Data from numerous sources show that some racial, ethnic, socioeconomic, and geographic groups suffer disproportionately from preventable or treatable chronic conditions. Assessing disparities in health care provided to U.S. residents at greatest risk of debilitating and costly illnesses enables policymakers to pinpoint areas where interventions can make the greatest overall difference in reducing national health care disparities.
The consequences of health care disparities for these groups reverberate throughout the society by lowering national health indicators and contributing to escalating health care costs. Moreover, characteristics and conditions of people within these populations are likely to overlap in ways that exacerbate disparities.
One goal of this report is to increase understanding of the independent effects of demographic characteristics on health care. While substantial work has been done on the impact of ethnicity and race on health care disparities, quality, and outcomes,i much less is known about the independent consequences of varying levels of income and education, as well as geographic location. In this chapter, the NHDR analyzes the role of socioeconomic position and geographic location—separate from racial and ethnic factors—on health care disparities within and across priority populations (e.g., comparing low-income women with high-income women). Disentangling these complex causal factors provides the fact-based foundation that policymakers need to devise and evaluate health care policies and programs. This chapter provides additional perspective on which factors are most important to implement effective programs to eliminate inappropriate disparities in care.
The statutory mandate for this report refers to "priority populations," a term that is defined in AHRQ's authorizing statute (section 901 (c) of the Public Health Service Act) as encompassing both specific population groups as well as geographically defined groups. In accordance with these guidelines, the NHDR includes data and analysis on the following:
i Institute of Medicine of the National Academies, Unequal Treatment.
ii Income categories for each of the populations described in this chapter are as follows: "poor" is defined as persons with family incomes less than 100% of the Federal poverty threshold; "near poor" is defined as persons with family incomes between 100% and 199% of the Federal poverty threshold; "middle income" is defined as persons with family incomes between 200% and 399% of the Federal poverty threshold; "high income" is defined as persons with family incomes of 400% or more of the Federal poverty threshold.
Rural populations, mentioned elsewhere in the authorizing legislation, are also included. Although other demographic groups may also suffer from health care disparities, they are not considered in this report. For each group, selected findings are highlighted for quality of and access to care.
The NHDR uses categories are compliant with OMB standards for reporting Federal statistics.1 These standards consider race and Hispanic origin to be two separate and distinct concepts. Racial categories are: white, black, Asian, Native Hawaiian or Other Pacific Islander (NHOPI), American Indian or Alaska Native (AIAN), and more than one race. Because Hispanics can be of any race, all racial categories can include persons of Hispanic and non-Hispanic origin. Ethnic categories are: Hispanic or Latino, non-Hispanic white (i.e., whites who do not list their ethnicity as Hispanic) and non-Hispanic black (i.e., blacks who do not list their ethnicity as Hispanic).
Both the reauthorization legislationi and the IOM guidance for the NHDRii direct AHRQ to assess the usefulness of the data available for analysis of health care quality (including health care disparities). The forthcoming analyses of the priority populations clarify both the limitations and potentialities of current data sets.
Priority populations are already subsets of the general population. Subdivision into smaller groups for analysis by different variables, factors, or conditions often results in sample sizes that are often too small to yield statistically valid results. As noted in earlier chapters, the omission of racial, ethnic, and socioeconomic data from information gathered in health care settings limits our ability to make comparisons.
For these reasons, this chapter does not provide a comprehensive assessment of health care disparities in priority populations. Instead, it should be seen as a starting point, identifying some problem areas and indicating gaps in current data and understanding.
i U.S., Congress, S.580, Sec. 901(b)1)(F), methods for measuring quality and strategies for improving quality; Sec. 901(c)(2) Process to Ensure Appropriate Research; Sec. 902 (a)(2)quality measurement and improvement; Sec. 912 (a) Support for Efforts to Develop Information on Quality; and Sec. 914, Information Systems for Health Care Improvement.
ii Institute of Medicine of the National Academies, Guidance for the National Healthcare Disparities Report. Washington, DC: National Academies Press, 2002, pp. 13-19; 27-35.
"We are in the midst of a revolution in women's health; a revolution that has brought new attitudes, awareness and an appreciation that women have some fundamentally different health needs that need to be addressed in their own right."2 -HHS Secretary Tommy G. Thompson |
Over the past decade, women's health has become a research and policy priority. By one rough measure, a Journal of the American Medical Association editorial notes that the number of indexed clinical trials under "women and health" grew from 66 in 1992 to 276 in 1998; and in 2000, the journal received one of the highest responses ever for a call for papers on the topic.3 While life expectancy for women is approximately 6 years greater than that for men, they suffer higher morbidity rates than their male counterparts. An oft-cited maxim is that "men die quicker, but women live sicker."
Women also comprise the majority of persons needing long-term care;4 three of four nursing home residents and two of three home care consumers are female.5 Additionally, women are at greater risk for Alzheimer's disease than are men and are twice as likely as men to be affected by major depression.6
The number of women in the United States who do not have health insurance has grown three times faster than the number of men without health insurance over a 5-year period. Uninsured women are nearly 20% more likely to have difficulty obtaining health care services than uninsured men.7 The majority (roughly 3 out of 5) of low-income uninsured patients are low-income mothers.8
Health care is particularly important to women during pregnancy, a critical period when care (or lack of it) can have far-reaching consequences for both mother and child. While approximately 17% of mothers-to-be do not start prenatal care in the first trimester,9 provision of prenatal care has increased for all women over the last decade.10 Black women, however, are more likely to die from complications of childbirth than white women. These findings suggest that, while prenatal care is essential, it is likely that as yet unresolved scientific issues may contribute to excess mortality among pregnant black women. In addition, these findings underscore the additional value of examining care provided to different subgroups: identification and clarification of critical scientific issues.
This section evaluates quality of health care for women in terms of effectiveness, safety, timeliness, and patient centeredness. It also identifies which areas health care quality varies among female subpopulations.
Health care is "effective" when appropriate, proven medical services and treatments are provided to all who could benefit, and not provided to those unlikely to benefit. In general, disparities observed in the general population were echoed in the female population. For example, few differences were observed in measures relating to chronic renal disease, heart disease, and frailty in old age.
The following discussion of health care effectiveness for women provides an overview of treatment patterns that can help identify data needs and suggest additional areas for future research.
Since the passage of the 2000 Breast and Cervical Cancer Prevention and Treatment Act, 40 States have begun to cover breast and cervical cancer treatment through Medicaid for uninsured women under age 65.11
In general, minority women and women of lower socioeconomic position are less likely to receive many cancer screening services. One exception may be cervical cancer screening among black women. Specifically, blacks (84%) are more likely than whites (82%) or Asians (66%) to report receipt of Pap smears in the past 3 years. Hispanic women (77%) are less likely than non-Hispanic whites (83%), and poori (73%), near poorii (75%), and middle incomeiii (81%) women are less likely than high incomeiv women (88%) to report recent Pap smears. Similarly, high school dropouts (74%) and high school graduates (81%) are less likely than women with any college education (86%) to report Pap smears in the past 3 years. As in the general population, black, Hispanic, and women of lower socioeconomic position are less likely to receive screening for colorectal cancer (go to Chapter 3) (NHIS, 2000).
i "Poor" is defined as persons with family incomes less than 100% of the Federal poverty threshold.
ii "Near poor" is defined as persons with family incomes between 100% and 199% of the Federal poverty threshold.
iii "Middle income" is defined as persons with family incomes between 200% and 399% of the Federal poverty threshold.
iv "High income" is defined as persons with family incomes of 400% or more of the Federal poverty threshold.
Women appear to fare better than men in diabetes care. While rates of retinal screening represent substantial improvements overall in the past few years, within the general population, the poor report annual retinal eye examinations (63%) less frequently than those with high incomes (74%). Among women, the differences are not significant (go to Chapter 3) (MEPS, 2000). Similarly, although black and Hispanic diabetics in the general population are less likely to receive influenza vaccination, these differences are greatly reduced among women.
In general, disparities related to HIV/AIDS are larger among women than among men or the general population. Overall, the rate of new AIDS cases is higher non-Hispanic blacks, Hispanics, and AI/ANs compared with non-Hispanic whites. These racial and ethnic disparities are also observed among women, but relative effect sizes tend to be larger. Among women age 13 and over, new AIDS cases are reported at a higher rate among non-Hispanic blacks (48 per 100,000 population), Hispanics (12 per 100,000), and AI/ANs (7 per 100,000), compared with non-Hispanic whites (2 per 100,000) (HIV/AIDS Surveillance System, 2000).
HIV/AIDS death rates in the general population are higher among blacks and patients with lower education. Among women, HIV/AIDS death rates are also higher among blacks (13 per 100,000 population) compared with whites (1 per 100,000) and among high school dropouts (13 per 100,000) and high school graduates (7 per 100,000) compared with persons with any college education (1 per 100,000), but again relative effect sizes are larger.
Black (74%), NHOPI (77%), and AI/AN (69%) women are less likely to receive prenatal care during the first trimester than white women (85%). Similarly, Hispanics (75%) are less likely than non-Hispanic whites (89%), while high school dropouts (70%) and high school graduates (83%) are less likely than women any college education (92%) to initiate early prenatal care (go to Chapter 3). In addition, minority and less educated mothers are more likely to die from obstetrical complications (NVSS, 2000).
Suicide rates are higher for female high school dropouts (6.5 per 100,000 population) and female high school graduates (7.3 per 100,000) compared with females with any college education (4.8 per 100,000) (NVSS-M, 2000).
The percent of male nursing home residents who ever received pneumococcal pneumonia vaccination is similar between blacks (30%) and whites (32%). However, among women, this measure is significantly lower among blacks (21%) compared with whites (36%) (NNHS, 1999). Sample sizes only allow for black-white comparisons.
The prime directive of medical care is to do no harm, but the Institute of Medicine report, To Err is Human, estimated that 44,000 to 98,000 Americans die each year as a result of medical errors, making it the eighth leading cause of death.12 This report also estimated costs attributable to medical errors totaling $29 billion annually. Two aspects of the patient safety among women are included in this section: inpatient safety and medication safety.
Minorities and residents of lower income ZIP Codes tend to have lower rates of obstetric trauma. For example, rates of serious lacerations or other obstetric trauma during instrument-assisted deliveries are lower among non-Hispanic blacks (193 per 1000 instrument-assisted deliveries) and Hispanics (200 per 1000) compared with non-Hispanic whites (235 per 1000). The rates are also lower among women who live in poor areasi (183 per 1000 instrument-assisted deliveries) and near-poor areasii (207 per 1000), compared with women who live in high-income areasiii (238 per 1000) (go to Chapter 3) (HCUP SID 16-State database, 2000). Differential rates of obstetric trauma may be due to differential rates of episiotomy.13
Among other inpatient safety indicators differences among women typically parallel variation in the general population. However, one exception is noted: among all hospitalized patients and among hospitalized men, rates of postoperative abdominal wound dehiscence are not significantly different between non-Hispanic blacks and non-Hispanic whites. Among women, this measure is higher among non-Hispanic blacks (1.3 per 1,000 abdominopelvic surgery discharges) compared with non-Hispanic whites (0.6 per 1,000).
In the general population, as well as among women, the percentage of patients whose provider does not usually ask about medications and treatments from other doctors is higher among blacks compared with whites and among Hispanics compared with non-Hispanic whites (go to Chapter 4). In addition, this measure is also higher among API women (MEPS, 1999).
i Poor areas" are defined as having ZIP Codes with median incomes of under $25,000.
ii "Near-poor areas" are defined as having ZIP Codes with median incomes of $25,000-$34,999.
iii "High-income areas" are defined as having ZIP Codes with median incomes of $45,000 and higher.
Health care cannot prevent death and disability if it is delivered too late, and delays can lead to complications that ultimately increase health care costs. In general, trends of disparities in measures of timeliness are similar between women and the population as a whole.
This section considers the health care system's obligation to provide care that is respectful of and responsive to individual patient preferences, needs, and values, and to ensure that patient values guide all clinical decisions. Health care recommendations must be personalized to ensure that they are concordant with patient values, properly understood, and followed. Measures of patient centeredness overlap with the concepts of patient-provider communication and patient-provider relationship discussed under Access to Health Care, below. In general, disparities in measures of patient centeredness are similar between women and men.
This section focuses on factors that relate to initial contact with the health care system that aid or hinder entry into the system. These measures cover health insurance coverage, usual source of care and unmet need.
Racial, ethnic, and socioeconomic disparities in health insurance coverage among women are similar to differences within the general population. Among women, blacks (28%) are more likely than whites (23%), and Hispanics (45%) more likely than non-Hispanic whites (21%) to spend any time during the year being uninsured. As with the general population, the probability of being uninsured follows an income gradient, such that poor (43%), near poor (40%), and middle income (23%) women are more likely to have spent time being uninsured than high income (11%) women. Female high school dropouts are considerably more likely (43%) to spend time being uninsured than those who attended at least some college (17%). Again mirroring general population trends, AI/AN (34%) and black women (30%) are considerably more likely than white women (13%) to receive public coverage, and Hispanic women (27%) are nearly twice as likely as non-Hispanic white women (14%) to have spent time being publicly insured. Finally, women with fewer than 12 years of education are more than 7 times as likely to have public insurance at any point during the year as those with more than 12 years (29% vs. 4%, respectively) (MEPS, 1999).
In the overall population, racial, ethnic, and SES disparities exist in access to a usual source of care. Similarly, female Hispanics (82%, vs. 92% of non-Hispanic whites) and poor (84%), near poor (85%), and middle income women (92%, vs. 95% of high income women) are less likely to have a usual source of care (NHIS, 2000).
Money is often a constraint. Women are more likely than men to cite financial reasons for having no usual source of care, and priority populations represent a disproportionate amount of these respondents. Specifically, Hispanic women (23%, vs. 14% of non-Hispanic white women), poor women and near poor women (22% and 23%, respectively, vs. 8% of high income women), and female high school dropouts (23%, vs. 11% of women with at least some college) report that financial limitations prevent them from having a usual source of care (MEPS, 1999).
Women and men are equally likely to rely upon a hospital, emergency room, or clinic as a usual source of care. Within the female population, however, differences exist: AI/ANs (47%) and blacks (26%) are more likely to depend upon these institutional sources of care than are whites (15%). Similarly, Hispanic women (25%) are more likely than non-Hispanic white women (14%), and poor, near poor, and middle income women (32%, 24%, and 16%, respectively) are more likely than high income women (11%) to depend upon these delivery sites for their primary care (NHIS, 2000).
Female respondents who are concerned about their ability to receive medical care if they should need it are frequently members of priority populations. For example, Hispanics (38%) are more likely than non-Hispanic whites (26%), poor (34%), near poor (33%), and middle income women (26%) are more likely than those with high income (20%), and women who did not finish high school (33%) are more likely than those who completed at least some college (24%) to be dissatisfied that they can obtain health care if they should need it (MEPS, 1999).
This dissatisfaction and concern is often borne out by experience. The following groups of women are less likely to "always obtain care for an illness or injury as soon as desired": APIs (36%, vs. 54% of whites), Hispanics (39% vs. 55% of non-Hispanic whites), and the poor (47%, vs. 55% of those with high incomes) (MEPS, 2000). Similarly, the following groups of women are less likely to always be able to schedule appointments for routine care as soon as they wanted: APIs (29%, vs. 42% of whites), Hispanics (34%, vs. 43% of non-Hispanic whites), 48% of the poor (compared with 41% of those with high incomes), and 44% of those with less than a high school education (compared with 40% of college attendees).
Access to off-hour care is restricted for many women. For example, poor women (43%, vs. 52% of high income women) and high school dropouts (39%, vs. 45% of women who attended college) are less likely to have the option of scheduling off-peak appointments. Interestingly, API women are actually more likely than white women (55% vs. 49%) to frequent clinicians that offer flexibility in scheduling off-peak appointments. Yet despite their ability to schedule appointments, API women tend to be less satisfied with the professional staff at their usual source of care's office (36%) than white women (25%) (MEPS, 1999).
Another important measure of getting care is the ability to see specialists. Hispanic women (69%, vs. 83% of non-Hispanic white women) and poor women (70%, vs. 84% of high income women) are less likely to report that they had no problems obtaining referrals to specialists (MEPS, 2000).
Once at the health care delivery site, certain groups of women are more likely to have to wait over 30 minutes before being seen by a clinician. Women who are black (20%, vs. 15% of whites), Hispanic (25%, vs. 14% of non-Hispanic whites), poor (25%, vs. 12% of those with high incomes), and who did not complete high school (23%, vs. 12% of college attendees) are more likely to experience long waits to see their health care provider (MEPS, 1999).
Even more pressing, black women are more likely both to wait over 1 hour for emergent/urgent care (15%, vs. 12% of white women) and for semi-urgent/non-urgent care (31%, vs. 24% of white women) in hospital emergency departments (NHAMCS-ED, 1999-2000).
Among women, some groups are more likely than others to report having poor communication with their health care provider at their last visit—specifically: women who are black (24%) and Asian (28%, compared with 18% of whites), Hispanic (29%, compared with 17% of non-Hispanic whites), poor (29%, compared with 15% of high income women), and have less than a high school education (30%, compared with 18% of college attendees) report problematic communication at their last doctor's visit (Commonwealth Fund Health Care Quality Survey, 2001).
Men and women have similar rates of satisfaction with care and similar ratings of their health care. However, racial, ethnic, and socioeconomic disparities among women exist. For example, APIs (31%, vs. 20% of whites) and Hispanics (24%, vs. 19% of non-Hispanic whites) are more likely to report dissatisfaction with the quality of care their provider offers (MEPS, 1999).
These same female subpopulations also report low levels of satisfaction with their overall health care. Specifically, among women, APIs (31%) and blacks (20%, vs. 16% of whites), Hispanics (19%, vs. 15% of non-Hispanic whites), the poor (25%, vs. 15% of those with high-incomes), and high school dropouts (21%, compared with 15% of women who attended college) reported lower satisfaction with care (MEPS, 2000).
Cultural values and ideas about what constitutes good health can vary between demographic groups. "Cultural competency" implies an awareness of health beliefs and behaviors, disease prevention and incidence, and treatment outcomes for different populations.14 It is, however, important to note that we have much to learn about how to operationalize cultural competency. Echoing findings in the general population, women who are black (17%), Asian (13%, compared with 2% of whites), Hispanic (16%, vs. 1% of non-Hispanic whites), and poor (11%, vs. 3% of those with high incomes) are more likely to believe that they would have received better care if they were a different race or ethnicity. Furthermore, high school dropouts (33%) are less likely than women who attended college (41%) to report that their doctors understand their background and values (Commonwealth Fund Health Care Quality Survey, 2001).
Women are more likely than men to understand and use health information. For example, women report that prescription medication instructions are difficult to understand less often (18%) than do men (23%). Consistent with findings in the general population, women who are Asian (39%, compared with 17% of white women) and Hispanic (35%, vs. 15% of non-Hispanic white women) are more likely to find prescription bottle information difficult to understand (Commonwealth Fund Health Care Quality Survey, 2001).
Due to sample size constraints, both data sources used for chronic care measures only provide black-white comparisons. In general, disparities in chronic care do not differ between women and men (NHHCS, 2000 and NNHS, 1999).
Mirroring trends in the general population, minorities tend to be less likely to receive mental health treatment or counseling, especially outpatient treatment. For example, black women (9%) and Hispanics (7%) are less likely than whites (15%) and non-Hispanic whites (16%) to have received mental health treatment or counseling in the past year (NHSDA, 2001). These data are particularly important given the fact that depression is the second leading cause of hospitalizations for women of reproductive age (HCUP Chartbook). Data on substance abuse treatment do not allow assessment of disparities among women.
Frequency of inpatient hospitalization of female HIV patients mirrors trends seen in the general HIV population: non-Hispanic black women (20%) and Hispanic women (19%) more frequently receive inpatient HIV care than non-Hispanic white women (13%) (HIV Research Network, 2000).
Avoidable admissions (also called "admissions of ambulatory care sensitive conditions") are hospitalizations for health conditions that, in the presence of comprehensive primary care, rarely require hospitalization. High rates of avoidable admissions suggest impaired access to or quality of ambulatory care.
Similar to the general population, non-Hispanic black women typically have higher rates of avoidable admissions, while API women typically have lower rates. One exception may be in admissions for chronic obstructive pulmonary disease, where rates are higher among non-Hispanic black men (345 per 100,000 population 18 and over) compared with non-Hispanic white men (264 per 100,000). In contrast, rates are lower among non-Hispanic black women (231 per 100,000 population 18 and over) compared with non-Hispanic white women (269 per 100,000).
The health and well-being of children has long been a research and policy priority in the United States and worldwide.15, 16 Children are recognized as deserving of special attention for at least three reasons: childhood is a unique developmental stage of life, child health must provide continuity with adult health, and the child health care system is distinctive.17
Childhood is a unique period of life. At least four distinguishing characteristics ("the 4 D's") have important implications for health care for children:
Reflecting these differences, a unique set of financing and organizational arrangements have evolved for children in the U.S. health care system. Part of society's obligation to children has been expressed in the special attention paid to children for public payment of their health care. For example, poor children are a mandatory population for the Medicaid program, and adolescents were added gradually over time as optional populations. In 1997, the State Children's Health Insurance Program (SCHIP) was passed to provide coverage to certain low-income children who were not eligible for Medicaid and otherwise not insured. The Maternal and Child Health Block Grant program and other public programs provide direct financing of some services for children. Families of some disabled children also can receive cash benefits from Social Security to help pay health care and related costs that otherwise remain uncovered.
This chapter summarizes NHDR findings on children in general. It also focuses on disparities within the child population for racial and ethnic minority children and children of different income levels.i Disparities affecting children with special health care needs are addressed in the chapter on individuals with special health care needs. Because many child-specific health care needs and services are not explicitly addressed and sample size limitations exist, these findings should not be interpreted as a comprehensive assessment of disparities in health care among children.
i Children are defined differently in different measurement sets. Unless otherwise noted, children are defined in the NHDR as they are defined in the MEPS and the NHIS: 0 through 17 years old.
This section evaluates quality of health care for women in terms of effectiveness, safety, timeliness, and patient centeredness. It also identifies which areas health care quality varies among female subpopulations.
Not all measures of effectiveness included in the quality of care measure set are applicable to children; some are applicable because children suffer from the disorders but have not been applied to children or to racial and ethnic minority children or children of different socioeconomic statuses. One reason that more quality of care measures have not been developed for children is the limited evidence on whether or not specific child health care interventions are effective. Eight aspects of the effectiveness of child health care are included in this section:
Among adolescents ages 13 to 17, new AIDS cases are being reported at a higher rate among non-Hispanic black (3.5 per 100,000 population) and Hispanic (1.0 per 100,000) adolescents compared with non-Hispanic white adolescents (0.2 per 100,000) (CDC HIV/AIDS Surveillance System, 2000).
Among children, HIV infection death rates are higher among black children (0.5 per 100,000) compared with white children (0.1 per 100,000) (NVSS-M, 2000).
Among adolescent mothers under 15, prenatal care is less often initiated by blacks (41%) and AI/ANs (43%) compared with whites (51%) but is similar between Hispanics (49%) and non-Hispanic whites (53%) (NVSS, 2000).
Blacks are less likely to receive childhood immunizations compared with whites, and lower income children are less likely to receive immunizations compared with more affluent children. For example, receipt of all recommended vaccinations is achieved by fewer black children (68%) than white children (75%) and by fewer children in poor (68%), near poor (71%), and middle income (75%) families, compared with children in high income families (79%) (go to Chapter 3). In contrast, receipt of the varicella vaccine is achieved more often by Asian (82%) and Hispanic (80%) children than by white (76%) and non-Hispanic white (75%) children, respectively (NIS, 2001). There is no evidence of disparity for adolescent immunizations (NHIS).
Black children receive fewer dental visits (32%) than white children (50%), and Hispanic children receive fewer (33%) than non-Hispanic white children (54%). Additionally, children who are poor (30%), near poor (36%), and middle income (50%) are less likely than high income children (63%) to see a dentist. Finally, uninsured children (24%) have fewer dental visits than privately insured children (54%) (MEPS, 1999).
Pediatric gastroenteritis, one of the most common reasons for pediatric hospitalizations, can be treated on an outpatient basis, and proper outpatient treatment can reduce admissions for gastroenteritis among children. Rates of hospitalization for pediatric gastroenteritis are higher among Hispanic children (133 per 100,000 population) and lower among API children (47 per 100,000) compared with white children (115 per 100,000). Because information on patient income and education is not available, median income of patient's ZIP Code is used as a crude proxy of socioeconomic position. Hospitalization rates are higher among children who live in poor areasi (172 per 100,000 population), near-poor areasii (157 per 100,000), and medium-income areasiii (124 per 100,000) compared with children who live in high-income areasiv (86 per 100,000) (HCUP SID, 2000).
Among children, rates of suicide are lower among black (1.0 per 100,000 population) and API (0.8 per 100,000) children than white children (1.7 per 100,000) and among Hispanic children (0.9 per 100,000 population) than non-Hispanic white children (1.8 per 100,000). However, AI/AN children have the highest rates of suicide of all groups (3.8 per 100,000 population) (NVSS-M, 2000).
Among children ages 12 to 17, substance abuse treatment is reported less frequently by non-Hispanic blacks (0.8%) compared with non-Hispanic white (1.6%) (National Household Survey on Drug Abuse, 2001)
i "Poor areas" are defined as having ZIP Codes with median incomes of under $25,000.
ii "Near-poor areas" are defined as having ZIP Codes with median incomes of $25,000-$34,999.
iii "Medium income areas" are defined as having ZIP Codes with median incomes of $35,000-$44,999.
iv "High-income areas" are defined as having ZIP Codes with median incomes of $45,000 and higher.
Data on treatment of upper respiratory infection come from NAMCS/NHAMCS in 1999-2000. Evidence of significant racial disparity is not present. Hospitalization rates among black children (60 per 10,000 population) and adults (21 per 10,000) tend to be higher than rates among white children (17 per 10,000) and adults (8 per 10,000) (NHDS, 2000).
Two aspects of the child patient safety are included in this section: inpatient safety and medication safety.
Again, because information on patient income and education is not available, median income of patient's ZIP Code is used as a crude proxy of socioeconomic position. One inpatient safety indicator is specific to children: birth trauma. Rates of birth trauma are lower among Hispanics (5.5 per 1000 live births) than non-Hispanic whites (8.2 per 1,000) and among children who live in ZIP Codes with median incomes $25,000-$34,999 (4.9 per 1,000 live births) and incomes $35,000-$44,999 (5.8 per 1,000) than children who live in ZIP Codes with median incomes $45,000 and over (8.0 per 1,000) (Figure 1) (HCUP SID, 2000).
Rates of obstetric trauma during cesarean delivery are lower among non-Hispanic black adolescents (1.2 per 1,000 cesarean deliveries) and Hispanic adolescents (1.3 per 1,000) than white adolescents (5.1 per 1,000). In addition, among children, rates of accidental puncture or laceration during procedures are higher among non-Hispanic blacks (2.5 per 1,000 relevant discharges) and Hispanics (2.3 per 1,000) compared with non-Hispanic whites (2.0 per 1,000) (Figure 2) (HCUP SID, 2000).
Figure 1. Birth trauma injuries per 1,000 selected live births
Notes:
ˆ Indicates reference group.
* p<0.05 and relative rate >10% for comparison of group with reference group.
Key: API=Asian or Pacific Islander
Source: Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project State Inpatient Databases 16-State database, 2000.
Figure 2. Accidental laceration or puncture during procedure per 1,000 discharges
Notes:
ˆ Indicates reference group.
* p<0.05 and relative rate >10% for comparison of group with reference group.
Key: API=Asian or Pacific Islander
Source: Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project State Inpatient Databases, 16-State database, 2000.
Among children, the percentage of persons with a provider who does not usually ask about medications and treatments other doctors may give is higher among blacks (86%) than whites (78%) and higher among Hispanics (85%) than non-Hispanic whites (77%) (MEPS, 1999).
Traditional measures and data sources on children's access to health care are limited. Not all traditional measures of mainstream health care (e.g., care provided in physicians' offices, hospitals, and emergency departments) are tailored to children's health care needs. Moreover, children sometimes get needed services outside the mainstream health care system, such as in schools, public specialty clinics, and from non-physician subspecialists. Mainstream efforts to collect data on children's health care rarely collect information about these sources of health care. As a result, little is known about how and where many children obtain many health services and about the extent to which there may be disparities among children.
This section focuses on factors that relate to initial contact with the health care system that aid or hinder entry into the system. These measures cover health insurance coverage, usual source of care, and unmet need.
Over 1 out of 5 children spends some time being uninsured. Hispanic children (41%) are considerably more likely than non-Hispanic white children (17%) to have been uninsured for part of the year (Figure 3). Three in 10 poor and near poor children and 19% of middle income children spend time being uninsured, compared with 10% of children in high income families. More than a quarter of all children experience public health insurance coverage for at least some period over the course of a year, and children are less likely than the general population to lack coverage either all year or at any point during the year (MEPS, 1999).
About one quarter of children, mostly poor and near poor, have public coverage any time during the year. Children who are AI/AN (55%) and black (46%) are far more likely to have public coverage than white children (21%). Hispanic children (39%) are more likely to be publicly insured than non-Hispanic children (23%). Nearly three quarters of poor children and 2 in 5 near poor children have public coverage, compared with 3% of children in high income families (MEPS, 1999).
Figure 3. Percent of children with any period of uninsurance during the year
Notes:
ˆ Indicates reference group.
* p<0.05 and relative rate >10% for comparison of group with reference group.
Key: API=Asian or Pacific Islander; AI/AN=American Indian or Alaska Native
DSU=Data do not meet the criteria for statistical reliability, data quality, or confidentiality.
Source: Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, 1999.
The American Academy of Pediatrics asserts that having a usual source of care or "medical home" is an essential element of effective pediatric care.18 About 93% of children are reported to have a usual primary care provider. However, API children (81%) are less likely to have a usual primary care provider than white children (88%). Hispanic children of any race (75%) are less likely than non-Hispanic white children (91%) to have a usual primary care provider. Finally, poor children (81%) are less likely than high income children (93%) to have a usual source of primary care (MEPS, 1999).
About 20% of children report that their source of ongoing care is a clinic, outpatient department or a hospital emergency department (Figure 4). AI/AN (49%) and black (28%) children are more likely than white (17%) children to report these settings as their source of ongoing care. Hispanic children of any race (28%) are more likely than non-Hispanic white children (15 percent) to report these facilities as their source of ongoing care, as are poor (30%) versus high-income (12%) children (NHIS, 2000).
The percent of children who can always get an appointment for routine care as soon as desired is lower among Hispanics (45%) than non-Hispanic whites (54%) and among poor (48%), near poor (45%), and middle income (52%) children compared with high income children (58%) (MEPS, 2000).
Figure 4. Percent of children who identify a hospital, emergency room, or clinic as their source of ongoing care
ˆ Indicates reference group.
* p<0.05 and relative rate >10% for comparison of group with reference group.
Key: NHOPI=Native Hawaiian or Other Pacific Islander; AI/AN=American Indian or Alaska Native
Source: National Center for Health Statistics, National Health Interview Survey, 2000.
Scheduling appointments in off-peak hours is a less frequent option for black (50%) and AI/AN (52%) children than white children (62%). Hispanic children (54%, vs. 64% of non-Hispanic white children) and poor children (52%, vs. 67% of high income children) are also less likely to have access to off-peak appointments (MEPS, 1999).
Satisfaction with professional staff at the usual source of care's office is an indicator of the ability to obtain care. Families of children who are APIs (38%) and AI/ANs (32%, vs. 24% of whites), Hispanic (29%, vs. 23% of non-Hispanic whites), and poor (29% vs. 22% of high income children) are more likely to report dissatisfaction with the professional staff at their child's usual health care delivery site (MEPS, 1999).
Among children, no racial or ethnic disparity is noted in difficulty obtaining referral to a specialist. However, poor and near poor children are more likely to report difficulty obtaining a referral compared with high income children.
Once at the doctor's office, children who are black (19%, vs. 15% of whites), Hispanic (25%, vs. 13% of non-Hispanic whites), and poor (26%, vs. 18% of near poor, 16% of middle income, and 10% of high income children) are more likely to wait more than 30 minutes to see their usual source of care. Black children (33%) are also more likely than white children (25%) to have an hour or more wait for semi-urgent or non-urgent emergency department care (MEPS, 1999). There are insufficient data to determine disparities among other groups of children for this measure and similarly inadequate data to determine if there are disparities in wait times at hospital emergency departments among different groups of children needing urgent or emergent care.
Families of black (86%, compared with 78% of white), Hispanic (85%, vs. 77% of non-Hispanic white), and poor (84%, vs. 78% of high income) children are more likely to report that their "usual source of care asks about prescription medications and treatments other doctors may give" (MEPS, 1999)
Data on overall satisfaction with health care provided to children are reported by parents or other adults responding on behalf of children. Eleven percent of parents overall rate their children's health care relatively low (<= 6 on a 0-10 scale). Black (14%) compared to white (10%) and poor (19%) compared with high income (6%) families are more likely to report dissatisfaction with their children's health care on this scale (MEPS, 2000).
Among children, many racial and ethnic minorities as well as persons of lower socioeconomic position are less likely to receive office or outpatient visits, prescription medications, and dental visits. For example, among children, blacks (58%) and APIs (62%) are less likely than whites (74%), Hispanics (61%) are less likely than non-Hispanic whites (76%), and poor children (65%) are less likely than high income children (79%) to receive an office visit (MEPS, 1999).
Significant racial disparities in referred care are not evident among children. However, visits in which the patient was referred by another physician or health plan are received less often by uninsured children (7%) than by children with private insurance (12%) (NAMCS/NHAMCS, 1999-2000).
Among children, rates of emergency department visits are higher among blacks (54 per 100 population) and lower among APIs (17 per 100) compared with whites (35 per 100) (NHAMCS-ED, 1999-2000).
Avoidable admissions (also called "admissions of ambulatory care sensitive conditions") are hospitalizations for health conditions that, in the presence of comprehensive primary care, rarely require hospitalization. High rates of avoidable admissions suggest impaired access to or quality of ambulatory care.
Among children, non-Hispanic blacks typically have higher rates of avoidable admissions while APIs typically have lower rates than non-Hispanic whites. In addition, rates of admission for bacterial pneumonia are higher among Hispanic children (200 per 100,000 population 18 and over) compared with non-Hispanic white children (147 per 100,000) (HCUP SID, 2000).
The aging of our population poses significant health care, economic, and social challenges. Studies document the growing prevalence of disease, disability, and dependency in the elderly population.19 Increased need for long-term care and treatment for chronic conditions are likely to place a strain on individuals, families, and the health care infrastructure, even without the added burden resulting from disparities associated with demographic characteristics.
About 1 in 20 Americans requires assistance with activities of daily living, and approximately 2 out of 3 disabled elderly persons live in community settings—which require timely long-term care as well as appropriate access to health care services. Today's elderly population is confronted with inadequate access to long-term care, and estimates of expected demand in the next 10 years depict a dire situation for those needing nursing home beds and home health care.
Since the 1965 enactment of the Medicare program, nearly all elderly persons have been covered by a core health insurance package. This insurance greatly reduces one significant barrier to health care—the ability to pay for acute care services. However, because Medicare does not completely eliminate cost as a barrier to health care, and other obstacles continue to affect the ability of the elderly to get timely and appropriate treatment of their health care needs, significant disparities in access to and utilization of health care among the elderly remain.
The impact of factors affecting access to and use of health care on racial, ethnic, and low income populations are well known.20 Among the elderly, the vulnerable subgroups include racial and ethnic minorities, individuals with low income, beneficiaries without supplemental insurance, and the oldest old (i.e., persons aged 85 or older). All of these groups have or are susceptible to disparities in access to and utilization of medically needed health care, particularly because many of these subgroups are growing faster than the Medicare population as a whole.21
Surveys of the general population often do not include enough elderly to permit assessment of many disparities. Therefore, the NHDR supplements its analysis with data from the Medicare Current Beneficiary Survey (MCBS, 1999). MCBS data allow comparisons of elderly beneficiaries who have Medicare and Medigap coverage with beneficiaries who have only Medicare fee-for-service coverage. The analysis presented here focuses on Medicare beneficiaries aged 65 and over who reside in the community.
This section evaluates quality of health care for the elderly in terms of effectiveness, safety, timeliness, and patient centeredness. It also identifies which areas health care quality varies among female subpopulations.
Health care is "effective" when appropriate, proven medical services and treatments are provided to all who could benefit and not provided to those unlikely to benefit. The following discussion of health care effectiveness for the elderly provides an overview of treatment patterns that can help identify data needs and suggest additional areas for future research.
As in the general population, minorities and low socioeconomic status elderly are less likely to receive many cancer screening services. However, while Hispanic women generally are less likely to receive mammograms and Pap smears compared with non-Hispanic white women, these differences disappear among seniors. The percent of elderly women who report they had a mammogram within the past two years is the same for both Hispanics and non-Hispanic whites (68%), while the percent who report they had a Pap smear in the past three years is similar for Hispanics (67%) and non-Hispanic whites (64%) (NHIS, 2000).
Within the general population, women who are black and of lower socioeconomic status tend to have higher rates of death from cancer, while API, AI/AN, and Hispanic women tend to have lower rates. Lung cancer and breast cancer death rates follow different trends among the elderly than among the population as a whole. Specifically, while blacks generally have higher mortality rates from lung and breast cancer than whites, death rates from lung cancer are not significantly different between elderly blacks (329 per 100,000 population) and elderly whites (323 per 100,000). Likewise, death rates from breast cancer among elderly black women (132 per 100,000) do not significantly differ from the rates among and elderly white women (123 per 100,000) (National Vital Statistics System-Mortality (NVSS-M), 2000).
Disparities in screening for cardiovascular risk factors differ between the general population and the elderly. While there is no evidence for disparity in cholesterol screening among the general population, elderly blacks (84%) are less likely than whites (88%) to have had their blood cholesterol checked in the preceding 5 years. There are no significant disparities in blood pressure measurement noted among the elderly (NHIS, 1998).
Data on management of congestive heart failure from the National Hospital Discharge Survey (NHDS) indicate that rates of potentially avoidable hospital admissions for congestive heart failure increase with age, and differences between blacks and whites tend to attenuate (Figure 5).
Figure 5. Hospital admissions for congestive heart failure per 100,000 population by age and race
Source: National Center for Health Statistics, National Hospital Discharge Survey, 2000.
Data on new AIDS cases indicate that disparities in HIV/AIDS are generally larger among the elderly than among younger age groups. Among the elderly, new AIDS cases are being reported at a higher rate by non-Hispanic blacks (15.1 per 100,000 population) and Hispanics (5.9 per 100,000) than by non-Hispanic whites (0.7 per 100,000), and relative rates tend to be higher than in the general population (CDC's HIV/AIDS Surveillance System, 2000).
Data on HIV-infection deaths indicate similar disparities. Overall HIV infection death rates are higher among blacks and Hispanics. Among the elderly, HIV infection death rates are also higher among blacks (9.4 per 100,000 population) compared with whites (0.7 per 100,000) and among Hispanics (2.8 per 100,000) compared with non-Hispanic whites (0.5 per 100,000); but again, relative effects sizes are larger (NVSS-M, 2000).
Research indicates that vaccination against flu among the elderly results in lower rates of hospitalization, mortality, and complications, as well as cost savings.22 However, elderly blacks and persons of lower socioeconomic position are less likely to receive influenza and pneumococcal immunization. Among persons 65 and over, receipt of influenza vaccine is less often reported by blacks (48%) than by whites (66%), by the poor (56%) and near poor (61%) than by high income families (69%), and by persons with less than a high school education (58%) and high school graduates (66%) than by persons with any college education (70%) (go to Chapter 3). Among persons age 65 and over, receipt of influenza vaccine is also less often reported by Hispanics (56%) than by non-Hispanic whites (67%) (NHIS, 2000).
Data on immunization of nursing home residents from the National Nursing Home Survey (NNHS) only allow for comparisons of blacks to whites due to small sample sizes. In the total adult nursing home population, the black-white difference in the percent of residents who have ever received pneumococcal vaccination is not statistically significant. However, this measure is significantly different between blacks (29%) and whites (39%) age 75 to 84 and between blacks (32%) and whites (41%) age 85 and over (NNHS, 1999).
Data on nursing facility care from CMS's Nursing Home Resident Profile Table contain information on all current Medicare nursing home residents (~1.4 million residents). In the total nursing home population, non-Hispanic blacks and Hispanics are more likely than non-Hispanic whites to have bed sores. As the prevalence of bed sores increases with age among minorities, these differences tend to grow larger (Figure 6).
Figure 6. Percent of nursing home residents with bed sores by age and race/ethnicity
Source: Centers for Medicare & Medicaid Services Nursing Home Resident Profile Table.
In the total nursing home population, Hispanics and APIs are more likely than non-Hispanic whites to be restrained but non-Hispanic blacks are not more likely to be restrained. In contrast, among nursing home residents age 85 and older, non-Hispanic blacks, Hispanics, and APIs are all more likely to be in restraints compared with non-Hispanic whites (Figure 7).
Figure 7. Percent of nursing home residents in restraints by age and race/ethnicity
API=Asian or Pacific Islander
Source: Centers for Medicare & Medicaid Services Nursing Home Resident Profile Table.
This report has previously noted that medical errors are the eighth leading cause of death23 in the Nation. Two aspects of patient safety among the elderly are discussed here: inpatient safety and medication safety.
When disparities among hospitalized elderly are identified, they typically parallel disparities in the general population. However, several exceptions are noted. Among all hospitalized patients, rates of decubitus ulcers are higher among non-Hispanic blacks compared with non-Hispanic whites. As the frequency of these pressure ulcers increases with age, differences tend to grow larger (Figure 8). In addition, among the elderly, Hispanics also have higher rates of decubitus ulcers compared with non-Hispanic whites (HCUP SID 16-State database, 2000).
Among hospitalized patients, rates of death in low mortality DRGs are not significantly different between non-Hispanic blacks and non-Hispanic whites. Among the elderly, this measure is higher among non-Hispanic blacks (3.9 per 1,000 admissions in low mortality DRGs) compared with non-Hispanic whites (3.5 per 1,000).
Data on medication safety from MEPS indicate that the use of inappropriate medications among the elderly do not indicate significant disparity.
Figure 8. Decubitus ulcers per 1,000 selected stays of 4 or more days
Source: Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project State Inpatient Databases 16-State database, 2000.
In general, the elderly encounter fewer problems with sources of care and delays in getting care. Seniors who are minorities or of low socioeconomic status, however, do wait longer to be seen in doctor's offices and emergency rooms.
The elderly generally report problems with patient-provider relationship and communication less often than younger age groups. Disparities related to patient centeredness tend to decrease with age.
The impact of factors affecting access to and use of health care on racial, ethnic, and low income populations has been widely documented.24 Many of the vulnerable elderly groups are growing faster than the Medicare population as a whole. Between 1992 and 1998, the Medicare population grew at an average annual rate of 1.5% compared to 6.3% for Hispanics, 9.8% for non-Hispanic minorities, and 3.4% for the oldest old.25 Other barriers also continue to affect the ability of the elderly to get timely and appropriate treatment for their health care needs.
Patterns of health insurance coverage among the population age 65 and over are considerably different from patterns among the general population. Since the 1965 enactment of the Medicare program, nearly all elderly persons have been covered by a core health insurance package; in contrast, only 83% of younger persons are insured. This insurance greatly reduces one barrier to health care—the ability to pay for acute care services.
However, Medicare does not completely eliminate cost as a barrier to health care. For example, elderly patients often have out-of-pocket expenses for prescription medications and doctor bills. In response to these gaps, many elderly opt to augment their Medicare coverage with supplemental private health insurance plans.
MCBS findings confirm the presence of race/ethnic and socioeconomic disparities in access to care among the elderly. For example, the percent of the elderly with no usual source of care is higher among blacks (7.8%) compared with whites (5.1%), Hispanics (8.4%) compared with non-Hispanic whites (4.9%), poor (7.5%) and near poor (6.3%) compared with those with high incomes (4.0%), and elderly with only Medicare (17.5%) compared with those with both Medicare and Medigap coverage (4.6%).
Similarly, the percent who delay care due to cost is higher among poor (6.8%), near poor (7.3%), and middle income (2.8%) elderly than high income elderly (1.2%), as well as among the elderly with only Medicare (12.1%) compared with elderly who have both Medicare and Medigap coverage (3.0%).
Finally, the percent experiencing problems getting to the doctor is higher among poor (7.8%), near poor (5.5%), and middle income (4.0%) elderly than high income elderly (2.4%), as well as among the elderly with only Medicare (6.3%) compared with those covered by both Medicare and Medigap (3.7%) (MCBS, 1999).
While the elderly are more likely than younger age groups to have a regular or usual source of ongoing care, prominent disparities are associated with race, ethnicity, and socioeconomic position. As in the general population, elderly minorities are more likely to have a hospital, emergency room, or clinic as their source of ongoing care (21% of elderly blacks vs. 13% of elderly whites, and 21% of elderly Hispanics vs. 12% of elderly non-Hispanic whites). Utilization of these potentially suboptimal sources of care is also higher among the elderly who are poor (21%), near poor (17%), or middle income (14%) compared with the elderly who have higher incomes (10%) (NHIS, 2000).
The elderly report fewer difficulties or delays in obtaining health care than younger persons, but there are discernible socioeconomic differences. For example, elderly persons with less than a high school education (6%) are more likely to report having difficulty in obtaining care than those who have attended college (2%). Similarly, the elderly who did not complete high school (23%) and those who ended their education after high school graduation (18%) are more likely to report dissatisfaction with getting needed health care than their peers who attended college (13%) (MEPS, 1999).
Compared with younger groups, the elderly are more likely to report having difficulties obtaining referrals to specialists than younger populations. These problems are reported more frequently by racial and ethnic minorities and by the poor. Elderly blacks (76%) are less likely to report no problems obtaining specialist referrals than elderly whites (89%), and elderly Hispanics (84%) are less likely to report no problems obtaining specialist referrals than elderly non-Hispanic whites (89%). Similarly, the elderly poor (84%) are less likely to report no problems obtaining specialist referrals than the elderly with high incomes (90%) (MEPS, 2000).
There is evidence of racial/ethnic and socioeconomic differences in waiting times among the elderly. The following elderly subgroups experience longer waiting times at their usual source of care: blacks (27%) and APIs (10%) compared with whites (16%); Hispanics (31%) compared with non-Hispanic whites (16%); and poor (24%) compared with high income elderly (16%) (MEPS, 1999).
Within the elderly population, racial, ethnic, and socioeconomic disparities exist in patient-provider communication. Black seniors (27%) are more likely than white seniors (13%) and elderly Hispanics (31%) are more likely than elderly non-Hispanic whites (11%) to have had poor communication with their provider during their last visit, as indicated by one or more indicators (Commonwealth Fund Health Care Quality Survey, 2001). Similarly, the elderly with less than a high school education (23%) are more likely than the elderly with any college education (13%) to report poor communication with their provider during their last visit, as measured by one or more indicators.
On the other hand, elderly Hispanics (88%) and blacks (81%) are more likely than elderly non-Hispanic whites (76%) or whites (77%) to report that their usual sources of care inquires about treatments prescribed by other doctors (MEPS, 1999).
While the elderly tend to rate their health care higher than younger groups, there are disparities among the elderly. Disparities among elderly subgroups in rating their health care less favorably (from 0 to 6) on a '0 to 10 scale' are seen as follows: elderly blacks (20%) and APIs (16%) compared with elderly whites (10%); the elderly poor (14%) compared with those with high incomes (7%); and the elderly with less than a high school education (15%) compared with those who attended at least some college (7%) (MEPS, 2000).
Disparities also exist between race/ethnic groups when respondents are asked if they felt as involved in decisionmaking as they wanted to be. Specifically, 30% of elderly blacks compared with 14% of whites and 36% of elderly Hispanics compared with 13% of non-Hispanic whites report not being involved in decisionmaking as much as they wanted (Commonwealth Fund Health Care Quality Survey, 2001).
Data from the Commonwealth Fund Health Care Quality Survey indicate that the elderly are more likely than younger groups to agree that their doctor understood their background and values, and significant disparities associated with demographic characteristics are not apparent.
Generally, the elderly are less likely to seek out health information and to have more difficulty understanding information provided by their doctors. Disparities among the elderly are similar to disparities noted among younger groups. For example, elderly Asians (55%) have more difficulty than elderly whites (47%) in understanding health care information provided by their physicians. Similarly, 63% of the elderly with less than a high school education compared to 45% of high school graduates and 40% of those who attended college have more difficulty obtaining health care information than those with college educations (Commonwealth Fund Health Care Quality Survey, 2001).
Disparities in the receipt of health care associated with race, ethnicity, and socioeconomic status observed in the general population tend to lessen among the elderly.
The elderly are more likely to receive most types of services, and disparities in the receipt of health care observed among younger populations are reduced. Disparities in ambulatory care are similar between the elderly and younger age groups. However, the significance is often lost due to small sample sizes. One exception may be rates of outpatient visits among blacks. In the general population, rates are lower among blacks (262 per 100 population) compared with whites (333 per 100). Among the elderly, rates of outpatient visits are higher. In addition, rates are similar between blacks (614 per 100 population) and whites (643 per 100) (NAMCS/ NHAMCS, 1999-2000). Data on referred care from NAMCS/NHAMCS indicate significant racial disparities in referred care among the general population, but these are not seen among the elderly.
In the general population, non-Hispanic blacks, Hispanics, and APIs typically have lower rates of high variation surgeries. Among the elderly, these disparities typically persist. One exception may be in rates of hysterectomy. In the general population, admission rates are higher among non-Hispanic blacks (560 per 100,000 female population age 18 and over) compared with non-Hispanic whites (487 per 100,000). In contrast, among the elderly, rates are lower among non-Hispanic blacks (131 per 100,000 female population age 18 and over) compared with non-Hispanic whites (191 per 100,000).
Data on chronic care from the National Home and Hospice Care Survey (NHHCS) and the National Nursing Home Survey only provide black-white comparisons due to small sample size. Among the elderly, rates of home health and nursing home discharge are similar between blacks and whites, while rates of hospice discharge are lower among blacks (890 per 100,000 population) compared with whites (1,425 per 100,000).
Data on mental health care and substance abuse treatment from National Household Survey on Drug Abuse (NHSDA) do not allow assessment of disparities among the elderly due to small sample sizes.
Data on HIV care from the HIV Research Network do not allow assessment of disparities among the elderly due to small sample sizes.
Non-Hispanic blacks typically have higher rates of avoidable admissions while APIs typically have lower rates compared with non-Hispanic whites in the general population. Among the elderly, these disparities typically persist. One exception may be in rates of admission for bacterial pneumonia. In the general population, admission rates of Hispanics and non-Hispanic whites are similar, but among the elderly, rates are lower among Hispanics (1,313 per 100,000 population) compared with non-Hispanic whites (1,630 per 100,000) (HCUP SID 16-State database, 2000).
Because surveys of the general population often include inadequate numbers of the elderly to assess many disparities, analyses of the MCBS were conducted to better understand disparities in access to health care among the elderly. The MCBS allows comparisons between elderly beneficiaries with Medicare and Medigap coverage and beneficiaries with only Medicare fee-for-service coverage. The analysis presented here focuses on Medicare beneficiaries age 65 and over who reside in the community.
In general, MCBS findings confirm the presence of racial/ethnic and socioeconomic disparities in access to care among the elderly. For example, the percent of the elderly with no usual source of care is higher among blacks (7.8%) compared with whites (5.1%), Hispanics (8.4%) compared with non-Hispanic whites (4.9%), poor (7.5%) and near poor (6.3%) compared with high income persons (4.0%), and elderly with only Medicare (17.5%) compared with those covered by both Medicare and Medigap (4.6%).
Similarly, the percent who delayed care due to cost is higher among the poor (6.8%), near poor (7.3%), and middle income (2.8%) elderly compared with high income elderly (1.2%) and among the elderly with only Medicare (12.1%) compared with those who have both Medicare and Medigap coverage (3.0%).
Finally, the percent experiencing problems getting to the doctor is higher among poor (7.8%), near poor (5.5%), and middle income (4.0%) elderly compared with high income elderly (2.4%) and among the elderly with only Medicare (6.3%) compared with those covered by both Medicare and Medigap (3.7%).