QUARTERLY FOCUS ISSUE: HEART FAILURE: CLINICAL RESEARCH
Geographic Disparities in Heart Failure Hospitalization Rates Among Medicare Beneficiaries
Michele Casper, PhD*,*,
Isaac Nwaise, MA*,
Janet B. Croft, PhD ,
Yuling Hong, MD, PhD*,
Jing Fang, MD* and
Sophia Greer, MPH*
* Division for Heart Disease and Stroke Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia
Division for Adult and Community Health, Centers for Disease Control and Prevention, Atlanta, Georgia
Manuscript received August 18, 2009;
revised manuscript received October 7, 2009,
accepted October 12, 2009.
* Reprint requests and correspondence: Dr. Michele Casper, Division for Heart Disease and Stroke Prevention, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, 4770 Buford Highway NE, Mailstop K-47, Atlanta, Georgia 30341 (Email: mcasper{at}cdc.gov).
 |
Abstract
|
|---|
Objectives: This study was designed to document local-level geographic disparities in heart failure (HF) hospitalization rates among Medicare beneficiaries.
Background: Although the burden of HF is well documented at the national level, little is known about the geographic disparities in HF.
Methods: The study population consisted of fee-for-service Medicare beneficiaries  65 years of age who resided in the U.S., Puerto Rico, or the U.S. Virgin Islands during the years 2000 to 2006. Using hospital claims data for Medicare beneficiaries, we calculated spatially smoothed and age-adjusted average annual county-level HF hospitalization rates per 1,000 Medicare beneficiaries for the total population and by racial/ethnic group (blacks, Hispanics, and whites) for the years 2000 to 2006. A HF hospitalization was defined as a short-stay hospital claim with a principal (first-listed) discharge diagnosis of HF using the International Classification of Diseases-9th Revision-Clinical Modification code 428.
Results: The average annual age-adjusted HF hospitalization rate per 1,000 Medicare beneficiaries was 21.5 per 1,000, and ranged from 7 to 61 per 1,000 among counties in the U.S. For the total study population, a clear East-West gradient was evident, with the highest rates located primarily along the lower Mississippi River Valley and the Ohio River Valley, including the Appalachian region. Similar patterns were observed for blacks and whites, although the pattern for Hispanics differed.
Conclusions: The evidence of substantial geographic disparities in HF hospitalizations among Medicare beneficiaries is important information for health professionals to incorporate as they design prevention and treatment policies and programs tailored to the needs of their communities.
Key Words: geographic disparities • heart failure • hospitalizations • Medicare beneficiaries
|
Abbreviations and Acronyms
| | AMI = acute myocardial infarction | | HF = heart failure |
|
The burden of heart failure (HF) has been well documented at the national level (1–4). The number of prevalent HF cases is approximately 5,700,000 and continues to increase, in part because of improved survival and better care of patients with acute myocardial infarction (AMI) and increased prevalence of hypertension and diabetes mellitus, as well as the aging of the U.S. population (1,3,5,6). It is estimated that the lifetime risk of developing HF is 20% among persons age 40 years or older (7). More than 1.1 million people were hospitalized with HF as the principal diagnosis in 2006 and >290,000 people died from HF-related diseases in 2004 (3). The economic cost of HF, including direct medical costs and indirect costs due to loss of productivity, is >$37 billion (3). The prevalence rate of HF increases with age and is higher among non-Hispanic blacks than among non-Hispanic whites and Mexican Americans, based on estimates from the National Health and Nutrition Examination Survey (3).
However, very little is known about geographic disparities in HF at the local level. One recent study examined regional differences in HF hospitalization rates among the 4 main regions of the U.S. (Northeast, Midwest, South, and West) (8), and another recent study published HF readmission rates among Medicare beneficiaries using hospital referral regions as the unit of analysis (9). However, we believe there have been no published studies of county-level HF hospitalization rates. In this brief report, we document the geographic disparities in the burden of HF among fee-for-service Medicare beneficiaries, by race-ethnicity, at the county level for the aggregated years 2000 to 2006.
|
Methods
|
|---|
The study population consisted of fee-for-service Medicare beneficiaries ages 65 years who resided in the U.S., Puerto Rico, or the U.S. Virgin Islands during the years 2000 to 2006. Beneficiaries were excluded if they were members of a health maintenance organization, died before July 1, or were younger than 65 years of age on July 1 of each of the study years.
We obtained hospital claims data included in the Medicare Provider Analysis and Review, Part A, for our study population from the Centers for Medicare and Medicaid Services. We defined a HF hospitalization as a short-stay hospital claim with a principal (first-listed) discharge diagnosis of HF using the International Classification of Diseases-9th Revision-Clinical Modification (10) code 428.
We calculated spatially smoothed and age-adjusted average annual county-level HF hospitalization rates per 1,000 Medicare beneficiaries for the total population and by racial/ethnic group (blacks, Hispanics, and whites) for the years 2000 to 2006. To produce more stable rates in counties with small populations, spatial smoothing was performed using a spatial moving average and the contiguity matrix from the Area Resource File. Hospitalization rates were directly age-standardized to the 2000 U.S. standard population, age 65 years. For the total population and each racial/ethnic group, county-level HF hospitalization rates were categorized into quintiles, and maps were produced using Environmental Systems Research Institute (ESRI) software (Redlands, California) to document the geographic distribution of HF hospitalization rates for the total population, blacks, Hispanics, and whites.
We identified the race/ethnicity of each beneficiary on the basis of the race code on the claim record for a patient's hospital stay. Because race and Hispanic ethnicity were not reported separately in the Medicare databases, the categories of black, Hispanic, and white are mutually exclusive (11). This reporting practice can result in misclassification of race, and there are data that suggest that Hispanics are under-reported in the Medicare datasets (11). We recognize that racial/ethnic categories are socially constructed and are not biologically based (12,13).
|
Results
|
|---|
The average annual age-adjusted hospitalization rate (per 1,000 Medicare beneficiaries) for HF in the study population was 21.5. The rate for blacks (32.9) was substantially higher than the rates for Hispanics (25.0) and for whites (20.4). Among counties in the U.S., the rate ranged from 7 to 61 per 1,000 (Fig. 1). A clear East-West gradient is evident in the geographic distribution of HF hospitalization rates. The map of HF hospitalization rates for the total population indicates that counties with the highest rates (top quintile) were located primarily along the lower Mississippi River Valley and the Ohio River Valley, including the Appalachian region. High-rate counties were also found in the southern and north-central regions of Texas and the southeastern corner of Oklahoma. Because whites are the majority race group in the Medicare-aged population, a similar pattern was observed for whites (Fig. 2). For blacks, high-rate counties were also observed along the Mississippi River Valley and in parts of the Appalachian region—particularly western Pennsylvania and southern West Virginia (Fig. 3). Northern Illinois also had a concentration of high-rate counties. Among Hispanics, the pattern was different, with high-rate counties located primarily in either the Northeast region or Texas (Fig. 4).
View larger version (65K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1 HF Hospitalization Rates Among Medicare Beneficiaries, Age 65 Years, 2000–2006: Total Population
The map depicts age-adjusted and spatially smoothed county-specific heart failure (HF) hospitalization rates per 1,000 among fee-for-service Medicare beneficiaries, age 65 years and older, for the aggregated years 2000 to 2006. HF hospitalization was defined as principal discharge diagnosis of HF using the International Classification of Diseases-9th Revision-Clinical Modification code 428. Counties labeled as "insufficient data" (n = 6) did not have large enough populations of Medicare beneficiaries to produce stable rates of HF hospitalization. A clear East-West gradient was evident, with the highest rates located primarily along the lower Mississippi River Valley and the Ohio River Valley, including the Appalachian region. *HF hospitalization rates are spatially smoothed to enhance the stability of rates in countries with small populations. Source: Medicare Provider Analysis and Review (MEDPAR).
|
|
View larger version (65K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2 HF Hospitalization Rates Among Medicare Beneficiaries, Age 65 Years, 2000–2006: Whites
The map depicts age-adjusted and spatially smoothed county-specific heart failure (HF) hospitalization rates per 1,000 among white fee-for-service Medicare beneficiaries, age 65 years and older, for the aggregated years 2000 to 2006. HF hospitalization was defined as principal discharge diagnosis of HF using the International Classification of Diseases-9th Revision-Clinical Modification code 428. Counties labeled as "insufficient data" (n = 10) did not have large enough populations of white Medicare beneficiaries to produce stable rates. A clear East-West gradient was evident, with the highest rates located primarily along the lower Mississippi River Valley and the Ohio River Valley, including the Appalachian region. *HF hospitalization rates are spatially smoothed to enhance the stability of rates in countries with small populations. Source: Medicare Provider Analysis and Review (MEDPAR).
|
|
View larger version (63K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3 HF Hospitalization Rates Among Medicare Beneficiaries, Age 65 Years, 2000–2006: Blacks
The map depicts age-adjusted and spatially smoothed county-specific heart failure (HF) hospitalization rates per 1,000 among black fee-for-service Medicare beneficiaries, age 65 years and older, for the aggregated years 2000 to 2006. HF hospitalization was defined as principal discharge diagnosis of HF using the International Classification of Diseases-9th Revision-Clinical Modification code 428. Counties labeled as "insufficient data" (n = 1,026) did not have large enough populations of black Medicare beneficiaries to produce stable rates. Counties with the highest rates were located primarily along the lower Mississippi River Valley, the mid-Appalachian region, and parts of Illinois. *HF hospitalization rates are spatially smoothed to enhance the stability of rates in countries with small populations. Source: Medicare Provider Analysis and Review (MEDPAR).
|
|
View larger version (63K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4 HF Hospitalization Rates Among Medicare Beneficiaries, Age 65 Years, 2000–2006: Hispanics
The map depicts age-adjusted and spatially smoothed county-specific heart failure (HF) hospitalization rates per 1,000 among Hispanic fee-for-service Medicare beneficiaries, age 65 years and older, for the aggregated years 2000 to 2006. HF hospitalization was defined as principal discharge diagnosis of HF using the International Classification of Diseases-9th Revision-Clinical Modification code 428. Counties labeled as "insufficient data" (n = 2,160) did not have large enough populations of Hispanic Medicare beneficiaries to produce stable rates. Counties with the highest rates were located primarily in Texas, along the Northeast shoreline, and in counties along the border between Canada and Michigan, Ohio, Pennsylvania, and New York. *HF hospitalization rates are spatially smoothed to enhance the stability of rates in countries with small populations. Source: Medicare Provider Analysis and Review (MEDPAR).
|
|
The magnitude of geographic disparity was substantial between the high- and low-rate counties. For the total population, the midpoint of the rates in the top quintile was 4 times larger than the midpoint for the bottom quintile. For blacks, Hispanics, and whites, the ratios between the midpoints of the top and bottom quintiles were 4, 6, and 4, respectively.
|
Discussion
|
|---|
These county-level maps of HF hospitalization rates indicate that parts of the lower Mississippi River Valley and Ohio River Valley, including Appalachia, stand out as regions of the country with the heaviest burden from HF. In light of the increasing hospitalization rates for HF at the national level (1) and the heavy economic burden that accompanies high rates of HF hospitalizations (3), it is strategically advantageous to identify areas of the country that are most in need of efforts directed to the prevention of HF. Furthermore, the race/ethnicity-specific maps presented here can guide public health professionals at the local level as they design prevention and treatment policies and programs tailored toward the needs of their communities.
A recent study of HF hospitalization rates among the 4 large regions of the U.S. (Northeast, Midwest, South, and West), using National Hospital Discharge Data, found that among patients age 65 years old and older, the lowest rates were in the West and there was no consistent pattern among the other 3 regions (8). The use of large geographic units in that study may have masked the underlying geographic disparities that we observed among the smaller geographic units in this study. Another recent study examined 30-day readmission rates for HF among health referral regions and found a pattern similar to the patterns presented here for HF hospitalization rates (9). The areas with the highest readmission rates were located in parts of the Mississippi River Valley and segments of Appalachia. The similarities of these findings further highlight the need to address the heavy burden of HF in these communities.
HF is one of the ambulatory care sensitive conditions identified by the Agency for Healthcare Research and Quality (14). These are conditions for which effective outpatient care and early interventions can potentially prevent the need for hospitalization, reduce readmissions, and prevent complications or more severe disease. As such, high rates of HF hospitalizations may identify counties that are in particular need of improved access to quality health care. Many of the risk factors for HF (e.g., hypertension, diabetes mellitus, obesity, and AMI) are largely preventable with a combination of access to quality medical care and living and working conditions that promote healthy options for all residents. Furthermore, cardiac rehabilitation has been shown to improve outcomes for cardiac patients, thereby contributing to the prevention of subsequent HF (15). Therefore, clinicians, medical care organizations, health policy analysts, state health department officials, and other public health professionals can use data provided in these maps of HF hospitalization rates to guide their efforts at reducing the geographic disparities in HF around the country (16).
The data presented here are subject to several limitations given that Medicare data are collected for administrative purposes rather than explicitly for epidemiological studies. For instance, the data in this study depend on the accuracy of physician or administration reporting and coding. If there are geographic differences in financial incentives to report HF as the first-listed diagnosis, then these results could be explained by that coding bias; however, there is no evidence to suggest that such coding bias exists. In addition, we have observed similar geographic patterns of stroke hospitalizations in the Medicare population (17), which provides additional evidence to suggest that the observed geographic disparities are more likely to reflect geographic clustering of hypertension and other biomedical and socioenvironmental risk factors for cardiovascular diseases in those areas. The study population includes only patients in fee-for-service Medicare and excludes patients who are enrolled in managed care. Overall,  17% of Medicare enrollees were members of managed care organizations in 2004 (18). Moreover, the study population includes only patients 65 years of age. However, >80% of HF patients in the U.S. are 65 years of age (1), suggesting that our study population includes the majority of HF hospitalizations. Furthermore, Hispanics may be under-represented in the study population because of the procedures for reporting race and ethnicity in the Medicare datasets (11). These procedures could introduce bias into the results for Hispanics. Finally, this study does not include hospitalizations for which HF is listed as any mention on the Medicare claim form. Consequently, the HF hospitalization rates presented here under-represent the full burden of HF. Nevertheless, the considerably large Medicare claims dataset represents the best available information for assessing the burden of HF in the elderly population at the national, state, and local level.
These maps provide a snapshot of the existing geographic and racial/ethnic disparities in the burden of HF in the U.S. To better understand the contributions of other diseases of the heart to the observed patterns, and to complete the picture of the geographic disparities in heart disease, the Centers for Disease Control and Prevention is producing an Atlas of Heart Disease Hospitalizations among Medicare Beneficiaries that will be published online and in hard copy in the spring of 2010 (www.cdc.gov/dhdsp). The atlas will contain county-level maps of hospitalization rates for all heart diseases combined, coronary heart disease, AMI, HF, and cardiac dysrhythmia, along with 30-day mortality rates and discharge destinations (e.g., skilled nursing facility, home, and so forth). It is our hope that these maps will be used by health professionals to tailor prevention and treatment policies and programs to the needs of communities with heavy burdens of HF and other forms of heart disease.
|
Footnotes
|
|---|
Dr. Greer is an employee of Northrop Grumman Corporation. The other authors are government employees, and none of the authors received external financial support to produce this paper. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
|
References
|
|---|
1. Fang J, Mensah GA, Croft JB, Keenan NL. Heart failure-related hospitalization in the U.S., 1979 to 2004 J Am Coll Cardiol 2008;52:428-434.[Abstract/Free Full Text]2. Hunt SA, Abraham WT, Chin MH, et al. 2009 focused update incorporated into the ACC/AHA 2005 guidelines for the diagnosis and management of heart failure in adults: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines J Am Coll Cardiol 2009;53:e1-e90.[Free Full Text] 3. Lloyd-Jones D, Adams R, Carnethon M, et al. Heart disease and stroke statistics—2009 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee Circulation 2009;119:e21-e181.[Free Full Text] 4. Schocken DD, Benjamin EJ, Fonarow GC, et al. Prevention of heart failure: a scientific statement from the American Heart Association Councils on Epidemiology and Prevention, Clinical Cardiology, Cardiovascular Nursing, and High Blood Pressure Research; Quality of Care and Outcomes Research Interdisciplinary Working Group; and Functional Genomics and Translational Biology Interdisciplinary Working Group Circulation 2008;117:2544-2565.[Abstract/Free Full Text] 5. From AM, Leibson CL, Bursi F, et al. Diabetes in heart failure: prevalence and impact on outcome in the population Am J Med 2006;119:591-599.[CrossRef][Web of Science][Medline] 6. Levy D, Larson M, Vasan R, Kannel W, Ho K. The progression from hypertension to congestive heart failure JAMA 1996;275:1557-1562.[Abstract/Free Full Text] 7. Lloyd-Jones DM, Larson MG, Leip EP, et al. Lifetime risk for developing congestive heart failure: the Framingham Heart Study Circulation 2002;106:3068-3072.[Abstract/Free Full Text] 8. Zhang W, Watanabe-Galloway S. Ten-year secular trends for congestive heart failure hospitalizations: an analysis of regional differences in the United States Congest Heart Fail 2008;14:266-271.[Medline] 9. Krumholz HM, Merrill AR, Schone EM, et al. Patterns of hospital performance in acute myocardial infarction and heart failure 30-day mortality and readmission Circ Cardiovasc Qual Outcomes 2009;2:407-413.[Abstract/Free Full Text] 10. Department of Health and Human Services International Classification of Diseases, 9th Revision, Clinical Modification: ICD-9-CM. DHHS publication no. (PHS) 80-1260. Washington DC: Department of Health and Human Services; 1980. 11. Arday S, Arday D, Monroe S, Zhang J. HCFA's racial and ethnic data: current accuracy and recent improvements Health Care Financ Rev 2000;21:107-116.[Web of Science][Medline] 12. In: Haynes MA, Smedley BD, editors. The Unequal Burden of Cancer: An Assessment of NIH Research and Programs for Ethnic Minorities and the Medically Underserved. Washington, DC: National Academy Press; 1999. 13. Kaplan JB, Bennett T. Use of race and ethnicity in biomedical publication JAMA 2003;289:2709-2716.[Abstract/Free Full Text] 14. Agency for Healthcare Research and Quality AHRQ Quality Indicators—Guide to Prevention Quality Indicators: Hospital Admission for Ambulatory Care Sensitive Conditions. AHRQ publication no. 02-R0203. Rockville, MD: Agency for Healthcare Research and Quality; 2001. 15. Suaya JA, Stason WB, Ades PA, Normand SL, Shepard DS. Cardiac rehabilitation and survival in older coronary patients J Am Coll Cardiol 2009;54:25-33.[Abstract/Free Full Text] 16. Brush JEJ, Rensing E, Song F, et al. A statewide collaborative initiative to improve the quality of care for patients with acute myocardial infarction and heart failure Circulation 2009;119:1609-1615.[Abstract/Free Full Text] 17. Casper ML, Nwaise IA, Croft JB, Nilasena DS. Atlas of Stroke Hospitalizations Among Medicare BeneficiariesAtlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2008. 18. Centers for Disease Control and Prevention Racial disparities in total knee replacement among Medicare enrollees—United States, 2000–2006 Morbid Mortal Weekly Rep 2009;58:133-139.[Medline]
Related Article
-
Inside This Issue
J. Am. Coll. Cardiol. 2010 55: A35-A33.
[Full Text]
[PDF]
This article has been cited by other articles:
|
|
|
|
 
M. A. Hlatky and P. A. Heidenreich
The Year in Epidemiology, Health Services Research, and Outcomes Research
J. Am. Coll. Cardiol.,
May 10, 2011;
57(19):
1859 - 1866.
[Full Text]
[PDF]
|
|
|
|
Top
Abstract
Methods