Givens JL, Tjia J, Zhou C, Emanuel E, Ash AS. Racial and Ethnic Differences in Hospice Use Among Patients With Heart Failure. Arch Intern Med. 2010;170(5):427-432. doi:10.1001/archinternmed.2009.547
Copyright 2010 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2010
Heart failure is the leading noncancer diagnosis for patients in hospice care and the leading cause of hospitalization among Medicare beneficiaries. Racial and ethnic differences in hospice patients are well documented for patients with cancer but poorly described for those with heart failure.
On the basis of a national sample of 98 258 Medicare beneficiaries 66 years and older on January 1, 2001, with a diagnosis of heart failure who had at least 1 physician or hospital encounter and who were not enrolled in hospice care between January 1 and December 31, 2000, we determined the effect of race and ethnicity on hospice entry for patients with heart failure in 2001 after adjusting for sociodemographic, clinical, and geographic factors.
In unadjusted analysis, blacks (odds ratio [OR], 0.52) and Hispanics (0.43) used hospice care for heart failure less than whites. Racial and ethnic differences in patients who received hospice care for heart failure persisted after adjusting for markers of income, urbanicity, severity of illness, local density of hospice use, and medical comorbidity (adjusted OR for blacks, 0.59; 95% confidence interval, 0.47-0.73; and adjusted OR for Hispanics, 0.49; 95% confidence interval, 0.37-0.66; compared with whites). Advanced age, greater comorbidity, emergency department visits, hospitalizations, and greater local density of hospice use were also associated with hospice use.
In a national sample of Medicare beneficiaries with heart failure, blacks and Hispanics used hospice care for heart failure less than whites after adjustment for individual and market factors. To understand the mechanisms underlying these findings, further examination of patient preferences and physician referral behavior is needed.
Hospice care is designed to provide comfort and emotional support to patients with terminal illness and their families and is most commonly (84%) provided through the Medicare hospice benefit.1 Although patients dying of cancer comprised 44% of hospice admissions in 2006,1 hospice care can also offer substantial benefit to patients with other terminal illnesses, such as end-stage heart failure (HF). However, patients in the terminal stages of noncancer conditions use hospice care less frequently than those with advanced malignant tumors.1
Currently, HF affects nearly 5 million people in the United States and is the leading cause of hospitalization among Medicare beneficiaries.2 Despite progress in treatment,3,4 patients with advanced HF have a 1-year mortality of 50% to 70%,5 and hospice care is increasingly recommended in guidelines for such patients.4,6- 8 Although advanced heart disease represents the second most common hospice diagnosis, comprising 11.8% of hospice enrollees,1 hospice services are generally recognized as underused by patients with HF.4,9
Underuse of hospice care is well documented, especially among racial and ethnic minorities.10- 14 Racial and ethnic differences in patients who use hospice care have been found across a spectrum of patients with cancer diagnoses15- 17 and may be more pronounced in patients with noncancer diagnoses.18 However, previous studies10,11,13- 15,18 of ethnic differences in patients who use hospice care have been limited to cross-sectional or retrospective analyses, which have had limited ability to assess life expectancy or severity of illness among potential hospice care recipients. In this study, we use data from a large, nationally representative, ethnically diverse cohort of Medicare beneficiaries with HF to estimate the independent effect of race and ethnicity on hospice care for HF in the coming year, after adjusting for patient demographics, in-hospital interventions, comorbidity, and geographic variation in hospice use.
We used a national sample of Medicare beneficiaries oversampled for nonwhite patients as previously described.19 The data source included merged Medicare claims files (denominator, inpatient, carrier, outpatient, and hospice files) from 2000 and 2001. An initial sample of 1 million beneficiaries 66 years and older was selected from the Medicare denominator file for 2001, with deliberate oversampling of beneficiaries who died in 2001 and racial and ethnic minorities categorized as black, Hispanic, and other. These files were merged with the National Death Index from 2001 to verify date of death.
We included Medicare beneficiaries who had complete claims data from 2000 and 2001, resided in the United States, were not enrolled in Medicare managed care organizations, and were not entitled to the Medicare end-stage renal disease benefit. Among those meeting these initial inclusion criteria (n = 603 128), we limited the study sample to those beneficiaries (n = 98 258) with a diagnosis of HF (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM]20 codes 428.xx, 398.91, 402.11, 402.91, 404.11, 404.13, 404.91, or 404.93) who had at least 1 physician or hospital encounter and who were not enrolled in hospice care between January 1 and December 31, 2000. On the basis of the sampling, these individuals reflect the experience of approximately 2.7 million beneficiaries.
Our primary outcome variable was entry into a hospice with an admitting diagnosis of HF between January 1 and December 31, 2001. We identified hospice entry from the first hospice admission date in 2001 and calculated hospice duration from first admission date until death or December 31, 2001, for nondecedents.
We used the Medicare denominator file categories of white, black, and Hispanic. We also collapsed all remaining categories (including Asian, North American natives, and unknown) into “other.”
To characterize the morbidity of participants, we calculated prospective diagnostic cost group (DCG) scores derived from outpatient, inpatient, and carrier claims for the year 2000 (DxCG version 6.1 for SAS Windows; SAS Institute Inc, Cary, North Carolina). The DCG score predicts Medicare costs for the period designated as next year, as calculated from 1 year of ICD-9-CM diagnosis codes, age, and sex, and is expressed as a relative risk.21 Thus, a score of 1.0 indicates an expected level of future health care use equal to the mean for Medicare beneficiaries, and 2.0 indicates expected costs that are twice as high. This score also predicts other outcomes, including mortality.22
To characterize each beneficiary's HF severity, we used the inpatient file from 2000 to calculate the number of emergency department (ED) visits and hospitalizations and the number of days spent in an intensive care unit (ICU) or coronary care unit (CCU). We used the denominator file to capture beneficiary age and sex and to define a geography-based socioeconomic status indicator (median income of zip code of residence) as determined from 2000 US Census data. We also used a marker for a state's Medicaid purchase (ie, Medicaid buy-in) of the Part B benefit as an indicator of low individual income.
We used our software's condition categories to identify other common morbidities present in 2000 that could lead to hospice care in 2001. These categories were cancer, dementia, stroke, and chronic obstructive pulmonary disease.
Because hospice availability23 and use4,10,24 differ by place of residence, we created 2 geographic variables from beneficiary zip code of residence. The first measures urbanicity because persons in rural settings typically have less access to hospice services.10,23 We used the Beale rural-urban continuum codes25 to categorize urbanicity as follows: metropolitan region with a population of 1 million or more, metropolitan region with a population of less than 1 million, nonmetropolitan region, and unknown.
We also developed a novel, hospice-specific, health service area (HSA)–based variable that we call local hospice density, derived from the larger data set of HF and non-HF beneficiaries who met our initial eligibility criteria (n = 603 128). This variable describes the local prevalence of hospice use among the 2001 decedents in this data set (n = 158 903). The HSAs are either single counties or clusters of counties that are relatively self-contained with respect to hospital care and delineate local health care markets for community-based primary inpatient care. We located each decedent in the HSA by his or her zip code of residence. We then defined the local density of hospice use as the percentage of the HSA's decedents who had entered into hospice care. For HSAs with fewer than 50 decedents in our data, we substituted the density of hospice use for the hospital referral region that contains it. Hospital referral regions consist of 1 or more HSAs and represent the tertiary market for medical care, including referral or specialty care.26
We used descriptive statistics to characterize the demographics, health characteristics, health care use (including hospice entry), and mortality of the study population by racial or ethnic category. All analyses were conducted using sampling weights to obtain population-based estimates and a linearized variance estimator based on a first-order Taylor series linear approximation to compute standard errors.27 The sample weights represent the reciprocal of the sampling probabilities for each of the 8 strata represented in the total sample of 1 million beneficiaries (2 based on decedent status and 4 based on race and ethnicity). To assess the statistical significance of bivariate associations, we used an adjusted Wald test for continuous variables, a design-based Pearson χ2 test for categorical variables, and a nonparametric equality-of-medians test for medians.
To identify the independent association of race and ethnicity on hospice entry for HF, we developed a series of weighted logistic regression models with the outcome of hospice entry for HF in 2001. Models successively added covariate sets: race and ethnicity alone (model 1); age and sex (model 2); urbanicity, income, hospitalizations, ED visits, number of ICU and CCU days, diagnoses of cancer, chronic obstructive pulmonary disease, dementia, or stroke, and DCG score (model 3); and local hospice density (model 4). To examine whether our findings were attributable to differences in local health care delivery systems not captured in our data, we also developed a fixed-effects regression model that compares members of different racial and ethnic groups only when they reside in the same HSA. Because the findings of this sensitivity analysis are similar to the more conceptually revealing original analysis described herein, we present the results of the original analysis.
Analyses were conducted using STATA statistical software, version 10.0 (StataCorp, College Station, Texas), and SAS statistical software, version 9.0 (SAS Institute Inc). This study was approved by the institutional review board of the Boston University School of Medicine.
The final study population included 98 258 Medicare beneficiaries (weighted population, n=2.7 million). The weighted percentages by race and ethnicity were as follows: white, 88.3%; black, 8.5%; Hispanic, 1.4%; and other, 1.8%. Many differences in sociodemographic characteristics, DCG scores, and severity of illness measures by race and ethnicity were statistically significant (Table 1). Beneficiaries who were of black, Hispanic, and other races and ethnicities were more likely to live in large metropolitan urban areas. Black and Hispanic beneficiaries had lower incomes and were more likely to have Medicaid buy-in. There were also notable age differences among the groups, with black and Hispanic beneficiaries being younger than whites and others. Differences in markers of health and health care use were notable, with Hispanics having the highest mean DCG score, number of hospitalizations, length of stays in the CCU and ICU, and number of ED visits. Unadjusted use of hospice care for HF and for any reason differed by race or ethnicity (Table 2). The percentage of beneficiaries using hospice care for any diagnosis in 2001 was small (3.9%). Of these, 18.2% entered because of HF. The percentage of decedents who had used hospice care was 19.9% overall, with a higher percentage of white decedents (20.4%) using hospice care compared with blacks (15.4%), Hispanics (16.9%), or those of other races or ethnicities (16.3%). Among those who used hospice care, a higher percentage of whites entered hospice care for HF treatment (18.5%) than blacks (14.1%), Hispanics (13.2%), and other racial and ethnic groups (15.8%). For HF, Hispanics had the longest median duration of hospice care (19 days), followed by blacks (14 days), whites (13 days), and other racial and ethnic groups (10 days).
Nonwhite patients used hospice care for HF less than whites (Table 3). After adjusting for sociodemographics, urbanicity, comorbidities, DCG score, use of medical services, and local hospice density (model 4), hospice use remained lowest for Hispanic beneficiaries (adjusted odds ratio [aOR], 0.49) compared with whites. Blacks (aOR, 0.59) and other nonwhite beneficiaries (aOR, 0.64) with HF were also less likely to use hospice care. Sex, income, Medicaid buy-in status, more days in the ICU or CCU, greater number of ED visits, and geographic residence in an urban area were not significantly associated with hospice entry. Higher DCG score, advanced age, ED visits and hospitalizations, and greater local hospice density were associated with hospice use. Competing hospice diagnoses of chronic obstructive pulmonary disease, dementia, and stroke were not associated with hospice use, but a diagnosis of cancer was protective against entry into hospice care for HF (aOR, 0.72).
This analysis of a national cohort of Medicare beneficiaries finds marked racial and ethnic differences among patients with HF who received hospice care. In our sample, blacks had 41.0% lower odds of hospice use and Hispanics 51.0% lower odds of receiving hospice care than whites. To our knowledge, our study is the largest longitudinal cohort study examining factors associated with hospice entry for a noncancer diagnosis. In addition, we use a prospective design and a novel measure of local hospice use.
Our finding that racial and ethnic minorities are less likely to use hospice care for HF is consistent with other hospice literature. Several studies have documented lower rates of hospice use among blacks, Hispanics, and Asian Americans compared with whites. However, these studies either have specifically focused on patients with cancer14- 16 or do not define the diagnosis determining hospice entry.12,13,28 There is concern that racial and ethnic disparities may be more pronounced among patients with noncancer diagnoses than among those with cancer.18 Our findings document significant racial and ethnic differences in hospice use and counter speculation that overall increases in the availability of hospice services in the 1990s may have erased racial and ethnic differences in hospice use.11
In this study, we have adjusted for numerous factors related to hospice entry, many of which vary substantially by race and ethnicity. Geographic variation in hospice availability is associated with patterns of hospice use,10 including hospice entry for HF.4 Hospice availability differs between urban and rural areas,10,23,24 and lower hospice use has been documented among patients living in predominantly minority vs white areas.29 On a state level, greater regional availability of hospice services is associated with less racial and ethnic disparity between blacks and whites in hospice use.28 We created a variable to represent the prevalence of hospice use among Medicare decedents within smaller local geographic units (HSAs) and found that our observed ethnic differences in hospice use for HF persisted after accounting for local hospice availability and urbanicity.
In addition to sociodemographic, clinical, and geographic characteristics, cultural beliefs and values may contribute to differences between blacks and whites in end-of-life care and hospice use.30,31 For example, compared with whites, blacks are less likely to complete advance directives, have less favorable beliefs about hospice care, opt for more aggressive treatments, and are more likely to have spiritual beliefs that conflict with the goals of palliative treatment.19,30,32 In addition, lack of trust between patients and physicians may be more pronounced for ethnic minorities33,34 and may contribute to ethnic differences in hospice entry.35,36 Blacks more often report receiving inadequate information regarding end-of-life care37 and are less likely to be informed about hospice services than whites.38 Our administrative data contained information on neither patient cultural beliefs and values nor physician behavior, factors that may also help explain differences in hospice use. Finally, health literacy, also not measured in this study, has been found to partially explain racial and ethnic differences in end-of-life treatment preferences.39
This study has several limitations. First, we could not measure HF severity using clinical markers such as ejection fraction or exercise tolerance, and thus, HF severity might vary by race and ethnicity. However, we adjusted for prior-year measures of ED visits, hospitalizations, and ICU and CCU use as proxies for clinical complications and disease severity. Second, although race and ethnicity coding in the Medicare denominator file has a good positive predictive value for identifying whites, blacks, and Hispanics (>94%), it performs less well for Asians and American Indian/Alaskan Native beneficiaries,40 making findings for the “other” race and ethnicity category difficult to interpret. Finally, although higher hospice use has been reported among enrollees from health maintenance organizations compared with patients with traditional Medicare coverage,41 we could not include such patients in our study sample because of incomplete capture of diagnoses and health care use.
This article describes the largest noncancer study of hospice entry to date. It prospectively examines a national probability sample of people with HF, oversampled for racial and ethnic minorities and decedents. Large differences in racial and ethnic minority use of hospice care for HF compared with whites remain largely unchanged after adjusting for differences in income, urbanicity, comorbidity, severity of illness, and hospice use density. It is not clear how many of these differences reflect access issues as opposed to considered patient preferences.
Correspondence: Jane L. Givens, MD, MSCE, Hebrew SeniorLife Institute for Aging Research, Beth Israel Deaconess Medical Center, 1200 Centre St, Boston, MA 02131 (firstname.lastname@example.org).
Accepted for Publication: September 25, 2009.
Author Contributions:Study concept and design: Givens, Tjia, and Ash. Acquisition of data: Emanuel and Ash. Analysis and interpretation of data: Givens, Tjia, Zhou, and Ash. Drafting of manuscript: Givens and Tjia. Critical revision of manuscript for important intellectual content: Givens, Tjia, Zhou, Emanuel, and Ash. Statistical analysis: Givens, Tjia, Zhou, and Ash. Obtained funding: Emanuel. Study supervision: Givens and Ash.
Financial Disclosure: Dr Ash is cofounder of and consulting senior scientist at DxCG Inc (Boston, Massachusetts), a company whose software was used to measure morbidity burden in this study. Otherwise, there are no conflicts of interest or financial disclosures.
Funding/Support: This study was supported by contracts with the Department of Bioethics of the Clinical Center of the National Institutes of Health. Dr Tjia is supported by National Institute on Aging Career Development Award K08AG021527.
Role of the Sponsors: The funding sources had no role in the design and conduct of the study; in the collection, management, analysis, and interpretation of the data; or in the preparation of the manuscript.
Drs Emanuel and Ash arranged for use of Medicare and Medicaid Services data.