4 shows the percentage of survival at 1 and 5 years for each ethnic group, stratified by disease etiology and treatment type_ Among disease etiologies, there is a consistent pattern of improved survival of African-Americans over non-Hispanic whites at both 1 and 5 years. Except for diabetic survival at 5 years, the survival of Mexican-Americans falls between that of AfricanAmericans and non-Hispanic whites, sometimes
Age (yr) 20-29 30-39 40-49 50-59 60-69 70-79 >80 All ages combined
5 Yr
NHW
MA
AA
Total
NHW
MA
AA
Total
91 85 84 81 73 66 61
96 91 87 83 75 65 63
90 89 90 86 82 81 69
92 88 87 83 76 69 63
73 52 47 42 26 16 7
83 69 54 39 23 18 17
61 68 61 46 38 33 16
72 60 53 42 28 21 10
78
81
86
38
40
49
Abbreviations: NHW, non-Hispanic whites; MA, Mexican-Americans; AA, African-Americans. * This proportion takes into account censored patients throughout the 5 years.
closer to non-Hispanic whites and sometimes closer to African-Americans. For treatment type, treatment received in a hemodialysis center presents the same pattern as among disease etiologies, with African-Americans having the best survival followed by Mexican-Americans and then non-Hispanic whites. For continuous ambulatory peritoneal dialysis the pattern is also similar, except that the 5-year survival for African-Americans is the same as that for MexicanTable 4. Cumulative Survival Rate for Each Ethnic Group at 1 and 5 Years, Stratified by Disease Etiology and Treatment Type" 1 Yr
Disease HTN GMN OM OTH All diagnoses Treatment CTR CAPO HOME TRANS All treatment
5Yr
NHW
MA
AA
Total
NHW
MA
AA
Total
75 88 74 78
82 88 76 83
87 91 82 84
82 88 77 80
31 50 25 42
44 60 24 49
50 61 35 51
43 55 26 46
78
81
86
38
40
49
76 81 91 91
79 85 68 96
85 93 95 90
32 52 60 81
35 62 41 88
47 62 59 83
78
81
86
38
40
49
79 84 90 92
37 55 59 83
Abbreviations: NHW, non-Hispanic whites; MA, Mexican-American; AA, African-American; HTN, hypertension; GMN, glomerulonephritis; OM, diabetes mellitus; OTH, other; CTR, center hemodialysis; CAPO, continuous ambulatory peritoneal dialysis; HOME, home hemodialysis; TRANS, transplant. * This proportion takes into account censored patients throughout the 5 years.
PUGH, TULEY, AND BASU
806
Table 5. Hazard Ratio for Death Compared With Non-Hispanic Whites Stratified by Disease Type, and Treatment Type, Age-Adjusted
Disease type All etiologies (n = 11,958) Diabetes (n = 3,260) Hypertension (n = 2,951) Glomerulonephritis (n = 1,840) Other (n = 3,907) Treatment type Center hemodialysis (n = 9,848) Home hemodialysis (n = 198) Transplantation (n = 1,168) CAPO (n = 730)
African-American HR(95%CI)
Mexican-American HR (95% CI)
0.71 (0.67,0.76)
0.93 (0.88, 0.99)
0.55 (0.48, 0.63)
0.77 (0.69, 0.86)
0.65 (0.58, 0.74)
0.82 (0.69, 0.96)
0.86 (0.70, 1.06) 0.85 (0.75, 0.95)
0.87 (0.72, 1.05) 0.90 (0.80, 1.01)
0.66 (0.61, 0.70)
0.89 (0.84, 0.95)
1.11 (0.53, 2.33)
2.33 (1.05, 5.15)
0.98 (0.65, 1.47) 0.80 (0.54, 1.19)
0.62 (0.39, 1.00) 0.81 (0.56, 1.16)
Abbreviations: HR, hazard ratio; CI, confidence interval; CAPO, continuous ambulatory peritoneal dialysis.
Americans. For home hemodialysis, AfricanAmericans maintain their survival advantage but Mexican-Americans have a much worse survival than either non-Hispanic whites or AfricanAmericans. For transplantation, MexicanAmericans have the best survival while AfricanAmericans are slightly worse at 1 year and slightly better at 5 years than non-Hispanic whites. These data are not age adjusted and there are known age differences in the ethnic groups, especially within the category of diabetic nephropathy. Since age and diabetic nephropathy are two of the strongest predictors of survival with ESRD, we performed age-adjusted analyses. Table 5 displays the hazard ratio for death compared with non-Hispanic whites stratified by disease or treatment type and age adjusted. African-Americans have a survival advantage overall, especially for hypertensive and diabetic disease. (The lower the hazard ratio, the better the survival in comparison to non-Hispanic whites. A hazard ratio greater than one would indicate a survival disadvantage.) Mexican-Americans have an overall advantage as well as better survival among both the hypertensive and diabetic categories. Both African-Americans and Mexican-Americans have a better survival than nonHispanic whites receiving hemodialysis in a center, which is the largest treatment group.
Mexican-Americans have somewhat worse survival on home hemodialysis. African-Americans and Mexican-Americans have survival similar to that of non-Hispanic whites for both transplantation and continuous ambulatory peritoneal dialysis, although Mexican-Americans just missed achieving an advantage with transplantation. To assess the relative contribution of ethnicity/ race as compared with traditional predictors of survival, such as age and disease type, we performed a multivariate analysis, which is displayed in Table 6. The hazard ratios displayed are in comparison to non-Hispanic whites with diabetic nephropathy, the group with the worst survival. Age and disease, as expected, were strong predictors of survival. Mexican-Americans were found to have a survival advantage. AfricanAmericans also have an advantage in the older age groups (African-American/age interaction). For the interaction between other diseases and race, African-Americans and Mexican-Americans have worse survival. Center size, transplantation, and continuous ambulatory peritoneal dialysis all showed a survival advantage. Table 6. Multivariate Model for Survival Comparing All Other Patients With Diabetic Non-Hispanic Whites (n = 11,556) Predictor Variable
HR
Probability Value
Age (1 O-yr interval) Mexican-American African-Americant Disease Other Glomerulonephritis Hypertension Center size Transplantation CAPO MA: age AA: Age MA: diseases, other AA: diseases, other
1.30' 0.47 0.90
0.000 0.000 0.501
0.57 0.53 0.66 0.93 0.37 0.68 1.07* 0.93' 1.19 1.35
0.000 0.000 0.000 0.010 0.000 0.000 0.012 0.000 0.021 0.000
NOTE. Required P = 0.10 to enter the model for the sample. Sex showed no univariate association and was not included in the model. Abbreviations: AA, African-American; MA, MexicanAmerican; HR, hazard ratio; CAPO, continuous ambulatory peritoneal dialySiS. , Hazard ratio calculated relative to each increasing age category. t African-Americans were kept in the model to preserve the structure of the variable race.
807
PROLONGED ESRD SURVIVAL FOR MINORITIES
DISCUSSION
The data reported here support a minority pattern of ESRD consisting of increased incidence of disease 2-4 and better survival on treatment. But what is its significance? First, for planning for the future, accurate projection of numbers of patients in need of renal replacement therapy is crucial for predicting the cost of the Medicare program. The cost of maintaining minorities on renal replacement therapy will continue to increase given the double impact of both increased incidence and increased survival. For Mexican-Americans, this will be compounded further by the rapid increase in the population itself. Second, the factors contributing to the increased survival of minorities need to be studied further. Can we learn from minorities how to improve survival for non-Hispanic whites? What is it about being African-American or MexicanAmerican that prolongs survival on renal replacement therapy? Eggers5 has pointed out that both increased voluntary withdrawal by nonHispanic whites, and possibly increased severity of illness among non-Hispanic whites, may contribute. The difference in voluntary withdrawal accounts for 23% of the difference in I-year survival rates between whites and African-Americans. The difference in severity of illness has yet to be documented, although African-Americans are transplanted less leaving healthier individuals receiving dialysis. One could also postulate other influences, such as social support, religion, extended family, or other cultural differences. Given the high mortality of the ESRD population in comparison to their age-matched peers, an effort at looking in greater depth at predictors of survival is warranted and has been supported by the National Institute of Diabetes, Digestive, and Kidney Diseases. The unexpectedly strong influence of minority status on survival could be used as a starting point for further study. Finally, prevention efforts for ESRD may need to be intensified in minority groups. Hypertensive and diabetic ESRD are predominant among minorities (not just African-American and MexicanAmerican, but also oriental and Native American). The renal consequences of both of these
diseases are potentially preventable. However, prevention is likely to require available, affordable long-term outpatient medical care. 17 Without universal health care coverage, these goals may never be achieved. REFERENCES I. US Renal Data System Annual Report: National Institutes of Health. Bethesda, MD, 1993 2. Eggers PW, Connerton R, McMullan M: The Medicare experience with end-stage renal disease: Trends in incidence, prevalence, and survival. Health Care Finane Rev 5:69-88, 1984 3. US Renal Data System Annual Report: National Institutes of Health. Bethesda, MD, 1990 4. Pugh lA, Stern MP, Haffner SM, Eifler CW, Zapata M: Excess incidence of treatment of end-stage renal disease in Mexican Americans. Am I Epidemiol 127:135-144, 1988 5. Eggers PW: Mortality rates among dialysis patients in Medicare's end-stage renal disease program. Am I Kidney Dis 15:414-421, 1990 6. Hutchinson TA, Thomas DC, MacGibbon B: Predicting survival in adults with end-stage renal disease: An age equivalence index. Ann Intern Med 96:417-423,1982 7. Held PI, Pauly MV, Diamond L: Survival analysis of patients undergoing dialysis. lAMA 257:645-663, 1987 8. Umen AJ, Le CT: Prognostic factors, models and related statistical problems in the survival of end-stage renal disease patients on hemodialysis. Stat Med 5:637-652, 1986 9. Port FK, Wolfe RA, Hawthorne VM, Ferguson CW: Discontinuation of dialysis therapy as a cause of death. Am I NephroI9:145-149, 1989 10. Shulman R, Price JDE, Spinelli J: Biopsychosocial aspects oflong-term survival on end-stage renal failure therapy. Psychol Med 19:945-954, 1989 II. Burton HJ, Kline SA, Lindsay RM, Heidenheim AP: The relationship of depression to survival in chronic renal failure. Psychosom Med 48:262-269, 1986 12. Held PJ, Brunner F, Odaka M, Garcia JR, Port FK, Gaylin DS: Five-year survival for end-stage renal disease patients in the United States, Europe, and Japan, 1982 to 1987. Am J Kidney Dis 15:451-457, 1990 13. Williams RB, Barefoot JC, Califf RM, Haney TL, Saunders WB, Pryor DB, Hlatky MA, Siegler Ie, Mark DB: Prognostic importance of social and economic resources among medically treated patients with angiographically documented coronary artery disease. JAMA 267:520-524, 1992 14. Associated Press wire report: Hispanics in US growing by one-third. San Antonio Light 232:1, 1988 (Sept 7) 15. Owen JR: Report paints a grim portrait of US Hispanics. San Antonio Light 19: I, 1992 (Feb 7) 16. BMDP: BMDP Statistical Software Manual, vol 2. Berkeley, CA, University of California Press, 1988 17. Pugh JA, Tuley MR, Stern MP, Hazuda H: The influence of outpatient insurance coverage on the microvascular complications of non-insulin-dependent diabetes in MexicanAmericans. J Diabet Complications 6:236-241, 1992