Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
2 The Impact of Geography on Health Disparities in the United States: Different Perspectives W here an individual chooses to live can have a profound effect on their short- and long-term health. âEight Americas: Investigating Mortality Disparities Across Races, Counties, and Race-Counties in the United States,â a paper by Murray et al. (2006), examines the gap in life expectancies found in different parts of the United States in order to more fully elucidate issues related to health disparities in this country. During the public workshop, Dr. Murray presented this paper, along with additional research investigating mortality and causes of death at the local level in the United States. Dr. Acevedo-Garcia further discussed the connec- tion between a personâs place of residence and health disparities, focusing her comments on the impacts of living in specific neighborhood settings or metropolitan areas. Their presentations are recounted here. Other discussion topics and general comments raised during the session by Roundtable members, sponsors, and audience members are included in the Addressing Health DisparitiesâDifferent Perspectives section at the end of the chapter. The Murray et al. paper appears in Appendix C. EIGHT AMERICAS The Eight Americas presentation is based on analyses of county-level mortality data from the National Center for Health Statistics (NCHS), collected between 1960 and 2001, explained Dr. Murray. Graphic repre- â This section is an edited transcript of Dr. Christopher Murrayâs remarks at the workshop.
CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES sentations of county-level mortality data from 1997â2001 show how life expectancy rates for men (ranging from age 62.0 to age 80.2) and women (ranging from age 71.8 to age 84.5) vary depending on an individualâs county of birth (Figure 2-1). Like a mosaic, the shades depicting differ- ent life expectancy rates appear random and unrelated at first; however, further examination reveals that mortality patterns seem to follow specific geographic patterns. Data comparing county-level life expectancy rates for men and women over time can be examined in several ways. Figure 2-2 shows the standard deviation of the distribution of life expectancy across counties for men and women between 1960 and 2000. Although the lines depicting the standard deviation of county life expectancies for men and women follow similar trajectories, the differences across counties began to steadily increase after 1980. Similarly, tracking life expectancies for counties in the top and Âbottom 2.5 percentiles over time for men and women shows a similar result (Figure 2-3). Although the gap in life expectancies for the counties in the top and bottom 2.5 percent of all U.S. counties remained fairly constant from 1960 until around 1980, it has been growing since that time. Among the bottom 2.5 percent of counties, little or no progress in increasing life expectancies has been seen over the past 20 years. In absolute terms, the differentiation in life expectancies in U.S. counties continues to widen. According to the U.S. national average, and as seen in data from coun- ties that have historically had the highest life expectancies, male life expec- tancy has been increasing faster than female life expectancy. The counties with the highest life expectancies in the United States are at levels that surpass those seen in Japan, the country with the highest life expectancies globally. Defining the Eight Americas In addition to summarizing county-level analysis, race-countiesâ referring to the county of death and the race of the deceasedâwere ana- lyzed using 5-year moving averages. Life expectancies were calculated for race groups in every county where mortality among members of a certain race was large enough for the analysis. Data show that the range of life expectancies seen in the United States is even larger when comparing race- counties. Life expectancies as low as 58 years of age were calculated for Native Americans in southwestern South Dakota, and Asian women in Ber- gen County, New Jersey, have average life expectancies reaching 91 years of age. There is no evidence that the magnitude of the gaps is closing. Further analysis was conducted using the amassed county and race- county data to identify which diseases accounted for the existing mortality
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES Males Females FIGURE 2-1â County life expectancy 1997â2001. 2-01.eps bitmap images
10 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES 2.5 Standard deviation of county life expectancies (years) 2 1.5 Male Female 1 1960 1970 1980 1990 2000 Year FIGURE 2-2â Width of cross-country distribution of life expectancy. 2-02.eps patterns and the age groups for which the greatest differences in mortality were seen. Using county-level mortality figures proved problematic because they lacked statistical power; too few people were included in the figures from each county to track individual causes of death. The objective in using the new Eight Americas analysis was to identify a discrete number of sub- groups, each consisting of a population large enough to statistically analyze mortality by age, sex, and cause. The choice of eight Americasâversus any other numberâwas to identify a discrete number of subgroups that would have the power to capture most of the broad variation that is seen across counties and race-counties. The Eight Americas are defined in Table 2-1 and represented in Figures 2-4 and 2-5. Mortality and Cause of Death: Comparisons of the Eight Americas Using statistical analyses, it is possible to explore life expectancy and causes of death in the Eight Americas. Comparing trends in life expectancy between men and women in the Eight Americas did not show significant changes, indicating on a broad level that disparities are not decreasing and
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 11 Male 85 Top 2.5% 80 Bottom 2.5% Life expectancy (years) 75 70 65 60 1960 1970 1980 1990 2000 Year Female 85 Top 2.5% Bottom 2.5% 80 Life expectancy (years) 75 70 65 60 1960 1970 1980 1990 2000 Year FIGURE 2-3â Life expectancy for top and bottom 2.5 percent of counties. 2-03.eps
12 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES TABLE 2-1â The Eight Americas Average Income General Population Per America Description (millions) Capita Definition 1 Asian 10.4 $21,566 Asians living in counties where Pacific Islanders make up less than 40% of total Asian population 2 Northland 3.6 $17,758 Whites in northern plains and low-income Dakotas with 1990 county-level rural white per capita income below $11,775 and population density less than 100Â persons/km2 3 Middle 214.0 $24,640 All other whites not included in America Americas 2 and 4; Asians not in America 1, and Native Americans not in America 5 4 Low-income 16.6 $16,390 Whites in Appalachia and the whites in Mississippi Valley with 1990 Appalachia county-level per capita income below and the $11,775 Mississippi Valley 5 Western 1.0 $10,029 Native American populations in Native the mountain and plains areas, American predominantly on reservations 6 Black Middle 23.4 $15,412 All other black populations living in America counties not included in Americas 7 and 8 7 Southern 5.8 $10,463 Blacks living in counties in low-income Mississippi and the Deep South rural black with population density below 100 persons/km2; 1990 per capita income below $7,500, and total population size above 1,000 persons (to avoid small numbers) 8 High-risk 7.5 $14,800 Urban populations of more than urban black 150,000 blacks living in counties with cumulative probability of homicide death between 15 and 74Â years greater than 1.0% SOURCE: Adapted from Murray et al. (2006).
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 13 America 2 America 3 America 4 FIGURE 2-4â Americas 2, 3, and 4. 2-04.eps America 6 America 7 America 8 FIGURE 2-5â Americas 6, 7, and 8. 2-05.eps in some cases they are on the rise (see Figure 2-6). However the subtleties found in the graphs do show some interesting patterns in the data. America One, comprised of Asian Americans living in communities in which Pacific Islanders make up fewer than 40 percent of the total Asian population, has a high life expectancy that continues to increase. America Two shows a dwindling advantage over America Three (Middle America) among men and is showing marked improvement among women. America Four has had a very slow but steady rate of increase. The Appalachian,
14 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES Males 90 America 1 America 2 America 3 America 4 Life expectancy at birth (years) America 5 America 6 America 7 America 8 80 70 60 2000 1990 1996 1984 1986 1988 1998 1994 1992 1982 Year Females 90 Life expectancy at birth (years) 80 70 60 2000 1990 1996 1984 1986 1988 1998 1994 1992 1982 Year FIGURE 2-6â Life expectancy at birth in the Eight Americas. 2-06.eps
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 15 Mississippi Valley white populations are increasingly falling behind the rest of white America. Among Native American populations there has been little or no increase in life expectancy for women and moderate increases for men. A similar story holds true for Americas Six and Seven, the African American populations. The large dip in the line depicting life expectancies for men living in America Eight reflects, for the most part, the increase in HIV-related mortality and its subsequent decline. Overall, comparing the graphs for men and women in Americas One through Eight shows that there is very little change in the net effect between the early 1980s and 2000. County Trends in Life Expectancy The Eight Americas mortality database contains county-level data d Â ating back to 1960, and an analysis of those data shows a pattern of growing inequalities since 1983. To examine this phenomenon, county-level life expectancy data from 1961 to 1983 were compared with analogous data from 1983 to 1999. Male life expectancy data from 1961 to 1983 (Figure 2-7a) show several areas that have a statistically significant increase in life expectancy greater than the national average; several areas that are equal to or indistinguishable from the national average; and several coun- ties that have life expectancies that are statistically significant below the national average. Counties in red show areas in which there has been no statistically significant decline in life expectancy at the county level for the 22-year time period. Analysis showing life expectancy for men from 1983 to 1999 (Figure 2-7b) shows less progress; there are many more counties with a rate of change that is indistinguishable from zero, and some coun- ties near the Mississippi Valley have life expectancies that dropped. Among women, similar findings are seen when comparisons are made between life expectancy data from 1961 and 1983 (Figure 2-8a); however, data from 1983â1999 show that life expectancy among women in several counties is dropping (Figure 2-8b). With the exception of the Spanish flu pandemic of 1918 and 1920, there have been constant increases in life expectancy in the United States for more than 100 years, a finding consistent with life expectancy rates seen in other high-income countries. Yet there is a subset of the United States for which life expectancy at the county level for women, in particular, is dropping. This finding is quite unusual in recent mortality history among high-income countries. To help explain why life expectancy has been decreasing in certain segments of the United States, the cause of death for men and woman, compared by age group, was analyzed using county-level data. Analyses show that, among both men and women, mortality attributable to cardio-
16 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES A 2-07a.eps B bitmap image 2-07a.eps bitmap image 2-07b.eps bitmap image FIGURE 2-7â Change in male life expectancy: (a) 1961â1983, (b) 1983â1999.
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 17 A 2-08a.eps B bitmap image 2-08a.eps bitmap image 2-08b.eps FIGURE 2-8â Change in female life expectancy: (a) 1961â1983, (b) 1983â1999. bitmap image 2-08b.eps bitmap image
18 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES vascular disease has been declining steadily at rates that surpass increases in m Â ortality for other noncommunicable diseases. In younger age groups, mor- tality from HIV, homicide (in men), and lung cancer are evident. However, in the counties in which life expectancy is dropping, cardiovascular disease is actually increasing, unmasking the rise in other causes. It remains to be seen whether the general decline in cardiovascular disease in the United States will result in more counties experiencing a background rise in these other causes of death. County-level data were also used to examine the relationship between health disparity and wealth to determine whether the gap in poverty levels between counties would match the gap seen in overall health. Income infor- mation gleaned from the 2000 census, together with county-level data from the tax return database of the Internal Revenue Service, revealed that while life expectancy in several counties is dropping, the counties with decreasing life expectancies are not getting markedly poorer. Global Perspective of the Eight Americas An alternate way to examine the U.S. data is to compare the findings from the Eight Americas with similar data collected from other countries around the globe. Examining the relationship between health disparities and wealth, it becomes evident that other high-income countries are expe- riencing analogous increases in income levels, yet they are experiencing these changes without health disparities increasing as they are in the United States. Further research must be conducted to investigate why. The only other countries in which life expectancy has fallen in the past 50 years have been in the former Soviet Union and countries in Eastern and Southern Africa with high HIV prevalence. In the 1930s, some low-income countries experienced a rise in life expectancy, but in the past 50 years this occurrence has been limited to high-income countries. The combination of the Eight Americasâ patterns and trends, along with the recent discovery of the subset in the United States in which life expectancy is decreasing, shows that disparities are not getting smaller in the nation and, in fact, there is every expectation that they will continue to grow. To gain this global perspective, biomedical data representing the cause of death among men from two groups, America One and America Eight, were graphed along with comparable data from Japan, the United ÂKingdom, Russia, and West Africa (Figure 2-9). Comparisons were made looking at mortality figures for HIV/AIDS, intentional injuries, unintentional injuries, other noncommunicable diseases, other communicable diseases, cardioÂ vascular diseases, and cancers. Among children ages 0â4 years, mortality in America Eight is much higher than America One and is about equivalent to the Russian Federa-
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 19 0.75 HIV/AIDS Intentional injuries Unintentional injuries 0.50 Other non-communicable diseases Other communicable diseases Cardiovascular diseases 0.25 Cancers 0.00 America 1 Japan UK Russia America 8 AFR-high-mortality America 1 Japan UK Russia America 8 AFR-high-mortality America 1 Japan UK Russia America 8 AFR-high-mortality America 1 Japan UK Russia America 8 AFR-high-mortality America 1 Japan UK Russia America 8 AFR-high-mortality 0 - 4 years 15 - 44 years 45 - 64 years 65 - 74 years 75 - 84 years FIGURE 2-9â Male causes of death in the Eight Americas compared to Japan, the United Kingdom, Russia, and West Africa. NOTE: AFR-high-mortality, made up 2-09.eps largely of countries in West Africa and excluding countries with very high mortality due to HIV/AIDS. SOURCE: Murray et al. (2006). tion. Child mortality in America Eight is almost 10 times lower than in West Africa, meaning that a child living under the harshest conditions in America will still fare considerably better than a child living in West Africa. These findings change dramatically when comparing young high-risk urban black American men to West African men ages 15 to 44, and middle-aged adults ages 45 to 64. In these instances the gap among America Eight, West Africa, and the Russian Federation is not large. In fact, rather than m Â ortality being 10 times lower for individuals in the United States as seen among children 0â4 years of age, mortality among the higher age groups is nearly equivalent. The difference in mortality seen among men cannot be attributed solely to increases in HIV rates. If HIV is not factored in for any of the countries, mortality patterns found in America Eight are still comparable to those found in West Africa. The excess mortality among men is related to communicable and noncommunicable diseases, cardiovascular disease, and to some extent cancers, rather than to homicide and HIV, as might be expected. The mortality pattern for women is very similar. Another way to examine these data is to compare mortality rates across the Eight Americas with the range found in high-income countries
20 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES 2.8 2.4 Relative mortality index 2 1.6 1.2 0.8 0.4 0 -0.4 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 America 1 America 2 America 3 America 4 America 5 America 6 America 7 America 8 2-10.eps FIGURE 2-10â Mortality in the Eight Americas by age compared to other high- income countries. (Figure 2-10). The relative mortality index measured on the Y axis ranges from â0.4 to 2.8, meaning that any mortality rate in America for a specific age group is 2.6 times higher (or lower) than the worst of the high-income countries. Therefore, any relative mortality index greater than 1.0 denotes a subgroup in the United States (Americas One to Eight) that fares worse than any other high-income country at the national level. While child mortality rates in the United States compare favorably to those found in other high- income countries, mortality for young and middle-aged adults in Americas Five to Eight fall dramatically behind other high-income countries. Risk Factors To investigate what risk factors contribute to biomedical and nonÂ communicable diseases, large databases were sought to provide the nec- essary input. The National Health and Nutrition Examination Survey (NHANES), typically the best source for biomedical measurements and biomarkers, did not yield sample sizes large enough for the data to be analyzed at the level of the Eight Americas. The Behavioral Risk Factor â In-depth survey compiled by the Centers for Disease Control and Preventionâs NCHS that combines in-person interviews with standardized physical examinations, diagnostic proce- dures, and lab tests with national rather than state representation (NCHS, 2008).
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 21 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1 2 3 4 5 6 7 8 America Male Female FIGURE 2-11â Difficult-to-explain disparities between Eight Americas on the basis 2-11.eps of reported health care access. Surveillance System (BRFSS) is a rich data source, but it does not provide biomarker information. In order to use the data from the two sources, a method was devised to combine BRFSS and NHANES data to determine how self-reported measures from both data sets were related, and then esti- mates were calculated for some of the risk factors that require biomarkers. Several interesting observations were made using this method. BRFSS data were used to examine health plan coverage across the Eight Americas (Figure 2-11). The lowest self-reported health plan coverage was seen in America Five (Western Native Americans); however, this may be an anomaly since health care is provided on reservations in addition to health plan coverage. Slightly lower health plan coverage is also seen in America Seven compared with the other subgroups. Further analyses looking at utilization measured by self-reported data in response to the question âDid you see a doctor or have a check-up in the last twelve months?â failed to show a dramatic gradient in financial access that is strongly related to the disparities seen in the outcomes. What about behavioral risk factors such as tobacco, diet, or physical activity? The recent World Health Organizationâs Comparative Risk Assess- ment (WHO, 2002) work highlighted that the likely candidates for explain- â State-based system of telephone-administered health surveys that collects information on health risk behaviors, preventive health practices, and health care access (BRFSS, 2008).
22 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES ing behavioral risk factors in the United States are tobacco, alcohol, obesity, high blood pressure, and blood sugar. Therefore, an attempt was made to quantify these risks across the Eight Americas (Figure 2-12aâe). Of the major noncommunicable risk factors that are important determi- nants of health in the United States, obesity, blood sugar, and hypertension appear to be strongly related to the gradients across the Eight Americas. Considerably more work must be done to determine what the net effect of addressing each of those risk factors would have on the differences that are seen across the Eight Americas. Policy Focus When considering potential policy implications related to reducing health disparities, it is important to speculate beyond the health insur- ance debate. In the policy arena there is an assumption that improving financial access would address many of the disparities seen in the United States. However, this is unlikely to be the case although providing financial access would undoubtedly reduce catastrophic spending and result in some improvements in health. A recent Institute of Medicine report (IOM, 2002) enumerated the number of deaths attributable to the lack of health insur- ance; however, while not inconsequential, the number of deaths attributable to the lack of health insurance would not be comparable to the increase in mortality attributable to health disparities seen across the Eight Americas. The analysis around the Eight Americas focuses on noncommunicable diseases particularly in young and middle-aged adults. Yet adults in these age groups do not benefit from Medicare and therefore their medical inter- ests are rarely considered in matters relating to public finance or national medical policy. To decrease mortality from noncommunicable diseases, fundamental public health principles would advocate for changes in diet, physical activity, tobacco and alcohol use, and biological risk factors that can be managed through primary care, such as blood pressure, blood sugar, and cholesterol. Can noncommunicable diseases be addressed even more effectively? It seems unlikely, given the experiences of other countries, that one solution will work for the entire U.S. population given the differences across the Eight Americas; both substantively in terms of health patterns and because of cultural variations among different groups. Although there has been considerable attention on reducing health disparities and innu- merable public and philanthropic programs have focused on reaching this goal, little success has been made toward decreasing health disparities at the national level. Dr. Murray believes the appropriate policy response would be to foster a broader spectrum of innovation in addressing both the behav-
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 23 ioral risks and the pharmacologically manageable biological risks for nonÂ communicable disease. It is also essential to have rigorous monitoring and ongoing evaluation when fostering innovation. With these measures in place, successful programs can be recognized, their results can be docu- mented, and the Âmethods can be shared and replicated nationally. This is an appropriate model to follow given how little is known about changing risksâpharmacologically manageable risks and othersâat the local level. By looking at health care from different perspectives, the management of risk factorsâblood pressure, cholesterol, and blood sugarâcan be evalu- ated in relation to other existing data sources. NHANES data, for example, show that only 50 percent of hypertensives are being managed, and about 25 percent of people with high cholesterol and about 45 percent of diabetics are being treated. Examining these data in terms of who is receiving appro- priate treatment reveals that the number of patients receiving appropriate care dropped to dramatically low levels in terms of outcomes. In a coun- try that spends the most per capita on health care by a large margin, the enormously important risk factors for which there are effective therapies are being managed by only 50 percent of hypertensives. The only way to broaden the reach of effective interventions is through innovations at the local level. There are many models to investigate for reducing health disparities in the United States. Pay-for-performance models have been proposed, as well as the idea of creating incentives for individuals who receive treatment to reduce their health risk. Conditional cash transfers or financial incen- tives, for example, could be given for getting blood pressure checked or for managing cholesterol appropriately. A paradigm for this progressive model can be seen in Mexico, where randomized trials looking at conditional cash transfers have shown promise for getting people to use preventive services. While there are no obvious answers for addressing these issues, there are several local innovations that should be tested. Murray noted that while disparities are increasing in the United States, infant, child, and adult mortality and life expectancy are consistently drop- ping the ranking of the United States among other countries when health outcomes are compared. Over the last 30 years, the U.S. position in a table of comparable health outcomes has steadily fallen. Alternately, other countries have a rate of improvement that is higher than that of the United States. In conclusion, he observed, it is imperative that the United States adopt an agenda for improving health for Americas Four, Five, Six, Seven, and Eight, as well as the majority of Middle America who are falling behind the rest of the high-income world. Perhaps the answers to addressing health disparities will also have direct relevance for understanding why the United States is falling behind other high-income countries around the world.
24 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES A Self-Reported Smoking 40% 35% 30% 25% 20% 15% 10% 5% 0% 1 2 3 4 5 6 7 8 America Male Female 2-12a.eps B Alcohol: Average Number of Drinks Per Day 0.8 0.6 0.4 0.2 0 1 2 3 4 5 6 7 8 America Male Female 2-12b.eps FIGURE 2-12â Quantifying risks across the Eight Americas: (a) Smoking: Tobacco is one of the most important risk factors for health and reducing smoking would markedly improve everybodyâs health. However, since no clear gradient is seen in the data, decreasing smoking rates may reduce some of the disparities, but ulti- mately would not have a large affect. (b) Drinks of alcohol per day: America Two, which has the best mortality for white Americans, has the highest average number of drinks per day. Drinks of alcohol per day is an important public health risk fac- tors, but it may not hold the key to understanding the huge gradients across the Americas. (c) Obesity: Obesity data were calculated from BRFSS data corrected for self-report files. For women there is a marked gradient across the Eight Americas in obesity, going up such that Americas 5â8 are greater than 45 percent obese for women and over 30 percent obese for men. Future work will use attributable mor- tality calculations to examine what the reduction in disparities and life expectancy would be if obesity could be reduced. (d) Uncontrolled hypertension: Substantial gradients exist across the Eight Americas where there are levels of hypertension for women, reaching nearly 30 percent in the bottom Americas 6â8, and a gradient that is smaller, but nevertheless still very substantial for men. (e) Blood sugar: A steady gradient exists, perhaps correlated to the obesity pattern. Blood sugar levels
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 25 C Obesity Corrected for Self-Report Bias 60% 50% 40% 30% 20% 10% 0% 1 2 3 4 5 6 7 8 America Male Female 2-12c.eps D Uncontrolled Hypertension 35% 30% 25% 20% 15% 10% 5% 0% 1 2 3 4 5 6 7 8 America Male Female E 2-12d.eps Blood Sugar 110 108 106 104 Mean FPG 102 100 98 96 94 92 90 1 2 3 4 5 6 7 8 America Male Female 2-12e.eps are markedly higher in Native Americans which may be caused in part by a genetic component. Again, the pattern resembles the gradient seen in outcomes across the Eight Americas.
26 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES Additional Clarification Dr. Murray was asked why Hispanics were not included in the Eight Americas, since they represent the largest growing minority group in the United States. He explained that Hispanic Americans were not represented because county-level life expectancy data comparing death certificates and census reports for Hispanic are unreliable; recorded life expectancies in some counties were found to be as high as 190 to 250 years of age. Fun- damental differences exist between how Hispanic status gets reported by physicians or relatives on death certificates and self-reported data on the census, and there has been little success in finding a way to reconcile these data sources. Compared with other groups in a similar place or with a simi- lar socioeconomic status, Hispanics seem to have higher life Âexpectancies. This would mean that the white Hispanic population would be comparable to that of America Two. The data for black Hispanic populations are undetermined. OTHER PERSPECTIVES: THE INFLUENCE OF GEOGRAPHY ON HEALTH DISPARITIES Where a person resides, or features of a particular area, can have a direct impact on his or her health outcomes and mortality, explained Dr. Acevedo-Garcia in her presentation to the Roundtable. When examin- ing geographic disparities, whether using the geographic boundaries defined in the Eight Americas or using other geographic boundaries, it is important to make connections between those geographies and the social determinants of a health framework. Specific areas should map onto socioeconomic fac- tors driving racial inequality, such as state income inequality or disparities in neighborhood environments in metropolitan areas. The distribution of opportunity across neighborhoods in metropolitan areas can have a pro- found effect on socioeconomic advancement and health outcomes. Metropolitan areas are defined as core urban areasâthe central cityâ and the surrounding suburbs. These distinctions are important because housing and labor markets, as well as other factors that shape opportunity in America, operate differently between and across metropolitan areas. The majority of the people in the United States live in metropolitan areas. Two-thirds of all of the children in the United States live in the 100 largest metropolitan areas, and 40 percent (18 of 45 million) of those children live in what are called majority/minority metropolitan areas in which minority children are actually the majority (Acevedo-Garcia et al., 2007). Examples from the 2000 census (Acevedo-Garcia et al., 2007) include the following: â This section is an edited transcript of Dr. Acevedo-Garciaâs remarks at the workshop.
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 27 â¢ Chicago (2.2 million children; 51% minority [nonwhite]) â¢ Dallas (1 million children; 53% minority [nonwhite]) â¢ Los Angeles (2.7 million children; 80% minority [nonwhite]) â¢ Washington, DC (1.3 million children; 50% minority [nonwhite]) Metropolitan areas have been examined very extensively from an eco- logical, demographic, and urban planning perspective, because they shape equality in the United States. Although large disparities are not reflected in U.S. child mortality rates, it is important to focus on children and adolescents because, from a devel- opmental life-course perspective, what happens in childhood is going to impact disease outcomes in adulthood. How opportunity is experienced in metropolitan areas will directly impact the environment in which children live. In turn, this will affect their life course and, subsequently, the long- term economic disparities extant in those metropolitan regions. Neighborhoods as the Focus of Health Disparities While to a large extent the nationâs health care policy is determined by policies instituted by individual states, research from such disciplines as social epidemiology and human development shows that neighborhoods have an impact on childrenâs health and their developmental outcomes, above and beyond individual- and family-level factors. Poor neighborhood conditions may put children at risk for developmental delays, teen parent- hood, and academic failure, resulting in long-term implications throughout the life course. Factors such as access to healthy foods and the safety of the environment will determine a neighborhoodâs influence on the residentsâ health. Disadvantaged neighborhood environments are associated with hazardous physical environments, poor-performing schools, and a lack of public safety (Brooks-Gunn et al., 1997). Consistent with the findings of the Eight Americas, research suggests that health and social determinants show large geographic variations in absolute terms and in the level of disparities. Opportunity-rich neighborhoods do exist in metropolitan areas, but not all children have access to them. Therefore, the neighborhoods to which everyone should aspire already exist in each metropolitan area. The challenge is making those neighborhoods accessible to everyone. Housing m Â arkets in metropolitan areas are structured so that there are vast dis- parities in access to neighborhoods with opportunity. Large disparities in opportunity in metropolitan areas have a substantial impact on the well- being of Americaâs children and, in turn, on economic and social prospects of entire metropolitan regions. Metropolitan areas with better health out- comes and smaller disparities should ideally serve as examples for other
28 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES communities, and their policies and procedures should be evaluated and replicated in other areas across the nation. To gain perspective on the variation that exists across metropoli- tan areas, comparisons were made between rates of low birth weight by e Â thnicity (Asian, black, Hispanic, and white) for the 100 largest metro- politan areas in the United States for the years 2001â2002 (Figure 2-13) using data from the National Center for Health Statistics Vital Statistics. Low-birth-weight babies, weighing less than 2.5 kilograms at birth, are considered at increased risk for negative health outcomes, including higher rates of infant mortality. The analysis compared low-birth-weight rates ranked from 3 to 6 percent of all births, up to 9 to 12 percent of births in each metropolitan area. According to Healthy People 2010 objectives, the nation should strive for a low-birth-weight rate of 5 percent. In over 90 percent of metropolitan areas, white children have low-birth- weight rates between 3 and 6 percent, a rate very similar to the distribution of low-birth-weight children among Hispanics. This means that there is not a large disparity when comparing low birth weights between whites and Hispanics. However, there are large disparities between ÂHispanics and whites when comparing socioeconomic outcomes. Consequently, although significant problems in terms of some health outcomes do not currently exist, the conditions under which Hispanic children are living are deterio- rating due to their declining socioeconomic status, and this will predictably and negatively affect outcomes early in their lives. In nearly 70 percent of 100 90 White Black Hispanic Asian 80 Percent of Metro Areas 70 60 50 40 30 20 10 0 0-3 3-6 6-9 9-12 12-15 15-18 18-21 21-24 Percent of Births Less Than 2.5 kg FIGURE 2-13â Low-birth-weight rates:2-13.eps by race/ethnicity. distribution SOURCE: Dr. Acevedo-Garcia slide presentation.
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 29 metropolitan areas, black children have low-birth-weight rates between 9 and 12 percent. Efforts are under way to quantify the differences seen among different racial and ethnic groups, but it should be noted that varia- tions result in entirely different worlds of opportunity and ranges of access to different positive influences and experiences in metropolitan areas. Asian children generally cluster closely to the white children in the best possible part of the distribution. This finding mirrors the findings of Dr. Murrayâs research. If one were to graph a theoretical equal representation of poverty rates for neighborhoods in a metropolitan area, the distribution would be similar to that seen in Figure 2-14a. Research has shown that residing in neighborhoods with 20â40 percent poverty rates can impact child develop- ment (Brooks-Gunn et al., 1997). If this hypothetical distribution existed, less than 10 percent of all children would live in neighborhoods in which poverty rates exceed 40 percent. There would also be a comparable distri- bution of black and white children. This is a theoretical metropolitan area that is not found anywhere. In the Chicago metropolitan area, fewer than 25 percent of all black children live in neighborhoods with low poverty rates between 0â10 per- cent; the remainder lives in neighborhoods with poverty rates between 10.1 and 40 percent (Figure 2-14b). In contrast, over 85 percent of white children living in the Chicago metropolitan area live in neighborhoods with poverty rates below 10.1 percent, and the majority of the remaining chil- dren live in areas with poverty rates â¤20 percent. There is very little overlap between the distributions of neighborhood quality for white children and black children. Opportunity neighborhoods exist for white children, but, on the whole, black children live in totally different neighborhoods. The distribution of neighborhood quality is not solely dictated by a familyâs socioeconomic status. Comparing the distribution of poor black children with poor white children in Chicago (Figure 2-14c) shows that less than 5 percent of poor black children live in low-poverty neighborhoods, and more than 95 percent live in high-poverty neighborhoods (Acevedo- Garcia et al., 2007). Black families, even those with a higher income, tend to live in high-poverty neighborhoods, while white families with lower incomes are more likely to live in higher income neighborhoods. Nearly 75Â percent of poor white children live in neighborhoods in which the pov- erty level is â¤10 percent. This means that white children do not live in areas in which they have to contend with familial and environmental pressures associated with living in high-poverty neighborhoods. When the poverty composition of neighborhoods is analyzed by com- paring nonpoor black children to poor white children, the distribution of poor white children remains more favorable than the distribution for nonpoor black children (Figure 2-14d). This finding is consistent with
30 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES A Pyramid Graph: Theoretical Equal Neighborhood Environment for 2 Groups: A Mirror Image Over 40% Black White Neighborhood Poverty Rate 30.1-40% 20.1-30% 10.1-20% 0-10% 100 75 50 25 0 25 50 75 100 Share of Children in Neighborhoods with Specified Poverty Rates 2-14a.eps B Metro Chicago Poverty Composition of Neighborhoods of Black v. White Children Over 40% Black White Neighborhood Poverty Rate 30.1-40% 20.1-30% 10.1-20% 0-10% 100 75 50 25 0 25 50 75 100 Share of Children in Neighborhoods with Specified Poverty Rates FIGURE 2-14aâdâ Distribution of poverty rates for neighborhoods in a metropoli- tan area. 2-14b.eps
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 31 C Metro Chicago Poverty Composition of Neighborhoods of Poor Black v. Poor White Children Over 40% Poor Black Poor White Neighborhood Poverty Rate 30.1-40% 20.1-30% 10.1-20% 0-10% 100 75 50 25 0 25 50 75 100 Share of Children in Neighborhoods with Specified Poverty Rates 2-14c.eps D Metro Chicago Poverty Composition of Neighborhoods of All Black v. Poor White Children Over 40% All Black Poor White Neighborhood Poverty Rate 30.1-40% 20.1-30% 10.1-20% 0-10% 100 75 50 25 0 25 50 75 100 Share of Children in Neighborhoods with Specified Poverty Rates 2-14d.eps
32 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES distributions from other metropolitan cities that were analyzed. When analyses were done comparing neighborhood poverty rates for black, white, and Hispanic children living in the 100 largest U.S. metropolitan areas in 2000, it was evident that black and Hispanic children consistently live in neighborhoods with much higher poverty rates than white children. In fact, the socioeconomic profile for Hispanic children at the family, neighborhood, and school levels is similar to the profile of black children. If Hispanic children continue to live in these high-poverty environments, outcomes for Hispanic children will eventually resemble the outcomes seen for black children in similar neighborhoods. The long-term implications of this should be considered, stressed Dr. Acevedo-Garcia, especially in light of the fact that Hispanics are the largest minority group. Using economic indicators compiled from the 100 largest metropolitan areas, the best and worst neighborhood environments for black, white, Hispanic, and Asian children were identified. Additional analyses were done to determine the worst neighborhoods as far as disparities for Asian, black, and Hispanic children compared with white children (see Table 2-2). In the areas designated as having the worst disparities for black children, the share of black children living in low-income neighborhoods was more than 10 times larger than the share of white children living in low-income neighborhoods. TABLE 2-2â Best and Worst Neighborhoods from the 100 Largest Metropolitan Areas Worst Neighborhoods with Best Neighborhood Neighborhood the Worst Disparities Race/Ethnicity Environments Environments Compared to Whites Asian Austin, TX Bakersfield, CA Milwaukee-Wausheka, WI Baltimore, MD Fresno, CA Minneapolis-St. Paul, MN Washington, DC New York, NY Black Denver, CO Buffalo, NY Mobile, AL Colorado Springs, CO Chicago, IL Detroit, MI Raleigh-Durham- New York, NY Chicago, IL Chapel Hill, NC Hispanic Ann Arbor, MI Bakersfield, CA Chicago, IL Cincinnati, OH Providence, RI Hartford, CT Washington, DC Springfield, MA Milwaukee-Wausheka, WI White Ann Arbor, MI Bakersfield, CA Boston, MA El Paso, TX San Francisco, CA New York, NY
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 33 Availability and access to health insurance can have a dramatic impact on people in different minority groups, especially for Hispanics. ÂHispanics are uninsured at rates that surpass other ethnic groups, and there are national and state policies currently in place that limit the access of Â Hispanics to some services. One example of this was recently seen during discussions of the State Childrenâs Health Insurance Program reauthorization, when coverage for illegal immigrant children was highly debated and eventually defeated. Policies such as these restrict access to health care for immigrant children, a group among which Hispanic children are the majority. The impact of health insurance can also be seen when this issue is analyzed with a regional or local focus. The Relationship Between Geography and Policy It is important consider whether health inequities identified between geographic areas can be rectified through policy change. When analyzing geographies areas, geographic entities should be identified that are action- able from a political and policy standpoint. Since political and policy systems in the United States are structured geographically, finding action- able solutions to health disparities and other disparities can be extremely challenging. Patterns of devolution to the states in significant areas of social policy result in, for example, large variations in state child welfare policies. The variations prevent convergence to uniform policies and impede efforts to reduce disparities. Political fragmentation in metropolitan areas also makes disparities issues very difficult to address. There is a need for socioeconomic interventions and public health or medical interventions, continued Dr. Acevedo-Garcia. Additional research is needed to determine how disparities can be reduced using both types of interventions. In the current political climate, instituting race-based solu- tions is going to be increasingly more difficult. The Supreme Court recently ruled against school integration programs, and other race-based initiatives are being challenged. For these reasons, it will be necessary to find new frameworks for looking at disparities. If the geography for opportunity framework was to be used and children were redistributed using indica- tors of opportunity only, the result would be a racially integrated society. Creative strategies such as these must be developed for addressing racially focused issues when efforts to enact policy changes are stymied by political inaction or a lack of political will. Policies to reduce residential segregation include expanding neighbor- hood choice in the Department of Housing and Urban Development Sec- tionÂ 8 Voucher Program, fair housing enforcement, inclusionary zoning, and increased availability of rental housing. If availability of rental housing is restricted or there are restrictions on density for certain areas, minority
34 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES households are directly affected because Hispanic and black Americans are disproportionately renters rather than homeowners. Even poor whites have a greater likelihood of being homeowners compared with minorities because of the unequal distribution of wealth and the propensity for white Americans to benefit from intergenerational bequests. In the housing policy arena, people are very much at ease with dis- cussing place-based approaches versus people-based approaches, interven- ing in disadvantaged neighborhoods versus moving people from failing neighborhoods into better neighborhoods. Organizations such as the B Â rookings Institution or the Urban Institute contend that the nation has to do both. The majority of public health professionals have tra- ditionally favored approaches that focus on improving disadvantaged neighborhoods, rather than moving people to new neighborhoods, but it is extremely challenging to try to improve neighborhoods with high poverty rates. Some economic interventions can make a difference, but the problems that exist in these neighborhoods are deeply entrenched and interconnected. Schools, for example, are integral to the neighborhoods, and improving schools can improve neighborhoods to an extent. Yet fix- ing one facet of a failing neighborhood does not guarantee that success and prosperity will follow. Baltimore is embarking on a program that has moved about 1,000 families out of the inner city and into the suburbs; 1,000 additional families are waiting to move. The program aims to provide education, employment, and health services to ensure that the move to the suburbs will be success- ful. Programs such as these show great promise. To address issues related to health, more researchers have been focus- ing on disparities among children, and there is now a great deal of evi- dence showing that brain development has a significant impact later in life, especially involving cognitive development (IOM, 2000). Not addressing disparities may have serious socioeconomic implications. Hispanics play a very significant part in the growth of the workforce, especially among the low-wage workforce. Overall productivity could be affected if these issues are not adequately addressed. Targeting children is beneficial because they are disproportionately minorities, compared with the U.S. population. The fact that the nation has younger age distributions for minorities than for whites means that going forward there will be greater reliance on minorities to finance the way people live, including who is going to be working and who is going to be paying for Social Security and so on. This needs to be emphasized, because the moral issues are not going to be compelling enough to foster change.
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 35 Additional Clarification Dr. Escarce questioned whether childrenâs access to quality educa- tion was more highly correlated with the findings that were shown in Dr. Acevedo-Garciaâs presentation rather than issues regarding racial seg- regation. In response, Dr. Acevedo-Garcia explained that the school data mimicked the patterns seen in the neighborhood data. This would be expected, she believed, since schools are neighborhood-based in the United States. Any kind of segregation and socioeconomic inequalities that exist at the neighborhood level would be translated into the school systems. Hispanic and black children who are disproportionately likely to be poor because their families are poor and live in poor neighborhoods are also going to be disproportionately likely to go to poor schools. Dr. Acevedo- Garcia referred to this as a system of triple jeopardyâpoor neighborhoods, poor families, poor schoolsâand stressed that this composite of problems was one facet of disparities that needs to be addressed. Dr. Escarce also pointed out that most of Dr. Acevedo-Garciaâs solu- tions, such as rental housing and Section 8, could be construed as trying to improve the way the housing market works. Furthermore, he noted that a great deal of the problem is actually race-based, not socioeconomically based, and that Dr. Acevedo-Garciaâs own data would show that the vast majority of white children who are poor do not live in communities in which a lot of people are poor. Even if one could decrease racial segrega- tion and give more minority children better opportunities, the number of children who could be affected would be quite limited. All of the minority kids who are living in very poor neighborhoods could not be moved to other communities. In response, Dr. Acevedo-Garcia discussed the intervention in ÂBaltimore. To settle a desegregation lawsuit, the courts have proposed that African American residents who choose to participate will be moved from public housing projects in Baltimore City to the suburbs of Baltimore. This is a regional initiative focused on moving people to more affluent suburban areas. Meetings have been held with the program coordinators to deter- mine how to integrate a health component into the program to ensure that people who move to the suburbs do not lose their health care coverage and to make certain that people who move have access to everything they need. However, it has been very difficult to get the sponsoring foundations to focus on adding a health component to their housing initiative. It is important to recognize that funding initiatives that are intersectorial tend to be unpopular, and people lack incentives to work on things that combine more than one sector. Dr. Acevedo-Garcia also stressed that there is a great deal of evidence showing that early childhood programs, more ambitious than Head Start,
36 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES are needed to try to eliminate some of the inequalities among children. Another solution would be to move children to better neighborhoods so that, ultimately, fewer expensive public health interventions or early child- hood education programs would be required. In response, Dr. Escarce commented that programs such as these seem unlikely to happen politically, although he agreed that this was probably the only feasible approach for children. New York has a similar program in which small high schools are being created in communities. He also mentioned programs funded by the Bill and Melinda Gates Foundation and their investment in the Global Alliance for Vaccines and Immunization (GAVI). As described by Dr. Murray, GAVI was created as a publicâprivate partnership through a grant from the Bill and Melinda Gates Foundation, to find a way to increase immunization rates using local innovation. The Global Alliance asked countries to apply and propose how they were going to raise childhood immunization; it did not say how they should do it. They simply said that after three years they would pay $20 for every child who is immunized. In order to develop programs to alleviate health disparities, explained Dr. Escarce, it will be necessary to stop trying to decide whether problems are caused by social determinants or whether or not they are public health problems. There is need to move beyond descriptive academic analyses to testing innovative solutions. There needs to be a national fund for innova- tive health improvement that has the same attributes that GAVI has shown will work. A publicâprivate partnership, a large pool of resources, local applications, payment for progress, and a strongly embedded monitoring and evaluation program are all necessary in order to learn what is work- ing, continued Dr. Escarce. Further academic debate is also very useful and important to make a shift from describing the inequalities that either stabilize or grow to actually narrowing them. The only way to accomplish this is to take models that foster innovation and subject them to rigorous assessment. ADDRESSING HEALTH DISPARITIESâDIFFERENT PERSPECTIVES Reaction to the discussions regarding the relationship between health disparities and geography was thoughtful and, at times, passionate. Several of the issues discussed by the panelists and audience membersâthe current state of politics in the United States, language and framing, institutional racism, data collection problems, collaborations and community innova- tions, and health disparities approaches in St. Louisâare detailed here. â The following discussions were edited and organized around major themes to provide a more readable summary and to eliminate duplication of topics.
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 37 Health Disparities and U.S. Politics Several audience members shared concerns about the difficulty of addressing health disparities concerns in the current political environment. President Bush plans to veto a bipartisan bill that would enhance cover- age for young people and improve access to basic health care, commented Dr.Â Suggs, because it would be a step toward what President Bush referred to as socialized medicine. However, in Dr. Suggsâ opinion, this kind of con- text to discuss problems inhibits the opportunity to find satisfactory conclu- sions or remedies. Using a highly charged term like âsocialized medicineâ polarizes the issue and refocuses the debate to one of ideology rather than finding appropriate solutions to complex issues such as health care reform. Some of the problems that are being dealt with could benefit from a more open, objective, and candid discussion, continued Dr. Suggs. For example, if socialized medicine is untenable, then how can the programs that are in place for elected officials in Congress, for members of the U.S. military through the Department of Veterans Affairs, or, to some extent, for people who benefit from Medicare or Medicaid be rationalized? Todayâs burgeoning health care costs have a tremendous effect on society and on general access to quality health care, Dr. Suggs continued. Having power and wealth concentrated among a few industries, such as the pharmaceutical industry and the professional health care industry, can make it difficult to try to enact the reforms necessary to alleviate health disparities in the United States. Medical advancements have increased life expectancies and led to medical interventions that save lives. This also means that individuals who may have died from an illness in years past, can now lead long lives with the aid of hospice or long-term medical care. Yet a disproportionate percentage of people needing these long-term health care options are not protected by the health care policies that are in place today, stated Dr. Suggs. Alternatively, there are people who benefit from the current system without having paid into it. When my mother was growing up as a poor black woman in Mississippi, Dr. Suggs said, the actuary said that she would be dead by the time she was 50 years old. Therefore, when the Social Security system was put into place, it did not include my mother. My mother is now 94 years old and has, in a sense, been a beneficiary of a program that was never intended for black people. This example demon- strates some if the unintended consequences that occur when people who shape public policy ignore the problems related to disparities, concluded Dr. Suggs. When a pronouncement is made that universal health insurance alone is not enough, it should be expressed with the caveat that people should not step back from advocating for universal coverage, commented another audience member. People need to abandon the notion that health insurance
38 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES is unnecessary because emergency room care is accessible to everyone. It is important to continue to advocate for insurance, but simultaneously to convince people that an insurance plan or universal coverage alone is not enough, he continued. The health industry in the United States should be based on a foundation of ensuring good health rather than administering sick care as it does now. True health care would take up issues like housing, healthy environments, employment, and income disparities. These are the important things that impact peopleâs lives. Many health issues are not going to have simple, clinical, insurable inter- ventions, added another member of the audience. Consideration should be given to the environment in which people live and the effects of the choices that people are able to make given the options that they have available. This would ideally lead to a general population approach that, along with changes to insurance policies, could help ensure that people are healthier and, ultimately, that maintaining good health would be less costly. Framing Health Disparity Issues for a Broader Audience Several workshop participants were concerned with finding a way to discuss or frame issues related to health disparities in a way that will reso- nate with policy makers and government workers, and also to capture the publicâs attention, both locally and nationally. Fifty years ago the world adopted national income and product accounting, such that the field of macroeconomics was created, explained Dr. Murray. After World War II, people started benchmarking income per capita, and the annual growth rate at income per capita became a central policy target. In the 1970s, when Japan had a higher income growth rate than the United States, it had an incredible affect on the media, the Ameri- can public, and the U.S. policy debate, continued Dr. Murray. People were very concerned about why the Japanese were pulling ahead of the United States. What is needed now is a situation in which the American public focuses on why the nation is falling behind other high-income countries in terms of health. Dr. Levi added that an abundance of data suggests that neither poli- ticians nor the American public like hearing or admitting that another countryâs systems or programs are superior. Americans want an American solution; embracing a French solution or a Canadian solution is not some- thing that resonates. The challenge is to find a way to frame health dispari- ties issues in this country so that people recognize that a problem exists, but to do so without making comparisons that could make people feel that the American way is inferior or that the proposed approach may not be a uniquely American approach. Dr. Lurie added that there is a great deal to learn from less developed countries as well as developed countries.
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 39 Framing issues related to health disparities is extremely hard to do in this country, added Dr. Murray, because when you focus on health out- comes, you get one of four apologies. The first apology is that the United States is more diverse than other countries, as if diversity is a sort of scourge that makes it impossible for all residents to be healthy. Even if this argument were taken seriously, it cannot explain why the trends are not very good for the United States. The second and third apologies are that the health problems are caused by HIV or homicide, but it can be easily demonstrated that this is not true. The fourth apology is that it is a lack of insurance. Although that is a component of the problem, the entire issue is much more complex. The solution to framing issues related to health disparities is to shift from benchmarking health problems to benchmarking the coverage or by tracking care, continued Dr. Murray. This causes the argument to shift from saying that the United States has really bad outcomes and obesity is getting worse, for example, to determining what fraction of Americans or Â Missourians or people in the Mississippi delta are getting appropriate management of their diabetes. What fractions of those people are receiving appropriate interventions focused on diet or physical activity? If analyses show what is happening and provide comparisons from other settings, either nationally or internationally, it is very difficult to shift the onus of responsibility for those types of performance measures to somebody else. Ms. Glover Blackwell agreed that it is important to find the appropriate language because the way in which disparities are discussed will determine whether or not there will ever be the political and public will to be able to eliminate disparities. It is also important to try to identify what is making a difference, what is working. At PolicyLink, she explained, we invest in learning how to frame something and how to talk about and understand why it is important to invest in framing. We rarely think that, once we have figured out what needs to happen, there is a need to go out and start a new initiative, because there are so many programs out there already. Yet if we could determine which programs are truly the most successful, we could lift up what works. After lifting up what works, we could determine the ele- ments that make it work. Once the successful elements are identified, those elements could be infused into policy so that the original programs can be expanded and copied. Framing is absolutely important. She went on: it is also important to join with other people in this coun- try who are committed to trying to transform society so that everybody can participate and everybody can prosper. This is a movement, which is com- prised of many people working toward similar goals. Some of these people are in politics or working on housing issues. Some of them are working in environmental health, and some of them are in the workforce. We have to figure out how to join all of these people together.
40 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES When legitimate discussions about some of these difficult issues do not take place, the issues sometimes get obscured because ideology keeps people who have vested interests from making necessary changes, com- mented Dr.Â Suggs. It makes it impossible to have the kind of honest and open discussions that will be necessary to address issues related to health disparities. It is, after all, a very daunting problem that needs a more aggres- sive kind of approach. Ms. Boyce spoke about the power of words. She said: there are con- cepts I cannot abide and one of them is evidence-based practice. If we knew what evidence was going to work, we would not have health disparities. Evidence-based practice has been used to exclude community-based agen- cies from funding. Another concept that should be changed is data-driven decision making, when we know that the existing available data are faulty and that individuals are looking at the data disproportionately because of disparities. The words that are being used to describe the system do not match what is really going on in the communities. In addition, existing policies do not match what needs to happen in communities, continued Ms. Boyce. People are talking about paradigm shifts, and we keep searching for the words to define something different that needs to happen. Dr. Murray said that there needs to be a large pool of money to fund innovation and that only then will the system help us legitimize what works and what will make a difference. But can new words or concepts be coined that will better mirror what needs to happen? OtherÂ wise, we are going to keep using words like âevidence-based practiceâ and the people who need to be funded will never get the money, because the words that are used for awarding grants and justifying that a program is successful do not fit reality. Most of us were motivated to come here because we view access to quality health care as a human right, commented Dr. Rhee. The health movement that we are talking about here today mirrors the history of the Civil Rights Movement. As I reflect on my own experience as a physician and medical director and the language that I have been trained to use, I realize that the language I use gives me a lot of power in my community, Dr. Rhee continued. In the world of clinical care, we talk much more about survival, rather than viability. We focus on disease and not wellness. Our emphasis is on immediate gratification or using pills to fix things. We do not necessarily track many of the value kind of outcomes that are really important. It is important to recognize that language is a major part of the power that we as health care providers wield, continued Dr. Rhee. Nearly 20 per- cent of gross national product will soon be devoted to health care. Yet the focus of medicine has been on the bench side or bedside, rather than on the curbside (in the community). Ultimately, when you are talking about health
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 41 disparities improvements, it really is about the curbside interventions and whether or not they work. Framing Issues About Race and Institutional Racism The topics of racism and institutional racism spurred a great deal of discussion and debate. It is challenging to find a way to talk about some- thing that causes so many people to recoil, explained Ms. Glover Blackwell. However, race problems will not be solved if we do not talk about them. There must also be recognition that this is a charged discussion that cannot be approached in a way that isolates, accuses, or causes people to want to stay away. The challenge is finding appropriate language to frame the dis- cussion, while also understanding that we have to call it what it is. Dr. Rhee suggested that the terminology used to discuss these issues must strike a balance. The term âhealth disparitiesâ might not resonate with the public, but terms such as âracismâ or âinstitutionalized racismâ can be very powerful. The language that is used must be forceful and spe- cific, yet it should not cause people to disengage or make them unwilling to join in the discussion to find resolutions. There is great opposition to changing the status quo, stated Dr. Suggs. Racism is not going to be eliminated using the kind of arguments presented today. Some people are increasingly marginalized on the basis of race or social class and the price that society has to pay for that is enormous. Racial disparities are a disgrace, but they are also enormously expensive for the country. Our discussions should not simply focus on the injustice that health disparities cause, but also consider that it is terribly inefficient and costly if large segments of the population are ignored. Ms. Wright shared her belief that the message about institutional rac- ism must be targeted to specific audiences, because the variations between audiences can be quite significant. There are multiple approaches to having a constructive dialogue and to bring more people into the conversation in a meaningful way. Ms. Schwartz noted that all races and ethnicity are going to have to pay attention to this problem because the demographics in this country are changing so rapidly. It used to be that New Jersey was the most diverse state in the country; there was parity in the ratio of African Americans and Hispanics. Today New Jersey is losing citizens, and the only source of new residents is through immigration. These trends are going to drive all of these issues and attention should be paid, because these issues are going to impact everyone. Dr. Bracho shared her belief that it would be dangerous to talk about geography without highlighting issues regarding poverty and race. Clini- cal workers and public health professionals must talk about these issues
42 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES and develop solutions for reducing disparities. The link between poverty and disparities, and racism and disparities has been established. Yet little is being done because health disparities are not on the national political agenda. Dr. Suggs added that there are some issues that need to be discussed regarding the relationship between blacks and Hispanics. Had blacks and poor whites been able to come together in the 1960s and form an effec- tive political coalition, the course of this nation may have changed. Today there is a similar opportunity. If Hispanics and blacks, the largest growing minority and the established minority in this society, respectively, could join together and form an effective coalition focused on addressing the problems of their collective communities, their political clout would be enormous and effective change might be realized. A member of the audience responded that there have been times during this countryâs history when disparate groups have successfully joined forces for the greater good, and these efforts significantly changed the political scene both at the time and into the future. People should look for lessons in the annals of history. It is also very important to begin to tie similar issues together, so that people do not think in terms of one isolated problem. Especially in health care, people should merge issues together and work to see the connections between disparate problems. Data Concerns Many of the workshop participants expressed concerns about issues related to data accessibility. According to Dr. Murray, his analysis would have gone beyond 2001; however, NCHS stopped releasing data from sub- sequent years, citing privacy concerns. Yet, continued Dr. Murray, because the NCHS data originate from death certificates that are in the public domain at the local level, this seems peculiar. Since NCHS collects and tabulates data from documents that are in the public domain, it stands to reason that these data should be available to the public. Despite multiple requests to NCHS, the county-level death files have not been made avail- able, stated Dr. Murray. He surmises that it will require pressure on NCHS from policy makers in Washington in order to make the agency reverse its current policy. In any event, he concluded, it is impossible to continue monitoring disparities at the local level unless this policy is reversed. Dr. Acevedo-Garcia agreed that the lack of data makes her work more challenging and expressed frustration that county data on mortality were no longer available. The county-level data are essential for tracking dispari- ties in the United States, she explained. Although disparities are apparent in the metropolitan-area data, no health surveys are specifically representa- tive of people living in metropolitan areas. Similarly, surveys done at the
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 43 county, state, and national level are not structured in a way that adequately captures data on health disparity, opportunity, or inequality by geographic region. Dr. Acevedo-Garcia believes that her greatest challenge will be to con- duct simulation work. Her group is planning to combine empirical esti- mates of neighborhood effects on health with analyses of census data to try to simulate what the impact of policy changes would be on metropoli- tan areas. These analyses, based on estimates on neighborhood effects on health, would look at such issues as the availability of rental housing to see how this could impact residential segregation and, in turn, how this change would affect some child health outcomes. This kind of simulation work is very hard to do, she said, because it is based on quite a few assumptions. Yet it is important because the information in the current data sets does not provide the data necessary to simulate the health effects of neighborhoods and segregation. Her group wants health data sets that are representative of metropolitan areas and have information about neighborhoods, because without them it is possible to lose track of the real issues that are shaping the unequal opportunity structures. Dr. Bracho also had concerns regarding data collection. Her group, Latino Health Access, collects data from census tracts and from communi- ties. She argues that these data need to be revised for public health environ- ments, so that health disparities affecting young communities, which may not be evident in life expectancy data, can be identified with data that use smaller numbers of cases as the unit of analysis. There is a need for local comparisons to evaluate school performance, public safety, and environ- mental indicators such as open space on various concentrations of dispari- ties. Those are the data that are useful to advance interventions. Statistics do not always give a clear indication of what is really h Â appening, continued Dr. Bracho. If someone analyzed the statistics for Orange County, California, for example, they would find that the life expectancy for Hispanics would seem quite high since it is a very young community. Yet the disparities are there if you know where to look. Accord- ing to the census tract, only 3 percent of the senior citizens in the county do not have health insurance. If you segment the population and look specifi- cally at Hispanics, however, you would find that 56 percent of Hispanic elders do not have health insurance. Collaborations and Community Innovation It is important to recognize that many of these problems involve more than one sector, stressed Dr. Acevedo-Garcia. She works with the housing policy community and the public health community in metropolitan areas, but these groups rarely participate at the same meetings. Although there
44 CHALLENGES AND SUCCESSES IN REDUCING HEALTH DISPARITIES is a great deal of discussion about how difficult those collaborations are, she is aware of few incentives to collaborate with other groups or sectors to reduce disparities. Any community innovation template should have at least eight mini- mum characteristics, said Mr. Dotson. It needs to be multidimensional rather than focused on one issue, and it must be accessible, affordable, and available. We need to think of local implementation and local control as part of that innovative template. The final two characteristics needed are constancyâCan the community in which we are trying to implement this innovation depend on the program being there at a certain period of time on a regular basis?âand sustainability. Addressing Health Disparities in St. Louis The St. Louis Health Department has been advocating for a coordi- nated comprehensive approach to reducing health disparities for nearly a decade, said workshop participant William Dotson. In the next few months, the city will release a report, Public Health: Understanding Our Needs, the third in a series of biennial reports. This series of reports provides commu- nity needs assessments examining 64 variables categorized by demographic and socioeconomic factors and issues related to access and equality, racial polarization, epidemics, environmental issues, and injury behavior related to mortality. People use this report to gain a deeper understanding of the challenges of reducing health disparities and as a guide for writing grants and advocating for new programs. Another effort by the city of St. Louis was the Racial and Ethnic Approaches to Community Health (REACH) 2010 project, which spon- sored a community program targeting heart disease prevention. Despite developing innovative programs, establishing strong community partner- ships, and countless hours of planning and hard work, St. Louis was not awarded one of the demonstration projects. There was an effort to continue the program with support from local foundations and private entities, but adequate funding did not materialize. Ultimately the decision has to be made whether or not to continue a program at a lower funding level, know- ing that the reduction in funding could ultimately compromise the integrity of the original effort. This is a very hard choice to make. Over the past three years there have been some changes in the city of St.Â Louis. Life expectancy has gone up, mortality from HIV/AIDS has declined, and more women have taken advantage of first trimester care. There have been improvements, but they are small. St Louis has a long way to go in terms of organizing in order to increase the momentum toward reaching set goals. The challenge now is focused on providing, implementing, and creating momentum for solutions that will address health disparities.
THE IMPACT OF GEOGRAPHY ON HEALTH DISPARITIES 45 A member of the audience representing the St. Louis Health Commis- sion, an organization created with a mission of increasing access, improving health outcomes, and reducing health disparities within the public safety net system, shared some information from their recent reports examining issues that affect communities in the city of St. Louis. Among their findings, the commission found that primary care access in the public safety net system has increased by 13 percent in the past three to four years due to regional collaborative efforts, and there has been an 85 percent reduction in the time that people have to wait for specialty care. References Acevedo-Garcia, D., N. McArdle, T. L. Osypuk, B. Lefkowitz, and B. K. Krimgold. 2007. Children left behind: How metropolitan areas are failing Americaâs children. Diversity Report Number 1. Harvard School of Public Health, and Center for the Advancement of Health, January 2007. Boston, MA. http://diversitydata.sph.harvard.edu/children_left_ behind_Âfinal_report.pdf (accessed July 11, 2007). BRFSS (Behavioral Risk Factor Surveillance System). 2008. About the BRFSS. http://www.cdc. gov/brfss/about.htm (accessed March 19, 2008). Brooks-Gunn, J., G. J. Duncan, and J. L. Aber. 1997. Neighborhood poverty. New York: Russell Sage Foundation. IOM (Institute of Medicine). 2000. From neurons to neighborhoods: The science of early childhood development. Washington, DC: National Academy Press. IOM. 2002. Care without coverage: Too little, too late. Washington, DC: The National Academies Press. Murray, C. J. L., S. C. Kulkarni, C. Michaud, N. Tomijima, M. T. Bulzacchelli, T. J. Iandiorio, and M. Ezzati. 2006. Eight Americas: Investigating mortality disparities across races, counties, and race-counties in the United States. PLOS Medicine 3(9):1â12. NCHS (National Center for Health Statistics). 2008. National Health and Nutrition Exami- nation Survey. http://www.cdc.gov/nchs/data/factsheets/nhanes.pdf (accessed March 19, 2008). WHO (World Health Organization). 2002. The world health report 2002: Reducing risks, promoting healthy life. Geneva, Switzerland: World Health Organization. http://www. who.int/whr/2002/en/whr02_en.pdf (accessed July 11, 2006).