Chapter 1. Introduction and Methods
National Healthcare Disparities Report, 2008
In 1999, Congress directed the Agency for Healthcare Research and Quality (AHRQ) to produce an annual report, starting in 2003, to track "prevailing disparities in health care delivery as it relates to racial factors and socioeconomic factors in priority populations." Although the emphasis is on disparities related to race, ethnicity, and socioeconomic status (SES), this directive includes a charge to examine disparities in "priority populations"—groups with unique health care needs or issues that require special focus. The National Healthcare Disparities Report (NHDR) was designed and produced by AHRQ, with support from the Department of Health and Human Services (HHS) and private-sector partners, to respond to this legislative mandate.
The first NHDR, released in 2003, was a comprehensive national overview of disparities in health care among racial, ethnic, and socioeconomici groups in the general U.S. population and within priority populations. This 2008 NHDR represents the sixth release of this report. It continues to focus on a subset of core measures that make up the most important and scientifically supported measures in the full NHDR measure set and reports on trends in health care disparities, as well as the degree to which health care disparities for racial and ethnic minorities and poor populations have lessened.
This chapter summarizes the methodological approaches taken by AHRQ in producing the 2008 NHDR. Issues related to changes in measures, additional data sources, and modifications to presentation format are summarized below. Material that is new in this year's report is specifically highlighted and includes:
- Additional data for breast cancer and colorectal cancer treatment.
- A new measure on mental health treatment for depression.
- A new measure on daily preventive asthma medication use.
- Additional discussion of patient safety measures, including:
- Surgical discharges with catheter-associated urinary tract infection (UTI).
- Accidental puncture or laceration.
- Postoperative wound separation.
- Iatrogenic pneumothorax.
- Deaths in low-mortality diagnosis-related groups (DRGs).
- A new measure on hospice care for referral to hospice at the right time.
- Additional data on diversity of Licensed Practical Nurses and Licensed Vocational Nurses (LPNs/LVNs).
- Additional data on Native Hawaiians and Other Pacific Islanders,ii including:
- People without health insurance.
- Influenza vaccination for older adults.
- Pneumococcal vaccination for older adults.
- Continued expanded focus on individuals with disabilities.
As in previous years, the 2008 NHDR was planned and written by AHRQ staff with the support of AHRQ's National Advisory Council and the Interagency Work Group for the NHDR, which includes representatives from every HHS operating component.
i Socioeconomic disparities include differences in education and income levels.
iiDue to scarcity of health care data for this population, this report has supplemented national estimates (which often do not meet statistical reliability criteria for reporting) with State-level data. More information can be found in Chapter 4.
How This Report Is Organized
The basic structure of the 2008 NHDR includes the following Chapters:
- Highlights summarizes key themes and highlights from the 2008 report.
- Chapter 1: Introduction and Methods documents the organization, data sources, and methods used in the 2008 report and describes major changes from previous reports.
- Chapter 2: Quality of Health Care examines disparities in quality of health care in the general U.S. population. Measures of quality of health care used in this chapter are identical to measures used in the National Healthcare Quality Report (NHQR) except when data to examine disparities are unavailable. Sections cover four components of health care quality:
- Chapter 3: Access to Health Care examines disparities in access to health care in the general U.S. population. Sections cover two components of health care access:
- Chapter 4: Priority Populations examines disparities in quality of and access to health care among AHRQ's priority populations, including:
- Appendix A: Data Sources provides information about each database analyzed for the NHDR, including data type, sample design, and primary content.
- Appendix B: Detailed Methods provides detailed methods for selected databases analyzed for the NHDR.
- Appendix C: Measure Specifications provides information about how to generate each measure analyzed for the NHDR. It includes measures highlighted in the report text as well as other measures that were examined but not included in the text. It also includes information about the summary measures used in the report.
- Appendix D: Data Tables provides detailed tables for most measures analyzed for the NHDR, including measures highlighted in the report text as well as other measures that were examined but not included in the text. A few measures cannot support detailed tables and are not included in the appendix.iii When data are available:
- Race tables and ethnicity tables are stratified by age, gender, residence location, and one or more socioeconomic variables (i.e., household income, education, insurance, and/or area income).
- Socioeconomic tables are stratified by age, gender, residence location, race, and ethnicity.
iii NHDR data can now be accessed through NHQRDRnet, an online tool that provides Internet users with an opportunity to specify dimensions of analysis and produce data tables. NHQRDRnet is available through the AHRQ Web site at http://nhqrnet.ahrq.gov/nhqrdr/jsp/nhqrdr.jsp#snhere.
Presentation of the Measure Set
For the 2005 NHDR and NHQR, the Interagency Work Group was convened to select a group of measures from the full measure sets on which the reports would present findings each year. In 2006, the work group made additional changes to the core measure set. For some topics, the group favored alternating sets of core measures. These measures, which relate to cancer prevention and childhood preventive services, are listed in Table 1.1.
Table 1.1. Alternating core measures
|Reported in 2007 NHDR and NHQR*||Reported in 2008 NHDR and NHQR|
|Breast cancer screening (mammography)||Colorectal cancer screening|
|Late stage breast cancers||Advanced stage colorectal cancers|
|Breast cancer mortality||Colorectal cancer mortality|
|Children who received advice about healthy eating||Children who received advice about exercise|
|Children who had dental care||Children who had a vision check|
*The measures listed in this column will be reported again in the 2009 reports.
All core measures fall into two categories: process measures, which track receipt of medical services, and outcome measures, which in part reflect the results of medical care (Table 1.2). Not all process and outcome measures are reported for all conditions due to data limitations and other limitations (go to 2005 NHDR Chapter 1 for core report measure selection criteria).
Table 1.2. Core process and outcome measures (Measures that include data for all racial and ethnic groups and that are included in the summary analyses in the Highlights to this report are in italics.)
|Section||Process measures||Outcome measures|
|Effectiveness—End Stage Renal Disease|
|Effectiveness—HIV and AIDS|
|Effectiveness—Maternal and Child Health|
|Effectiveness—Mental Health and Substance Abuse|
|Effectiveness—Nursing Home, Home Health, and Hospice Care|
Policymakers and others have voiced their support for composite measures because they can be used to facilitate understanding of information from many different measures. The effort to develop new composites is ongoing, and in 2006, a number of new composite measures were added.iv Composite measures, which now make up about 20% of the core measures, are listed in Table 1.3.
Composite measures in the NHDR and NHQR are created based on two different models—appropriateness model or opportunities model. When possible, an appropriateness model is used to create composite measures. It is sometimes referred to as the "all-or-none" approach because it is calculated based on the number of patients who received all appropriate services. One example of this model is the diabetes composite, in which a patient who receives only one or two of the three services would not be counted as having received the recommended care.
In cases where insufficient data are available to apply an appropriateness model, an opportunities model may be applied. The opportunities model assumes that each patient needs and has the opportunity to receive one or more processes of care but not all patients need the same care. Composite measures that use this model summarize the proportion of appropriate care that is delivered. The denominator for an opportunities model composite is the sum of opportunities to receive appropriate care across a panel of process measures. The numerator is the sum of the appropriate services that are actually delivered. The composite measure of recommended hospital care for heart attack is an example where this model is applied. The total number of patients who actually receive treatments represented by individual components of the composite measure (e.g., aspirin therapy within 24 hours, beta blocker within 24 hours, smoking cessation counseling) is divided by the sum of all of these opportunities to receive appropriate care.
Table 1.3. Composite measures in the 2008 NHQR and NHDR
|Composite measure||Individual measures forming composite|
|Receipt of three recommended diabetes services|
|Recommended hospital care for heart attack|
|Recommended hospital care for heart failurea|
|Recommended hospital care for pneumoniab|
|Timing of antibiotics to prevent postoperative wound infection|
|Complications of central venous catheters|
|Patient experience of care|
a Use of angiotensin-converting enzyme (ACE) inhibitors in patients with left ventricular systolic dysfunction was changed to also include angiotensin receptor blockers (ARBs) as an acceptable alternative.
b Appropriate antibiotic selection was changed to exclude patients with health-care-associated pneumonia from the denominator used in the calculation. Collection of samples for blood culture within 24 hours of hospital arrival was changed so that only those patients who were admitted to the intensive care unit within 24 hours of hospital arrival are included in the denominator.
c The individual measure for postoperative urinary tract infection was refined to include only patients with catheter-associated urinary tract infections.
Presentation of Disparities
The 2008 NHDR and its companion NHQR continue to be formatted as chartbooks. The 2008 reports have been improved to show charts and data in a more readable format and to provide more concise summaries of the findings in each chart.
Each section in the 2008 report begins with a description of the importance of the section's topic in a standardized format. After introductory text, chart figures and accompanying findings highlight a small number of core measures relevant to this topic. When data are available, these charts typically show contrasts by:
- Race—Blacks, Asians,v Native Hawaiians or Other Pacific Islanders (NHOPIs), American Indians and Alaska Natives (AI/ANs), and people of more than one race compared with Whites.
- Ethnicity—Hispanics compared with non-Hispanic Whites.vi
- Income—Poor, near poor, and middle income people compared with high income people.vii
- Education—People with less than a high school education and high school graduates compared with people with any college education.
Almost all core measures and composite measures have multiple years of data, so figures typically illustrate trends over time. When data support stratified analyses, a figure showing racial and ethnic differences stratified by SES is included. These data are summarized in bullet format. Figures include a note about the reference group for population-based measures and the denominator for measures based on services or events.
For some measures with supporting data, regression models were run and used to help interpret bivariate and stratified results. (These are discussed in more detail in "Bivariate and Multivariate Analyses" below)viii
Gaps between priority populations and the reference group are characterized as growing larger, getting smaller, or not changing. This is done as part of the effort by HHS and AHRQ to provide information on where the Nation is——and is not—making progress in reducing disparities in health care.
Findings presented in the text meet report criteria for importance;ix comparisons not discussed in the text do not meet these criteria. However, absence of differences that meet criteria for importance should not be interpreted as absence of disparities. Often, large differences between groups did not meet criteria for statistical significance because of small sample sizes and limited power. In addition, significance testing used in this report does not take into account multiple comparisons.
Effectiveness measures for each condition or care setting area are organized further into categories that reflect the patient's need for preventive care, treatment of acute illness, or management of chronic conditions. Further detail on each of these categories and the measures included can be found in Chapter 2, Quality of Health Care.
v "Asian" includes "Asian or Pacific Islander" when information is not collected separately for each group.
vi Not all data sources used in the NHDR collect data by race and ethnicity separately (e.g., allowing for comparisons of Blacks with Whites and Hispanics with non-Hispanic Whites). When this is the case, comparisons are made by combined racial/ethnic categories (e.g., comparing non-Hispanic Blacks and Hispanics with non-Hispanic Whites).
vii Throughout this report, "poor" is defined as having family income less than 100% of the Federal poverty level; "near poor," between 100% and 199%; "middle income," between 200% and 399%; and "high income," 400% or more of the Federal poverty level.
viii The measures are obese adults given advice about exercise and individuals having a usual primary care provider.
ix Criteria for importance are that the difference is statistically significant at the alpha=0.05 level, two-tailed test and that the relative difference is at least 10% different from the reference group when framed positively as a favorable outcome or negatively as an adverse outcome.
Trends in Health Care Quality and Access
As in previous NHDRs, the 2008 report uses the earliest and most recent available NHDR data estimates for each measure to calculate average annual rate of change for the general U.S. population and for each racial, ethnic, and socioeconomic group. Consistent with Health, United States,1 the geometric rate of change, which assumes the same rate each year between the two time periods, has been calculated for the 2008 NHDR and NHQR.x
Two criteria are applied to determine whether a significant trend exists:
- First, the difference between the oldest and most recent estimates must be statistically significant with alpha=0.05.
- Second, the magnitude of average annual rate of change must be at least 1% per year when the measures are framed as a favorable outcome or as an adverse outcome.
Only changes over time that meet these two criteria are discussed in the 2008 reports. They are categorized as the following:
- Core measures for which the relative differences are changing less than 1% per year are identified as staying the same.
- Core measures for which the relative differences are becoming smaller at a rate of more than 1% per year are identified as improving.
- Core measures for which the relative differences are becoming larger at a rate of more than 1% per year are identified as worsening.
- Changes of greater than 5% per year are also differentiated from changes of between 1% and 5% per year in some figures.
An additional constraint relates to trends among specific racial and ethnic groups. Different Federal databases completed transition to the new Federal standards for racial and ethnic data that were required by 2003 at different times. These new standards created two separate racial categories: "Asian" and "Native Hawaiian or Other Pacific Islander." In addition, individuals could report more than one race. This results in underestimates for the "American Indian and Alaska Native" category, since a large proportion of this group identify as mixed race. In contrast, effects on estimates for Whites, Blacks, and Hispanics were proportionately much smaller. Consequently, the 2008 NHDR, as in the previous year, shows shorter trends (i.e., fewer years of data) for groups directly or significantly affected by the new standards, such as Asian, NHOPI, AI/AN, and multiple-race individuals.
x The geometric rate of change assumes that a measure increases or decreases at the same rate during each year between two time periods. It is calculated using the following formula: [(VY/VZ)1/N-1] × 100, where VY is the most recent year's value, VZ is the most distant year's value, and N is the number of years in the interval. See the entry for Average Annual Rate of Change in Appendix II, Definitions and Methods, Health, United States, 2007 (available at http://www.cdc.gov/nchs/hus.htm).
Bivariate and Multivariate Analyses
Bivariate analyses are included for some measures for which data are available to examine the interrelationship between race/ethnicity and SES. These analyses appear in Chapter 3, Access to Health Care, as well as in the section on low-income groups in Chapter 4, Priority Populations. This year, the NHDR also examines the question of the interrelationship between insurance status and income for quality of care in more depth for some selected measures through the use of bivariate analyses in Chapter 4, Priority Populations.
In multivariate models, estimates for a measure are controlled for multiple factors, including race, ethnicity, income, education, insurance, age, gender, and residence location, to show the extent to which these factors affect an outcome. To account for Medicare, the analyses were done separately for people under age 65 and age 65 years and over. Finally, to ensure that the findings were not biased by the sequence in which each factor was entered into the analysis, 12 separate analyses were done for each of the 12 measures. Adjusted odds ratios are shown to quantify the relative magnitude of disparities after controlling for a number of confounding factors. Two selected measures—one quality measure (obese adults given advice about exercise) and one access measure (persons who have a usual primary care provider)—were selected to conduct multivariate analyses.
In the Highlights and in Chapter 4, Priority Populations, the extent of disparities across the core measures is summarized for Blacks, Hispanics, Asians, NHOPIs, AI/ANs, and poor populations. Racial, ethnic, and socioeconomic groups are compared with a designated reference group for each core measure; each group could receive care that is worse than, about the same as, or better than the reference group. For each group, the percentages of measures for which the group received worse care, similar care, or better care were calculated.
Health care utilization measures are difficult to interpret and were excluded when summarizing disparities in access to care.xi In Chapter 4, Priority Populations, which presents information on each population separately, all core measures are used when summarizing trends in disparities for each group. However, in the Highlights, where multiple groups are presented side by side, only core measures with estimates for all racial and ethnic groups over time are used to facilitate comparisons across the groups. As noted above, an exception is made for income comparisons of quality measures because much less information is available for income groups than for racial and ethnic groups.
Beginning with the 2005 NHDR, rates relative to standard reference groups are used to quantify the magnitude of disparities and to identify the largest disparities faced by specific groups. For each group, the group rate was divided by the reference group rate to calculate the relative rate for each core measure. Relative rates of selected core measures are presented in the Highlights section of this report.
xi Interpreting health care utilization data is more complex than analyzing data on patient perceptions of access to care. Along with access to care, health care utilization is strongly affected by health care need and patient preferences and values. In addition, greater use of services does not necessarily indicate better care. In fact, high use of some inpatient services may reflect impaired access to outpatient services. For these reasons, measures of health care utilization are excluded from summaries of access to health care.
Changes to the Measure Set
The measure sets used in the 2008 NHDR and NHQR have been improved in several ways. As in previous years, a handful of measures were modified to reflect changing standards of care or improved information about care. Although no additional core measures were added, some noncore measures are being presented in the reports for the first time in 2008.
Modifications of Existing Composite Measures
The changes applied to existing measures this year were for individual component measures that make up composite measures. The changes affect the comparability of data over time to varying degrees for each measure. Beginning in the 2007 reports, the following core composite measures of effectiveness and patient safety underwent modifications:
- Recommended hospital care received by Medicare patients with acute myocardial infarction composite—The individual measure on use of angiotensin converting enzyme (ACE) inhibitors in patients with left ventricular systolic dysfunction was changed to also include angiotensin receptor blockers (ARBs) as an acceptable alternative.
- Recommended hospital care received by Medicare patients with heart failure—The individual measure on use of ACE inhibitors in patients with left ventricular systolic dysfunction was changed to also include ARBs as an acceptable alternative.
- Recommended hospital care received by Medicare patients with pneumonia. Two component measures underwent revision:
- The individual measure of appropriate antibiotic selection for community-acquired pneumonia was changed to exclude patients with health-care-associated pneumonia from the denominator used in the calculation.
- The individual measure for the collection of samples for blood culture within 24 hours of hospital arrival was changed so that only those patients who were admitted to the intensive care unit within 24 hours of hospital arrival are included in the denominator.
- Postoperative care. The individual measure for postoperative urinary tract infection was refined to include only patients with catheter-associated urinary tract infections.
New Noncore Measures
Non-core measures are measures identified by the Interagency Work Group to provide additional information in the NHDR to fill a specific data gap in a particular topic area. Each year, the NHDR features supplemental measures in special focus sections of the report. The following new non-core measures have been included in the 2008 NHDR to fill identified gaps:
Non-core measures are measures identified by the InteragencyWork Group to provide additional information in the NHDR to fill a specific data gap in a particular topic area. Each year, the NHDR features non-core measures in special focus sections of the report. The following new non-core measures have been included in the 2008 NHDR to fill identified gaps:
- One measure of workforce diversity from the American Community Survey (ACS):
- U.S. LPNs/LVNs by race and ethnicity.
- A refined measure of mental health treatment for depression from the Collaborative Psychiatric Epidemiology Surveys:
- Receipt of minimally adequate treatment by adults with mental disorders.
- One measure of hospice care quality from the National Hospice and Palliative Care Organization Family Evaluation of Hospice Care:
- Referral to hospice at the right time.
- Five additional measures of patient safety highlighted from the NHDR noncore measure set:
- Postoperative wound separation.
- Reclosure of postoperative abdominal incisional separation.
- Accidental puncture or laceration during procedures.
- Iatrogenic pneumothorax.
- Deaths per 1,000 admissions in low-mortality DRGs.
Measure revisions were proposed and reviewed in meetings of the Interagency Work Group for the NHDR, which includes representation from across HHS.
Databases Used in the 2008 Reports
Table 1.4 lists the databases used in the 2008 reports. This year the Collaborative Psychiatric Epidemiology Surveys (CPES) is added to the set of databases that were used in the 2007 reports.
Table 1.4. Databases used in the 2008 reports (new databases in italics)
|Survey data collected from populations: |
|Data collected from samples of health care facilities and providers: |
|Data extracted from data systems of health care organizations: |
|Data from surveillance and vital statistics systems: |
a This file is designed to provide national estimates of disparities in the AHRQ Quality Indicators using weighted records from a sample of hospitals from the following 23 States: Arizona, Arkansas, California, Colorado, Connecticut, Florida, Georgia, Hawaii, Kansas, Maryland, Massachusetts, Michigan, Missouri, New Hampshire, New Jersey, New York, Oklahoma, Rhode Island, South Carolina, Tennessee, Texas, Vermont, and Wisconsin.
Expanded Analysis of Trends in Disparities
Starting with the 2007 NHDR, an additional dimension of trends is emphasized in the reports. In previous years, the NHDR included discussion of change over time for each population from baseline to most recent data year, where available., The NHDR also reported separately on statistically significant differences between a comparison group and reference group. The 2008 NHDR continues to combine the discussion of change over time and differences between groups by focusing on the change over time in the difference between groups. Thus, the NHDR can now show not only where disparities exist in health care quality and access to health care, but also how disparities have changed over time.
The change over time in the disparity between the comparison group and reference group is reported if the following two criteria are met:
- There is a statistically significant difference between the estimate for the baseline year and the most recent data year for at least one group.
- There is a statistically significant difference between a comparison group and reference group at baseline year, at most recent data year, or at both baseline year and most recent data year.
The change is reported as increased if the absolute difference of the gap between the comparison group and reference group in the most recent data year is greater than the gap in the baseline year. The change is reported as decreased if the absolute difference in the gap between the comparison group and reference group in the most recent data year is less than the gap in the baseline year. Change is not reported if there is no change in either comparison group or reference group, if both the reference group and comparison group show significant differences between the baseline year and the most recent data year, or if data are not available for more than one data year.
No statistical test was performed for the difference from the baseline year to the most recent data year in the difference between the comparison group and reference group. Since most differences are relatively small, performing an additional statistical test would eliminate most reports of change in disparities.
The NHDR continues to report statistically significant differences between the comparison group and reference group where data are available for the most recent data year. Since the NHDR now contains estimates for several data years for most of the measures, for simplification in reporting on disparities, only the most recent data year is discussed.xii
xii For estimates from prior years, go to Appendix D: Data Tables for previous releases of the NHDR.
Individuals With Disabilities
The Interagency Work Group Subcommittee on Disability Statistics convened to develop a broad definition of disabilities that can be applied across different national data sources to obtain data on the quality of care for people with disabilities. Beginning with the 2007 NHDR, AHRQ has used a broad, inclusive measure of disability. This measure is intended to be consistent with statutory definitions of disability, such as the first criterion of the Americans with Disabilities Act of 1990, 42 U.S.C. §§ 12181 et seq., (i.e., having a physical or mental impairment that substantially limits one or more major life activities) and other Federal programs' definitions based on the Americans With Disabilities Act.
For the purpose of the NHDR, people with disabilities are those with physical, sensory, and/or mental health conditions who also have an associated decrease in functioning in such day-to-day activities as bathing, walking, doing everyday chores, and/or engaging in work or social activities. In displaying the data on disability, paired measures are shown to preserve the qualitative aspects of the data:
- Limitations in basic activities represent problems with mobility and other basic functioning at the person level.
- Limitations in complex activities represent constraints encountered when people, in interaction with their environment, attempt to participate in community life.
The use of the Subcommittee's recommendation of these paired measures of basic and complex activity limitations is conceptually similar to the way others have divided disability. It is also is consistent with the International Classification of Functioning, Disability and Health (ICF) separation of activities and participation domains. These two categories are not mutually exclusive; persons may have both limitations in basic activities and limitations in complex activities. Further information regarding the definition and methods can be found in the Individuals with Disabilities and Special Health Care Needs section in Chapter 4, Priority Populations.