NINDS Advisory Panel on Health Disparities Research
Lewis Morgenstern, M.D., Professor, University of Michigan (Chair)
Nabila Dahodwala, M.D., Assistant Professor, University of Pennsylvania
Walter Frontera-Roura, M.D., Dean, University of Puerto Rico School of Medicine
Avindra Nath, M.D., Professor, Johns Hopkins University
Bruce Ovbiagele, M.D., Associate Professor, University of California Los Angeles
Amelie Ramirez, Dr.P.H., M.P.H., Professor, University of Texas Health Science Center San Antonio
Edwin Trevathan, M.D., M.P.H., Dean, Saint Louis University School of Public Health
Barbara Vickrey, M.D., M.P.H., Vice Chair, University of California Los Angeles
In 2010 the NINDS Director convened an external panel to review and provide recommendations to NINDS with respect to the priority, administration and oversight of health disparities research conducted by NINDS. The panel met in person for a two-day meeting in May 2010, and had access to detailed records of NINDS efforts involving health disparities research in the past. The current panel’s report was divided into six sections: vision, strategic plan and priorities; organizational structure, scope and collaboration; evaluation and accountability; training; recruitment, enrollment and retention into clinical trials; and community education. Two members of the panel were assigned to write drafts of each respective section. These section drafts were circulated for feedback to the entire panel and revised accordingly. The revised section drafts were organized into a complete draft report by the panel Chair and circulated to the entire panel. In October, the panel met by conference call to discuss each section. Based on the written and verbal comments from the panel, the Chair revised the report. The report will be presented to NINDS Council in February, 2011
This report will first summarize the findings and provide the 20 major recommendations included in the six sections described above. Following this, a detailed review of health disparities research at NINDS and more detailed suggestions will be provided.
At the meeting in May, 2010 and after extensive review of the documents provided by NINDS there does not appear to be a clearly stated vision or strategic plan by which NINDS conducts health disparities research. It is obvious, however, that the Director feels that this is a high priority area from the fact that she has convened this panel and clearly speaks to the importance of this line of research going forward in the NINDS portfolio.
Based on information provided to the committee, it appears that health disparities research is sparse at NINDS, and what little there is spreads throughout NINDS across Offices and Clusters. The role that the Office of Minority Health and Research (OMHR) has played in designing initiatives and fostering programs in health disparities research has been relatively isolated from other parts of the institute.
A structured approach to track and evaluate the success of health disparities research initiatives is not apparent.
While structures exist for workforce diversity, there are no specific mechanisms for training future health disparities researchers.
There have been improvements in the enrollment of African Americans and women in NINDS funded clinical trials. There remains low enrollment of Hispanic Americans in these trials, and poor reporting of race/ethnic specific data in publications from these studies.
Community education should be evidence-driven and provide community residents with education and motivation to seek improved neurologic health. There are key scientific behavior change principles that should drive community education. Findings from NINDS-funded studies should be the focus of the messages. The audience for these research findings should specifically include not only the public but also healthcare professionals and administrators, and policy-makers – that is, all stakeholders who can use NINDS-supported research findings to effect changes in behavior and in policy to effect reduction or elimination of disparities in the burden of neurologic disease in the population.
a. Background and context
Our panel was charged with making specific and general, “practical and actionable,” recommendations on health disparities research that can guide NINDS in meeting its mission, “to reduce the burden of neurological disease through research.” NINDS developed a prior strategic plan on health disparities in 2001, which was expanded in 2003 and is analyzed in this report (National Institute of Neurological Disorders and Stroke, 2003).
The Advisory Panel as a whole received its charge from the Director, Dr. Landis, who asked us to be candid and to, “think as broadly as you can for the whole community,” with respect to recommendations regarding the unique role of NINDS in health disparities research. The panel was also asked to address strategies for broadly increasing minority recruitment into its clinical trials. The panel was asked to provide critical analysis of the extent of success of current programs and investments, and to make recommendations about how it, “can more effectively support its mission in relation to the reduction of health disparities,” specifically: (1) short-term and long-term priorities and strategies in research and training, including re-organization of current internal processes, if needed, and (2) ongoing evaluation.
Finally, the panel is mindful that all this is in the setting of a budget climate unlikely to see significant increases in the near future. The panel was advised that strategies that would leverage resources or infrastructure across institutes and agencies would be particularly welcomed.
b. Definitions and conceptual model
The panel judged that “health disparities” and “health disparities research” needed to be defined, and that how health disparities research relates to the NINDS mission needed to be made explicit. New knowledge on causes and effective interventions that reduce health disparities (“studies whose purpose is to address a priori research questions uniquely affecting ethnic minority and medically underserved groups”) was the Institute of Medicine’s recommended definition for what constitutes health disparities research (Institute of Medicine, 1999), according to the 2004 Final Report of the NIH Committee on Minority Health and Health Disparities Research Definitions and Application Methodology. The framework described in the King’s Fund report from the UK (Benzeval et al., 1995) posits four possible causal pathways for health disparities. Two of these – inadequacies in the physical environment (such as housing) and inequalities in social and economic factors (such as income) – are in general beyond the scope of health-related research to address. However, health disparities research that measures differences in health across disadvantaged populations and analyzes whether they are attributable to decreased access to needed health care services or due to higher occurrence of behavioral risk factors for disease are prime targets of health-related research (Institute of Medicine, 2003); designing, testing, and disseminating effective behavioral, healthcare delivery, or policy interventions to redress these disparities and improve health outcomes for neurologic conditions is thus clearly within the scope of NINDS’ mission (Cooper et al., 2002; Vickrey and Shapiro, 2009). Further, there are some clear biological underpinnings of different disease burden and presentation. For example, atherosclerosis appears to be deposited differentially in the intra-cranial and extra-cranial circulation among different race/ethnic groups. This argues for the need of basic science investigations in addition to epidemiologic and health services research approaches to health disparities.
One of several conceptual models for understanding disparities in health related to race/ethnicity is shown below (King and Williams, 1995):
A Framework for Understanding the Relationship Between Race and Health
It is clear from this model that there are many social, economic, and behavioral factors that drive associations of race and health; these associations are not solely biological. Given that, there needs to be research that involves social science and related disciplines, in order to redress health disparities. As the NIH Health Disparities Strategic Plan of 2004-2008 stated (National Institutes of Health, 2008; page 21), “ICs that are traditionally biological must develop an increased awareness of, and an integration with, the non-biological disease factors contributing to disease.”
What is the Value of Research on Health Disparities?
Such research should be conducted in order to acquire new knowledge that will lead to elimination of racial/ethnic and other disparities in health outcomes for persons with neurologic disease. In 2004, Steven Woolf, David Satcher, and others published a brief that projected the number of lives saved in the US from 1991 to 2000 due to advances in medical drugs and devices, and compared this to the number of deaths attributable to higher mortality in African-Americans compared to whites over the same period: while an estimated 176,633 lives were saved by the new drugs and devices over this period, an estimated 886,202 deaths would have been averted had the disparities in health of African-Americans in this time period been eliminated (Woolf et al., 2004).
Example of High-Impact Disparities Research Outside of Neurology
An example from research on racial/ethnic disparities in HIV care in the US that had a substantial impact on policy and healthcare was the HealthCare Services and Utilization Study (HCSUS). This observational cohort study identified a national probability sample of people in care for HIV in the US and followed them for two years. During the study period, antiretroviral triple therapy – a life-saving treatment for HIV/AIDS – became available. While access to therapy improved substantially in all groups over time, differences between blacks and whites in access to this therapy persisted throughout. Overall, HCSUS’ team of multi-disciplinary researchers from across the US produced high-impact papers in top-tier journals and hundreds of scientific publications altogether, to inform the research community, providers, and patient advocates. Equally important, the study’s findings also were used to guide a large roster of policy changes, including those pertaining to coverage of treatment and mechanisms to maintain vulnerable individuals on treatment, in order to reduce disparities in receipt of life-saving therapy and associated mortality disparities (Shapiro et al., 1999). Could there be differences in access to appropriate anti-epileptic drug treatment across disadvantaged groups for epilepsy? For multiple sclerosis? Without studies that measure to what extent such disparities exist and mechanistic research to elucidate what factors are associated with them, it is not possible to direct societal resources appropriately to interventions to reduce or eliminate disparities in health for these disorders.
Emerging Findings in Genomics in Neurology and Health Disparities Research
The investment that NINDS is making in biomarker research to predict treatment response and in genomics are all exceedingly relevant to health disparities research. To successfully leverage these new research findings in translational research and genetics into improved population health, there must be a complementary body of research on behavioral and healthcare organizational factors influencing racial/ethnic disparities in access to these new genetic tests. As an example of the need for such research, a recent study by Shields and others (2008) from the Harvard/MGH Center on Genomics, Vulnerable Populations, and Health Disparities surveyed a large, national sample of primary care physicians. They found that those physicians’ willingness to order genotyping tests in order to, “offer genetically-tailored treatments to their patients,” (in their study, to individually tailor smoking cessation treatment) was significantly affected by whether the test was framed as genetic or non-genetic, whether there were racial differences in frequency of key risk alleles, and whether the same genotypes that would be used to tailor the therapy were also associated with psychiatric conditions (Shields et al., 2008).
c. Delineation of the different purposes of the disparities strategic plan; relationship to the Panel’s recommendations
Logically from the prior section, research that directly acquires new knowledge on what causes, and on interventions that are tested and shown to reduce disparities, is likely to have a large impact in society. Health outcome improvement through reduction in disparities is the common goal at the intersection of the NINDS mission and federal law/DHHS regulations. Ideally, relevant audiences for research on health disparities would use the findings (1) to shape policy-making (at the healthcare delivery system or at the local, state, or national government level); (2) to change behavior of the public and/or individuals affected by these disorders (for example, lifestyle behavior change); and (3) to change behavior of health professionals (for example, more effective elicitation of patient preferences around alternative approaches to management, through greater knowledge of specific cultural issues and better decision tools for clinical practice).
The NIH Committee on Minority Health and Health Disparities Research Definitions and Application Methodology’s final recommendation was that each NIH institute should include in their report each year not only research and research training dollars that are “targeted” at health disparities research questions, but also “non-targeted” clinical research and research training. The definition of non-targeted clinical research dollars that are counted toward both Minority Health and Health Disparities Research accounting are the, “percentage of the award based on the ratio of minority participants to total participants” (for studies exceeding a minimum of 25% minority participation). While some of this research may provide new knowledge that can be directly disseminated and used to reduce racial and ethnic disparities in health, much or most of it is unlikely to have direct applications to policy or to patient or provider behavior change, based on review of titles of studies that are enumerated in Appendix B of the Panel’s notebook.
Further, the panel found it noteworthy that in FY08, a total of 8 clinical research awards comprise the entirety of targeted health disparities research funded by NINDS at present, relative to nearly 300 studies identified as non-targeted clinical research. Dollar-wise, a little over 11 million dollars supports the 8 targeted studies, compared to a calculated 90 million dollars for the portions of the other 300 studies that are counted towards health disparities investment. This is a 9:1 ratio in non-targeted to targeted research dollars in the NINDS portfolio of what is designated as clinical research in health disparities (excluding training and outreach). While 300 studies would appear to be a laudable volume of clinical research studies that have enrolled or plan to enroll a sizeable minority component into the study sample, even a small increase in “targeted” research - studies that directly address what causes and how to intervene to redress health disparities for people affected by neurological disorders - could have an enormous impact on the yield of the institute’s investment in producing new knowledge that can directly lead to reductions in health disparities. The FY 2008 investment in targeted clinical research in health disparities (excluding training and outreach) was approximately 0.7% of the entire FY08 NINDS budget.
While some general research on health disparities that is conducted elsewhere at NIH or in other agencies, such as AHRQ, NHLBI, and CDC, may yield findings that could apply to groups with neurologic disorders, there are several issues unique to people affected by neurologic disorders, and disparities research for other, non-neurological disorders may not be applicable or effective. One of these issues is stigma. Outside of mental health disorders, neurological disorders may be the most heavily stigmatized chronic conditions, and there is research to suggest that having a stigmatizing disorder and being in a disadvantaged minority population is a highly vulnerable context for which “standard” approaches may not be effective for overcoming disparities in access to or quality of care received. Second, neurologic disorders may often be of a nature that family caregivers or care partners need to be involved, as many of these diseases progress over time and may impair communication ability as well as cognition and mobility; the type of interventions that may be effective in improving quality of care with the addition of informal caregivers who may also be disadvantaged in terms of education, language concordance with providers, and wealth, is relatively unique (outside of dementia, where research is primarily supported through the National Institute on Aging). Finally, the process of championing research findings to push policy and promote behavior change by providers and patients requires time and effort.
National Healthcare Disparities Report: No Neurological Diseases Included
The 2009 National Healthcare Disparities Report issued by the AHRQ is mandated by Congress and is the seventh one that has been prepared by AHRQ and a DHHS-wide Interagency Work Group. The purpose of this comprehensive report is to enable tracking of how differences in healthcare delivery and outcomes due to disparities by race and socioeconomic status are changing over time. This important report includes metrics on quality and access to care and includes tracking of seven key conditions; none of these are neurological disorders.
d. Analysis of formulation and impact of prior strategic plan
In 2001 NINDS developed its first strategic plan in health disparities and then updated it in 2003 (NINDS OMHR, 2003). The Office of Minority Health and Research was charged with the development of the strategic plan. Implementation of its various components was described by Dr. Gordon at the in-person panel meeting in spring 2010 as moving from this office to the appropriate scientific cluster in which that disease or symptom was primarily associated, for development of RFAs and so forth.
Across the two strategic planning efforts in 2001 and 2003, separate strategic plans were developed for a total of seven diseases or populations: stroke, HIV-associated neurological diseases, neurological complications of diabetes, chronic pain disorders, cognitive and emotional health of children, epilepsy and status epilepticus, and injury to the developing brain, plus three cross-cutting areas: research capacity building and training, dissemination of public information, and inclusion of minorities in clinical research. Materials provided to the subcommittee included summaries of the 10 areas from the latest public version of the strategic plan (circa 2003), links to planning panel summaries for some of these, and a spreadsheet with a listing of planned activities in each of the ten focus areas, the outcome in terms of what happened with respect to that specific activity, and links to selected relevant materials (for example, a workshop summary or meeting report, a program announcement, a center website, others).
There is no description of the criteria for selecting these ten areas, particularly the seven areas focused on diseases, symptoms, or subpopulations. The process of priority-setting did appear to include public review of documents and elicitation of comments at several points, but a record of the criteria for selection of the ten areas, the data that were synthesized relevant to those criteria, and the process of deliberation and selection are not described. On the surface, several high volume conditions appear to have been prioritized – certainly stroke, epilepsy, and neurologic complications of diabetes. However, within those selected conditions, there is no information evident about how different goals or activities were selected as high-priority targets. For example, was it judged that stroke prevention needed more research on mechanisms of disparities or was ready for interventions to be developed and tested? The report of the 2002 Epilepsy Disparities Planning Panel (National Institute of Neurological Disorders and Stroke, 2002) stated that the panel recommended research to acquire evidence on the extent of the problem and mechanisms, but the only RFA released in response to this report requested studies on interventions. Six applications were received and none funded; but the RFA was asking for proposals addressing a stage of disparities research that may not have been ready for proposals on interventions, because the research on levels of disparities in epilepsy care and outcomes - and the factors and contexts in which they occurred - had not yet been done.
The most updated version that is described as, “the latest plan that was made publicly available,” appears to be from 2003. It is not clear who is charged with evaluating the impact of the strategic plan activities, or if there is any internal procedure within NINDS for back-and-forth feedback on implementation of the strategic plan. (If there were, one would anticipate updates or new activities in the last six years, but the most recent listed start date for any activity outside of stroke is 2004). Are activities in the plan reviewed annually, and progress relative to benchmarks assessed? If so, by whom is this done and with who is it shared? How are revisions to the implementation of the plan made?
The indicators that are described for performance measures in the plan are sometimes very non-specific (“establishment of the necessary structure to support the above lines of research…”), or are stated in terms of a measure (number of NINDS publications translated into Spanish) but not a target (like how many publications or what proportion would be the ideal goal?). Some of the performance measures are specific (“development and release of Program Announcements (PAs)/RFAs targeted on the priority area”), but the most distal ‘outcome’ described for any performance measure is the number of applications funded on that topic; there is no distillation of the yield of new knowledge acquired or to-be-acquired, or its projected impact on differences in health across racial/ethnic groups with the particular disorder.
Review of Summary of NINDS Health Disparities Strategic Plan and Outcomes
A spreadsheet in Section IIID of the panel notebook lists for each of the ten areas an activity, its objective, year initiated, and what happened out of that activity. Key implementation and outcome features tallied from those spreadsheet data are summarized below:
# of activities acted on/# of activities
Year of Most Recent Activity Start
Any PAs/ RFAs ever released
Any grants ever funded by NINDS
Yes but not from PA, but via 2 SPIRPs
Yes via SNRP
Yes but only 3 of 73 routed to NINDS were funded & on disparities
|Cognitive/ emotional health-peds|
|Dissemina-tion and Outreach|
|Minority inclusion in clinical research|
*some activities have start dates listed as years before the strategic plan
^held workshop in 2008 but unclear if arose from strategic plan
Notable findings are:
In summary, the prior strategic plan for health disparities at NINDS could have had greater impact if it had a dynamic process of review, adjustment, and follow-through on implementation. The priority-setting criteria are not specified, and the outcomes in terms of new scientific knowledge are unlikely to be extensive, given the relatively small output of intermediate outcomes such as studies funded in response to announcements.
Examples/”Best Practices” from Other ICs
The panel reviewed reports from selected other institutes available through Sharepoint links, including those of NCI, NHLBI, NIDDK, NIMH, and NIAMS. These also date to 2002-2003, and we did not examine other sources for comprehensive updates, although staff did provide brief summaries of several other institutes’ more recent descriptions of their staffing and ongoing efforts in health disparities. Of the older reports, it is notable that like NINDS, most also only describe their outcomes in terms of funded studies or even more non-specifically, with the exception of NHLBI: NHLBI provided a paragraph under outcomes summarizing key publications in high-impact journals and brief synopses of the findings and their implications. At a minimum, this is a model that enables Institute leadership to assess an outcome that is closer to estimating potential impact on reduction in a health disparity. NIMH described its priority-setting process in more detail, and included ranking of the objectives. The National Cancer Institute (NCI) appeared to be the only institute reporting (at that time) a calculator to enable estimation of the extent of disparities from SEER (Surveillance, Epidemiology and End Results program) data.
a. Vision, strategic plan and priorities
An important consideration for a NINDS health disparities research office is the development of a priority-setting process. The priority-setting process for health disparities research should be data-driven, standardized, transparent, and bring to the table a range of perspectives. Data-driven means judicious use of systematic literature reviews. The process should include public comment; standardization means use of formal, scientific methods for optimizing judgments, to rank research priorities. Priority-setting of any kind is not static or once-in-a-decade, but ideally should be updated/re-assessed – at least components – annually by NINDS. It will require a change in the current procedure and “culture” to do this.
A Priority-setting Process for Health Disparities Research Has Three Stages:
I. Surveillance - What is the magnitude of disparities in health for the condition or subpopulation? Do any exist? If so, is the magnitude or uniqueness of occurrence in that condition or subpopulation sufficient to warrant investment in mechanistic research?
II. Mechanistic - For conditions or subpopulations identified in stage I, what are the factors (biological, social or environmental) that explain the disparities, focusing in particular on those that have the potential for modification through individual, provider, healthcare delivery system, neighborhood, or policy change?
III. Intervention Design and Testing - Once a body of explanatory knowledge is sufficient to inform intervention strategies, research is needed to provide evidence as to what behavioral, clinical or policy interventions work.
IV. Dissemination/Translation of Evidence - Once successful interventions have been identified, what strategies are effective for widespread uptake of those interventions? Note: Some effort at this stage will be focused on research, but some of the effort for this stage may be to disseminate findings generated in prior stages to relevant policy and administrator audiences, as well as dissemination to the public and to professional societies.
b. Organizational structure, scope, and collaborations
The Office of Minority Health and Research (OMHR) within NINDS, established in 1998, is primarily responsible for special programs, specific grants and fellowships involving health disparities research. The organizational structure of OMHR consists of three program directors and a program specialist working in close proximity. OMHR staff-members are deployed around the nation carrying a message of NINDS priorities, particularly at scientific conferences.
However, based on information provided to the committee, it appears that health disparities research is sparse, and what little there is spreads throughout NINDS across Offices and Clusters, and the role that OMHR plays in designing initiatives and fostering programs in health disparities research is extremely limited. Health disparities research projects are not housed in a single silo, i.e. they are principally associated with scientific topic. OMHR seems to be mainly involved in efforts targeting the creation of diversity in the workforce and clinical trial populations. There is no discernible systematic approach towards originating or managing health disparities research at OMHR or Institute-wide. Although there are some important large epidemiological and health behavior studies (U-awards) geared at understanding and remedying the basis of health disparities (at OMHR, OCR), these are clearly too few.
Historically it seems that as concerns about health disparities rose, NINDS took advantage of the available SNRP to rapidly initiate programs in underserved communities that did not have major epidemiological expertise. However SNRPs were originally designed to diversify the workforce and facilitate the successful transition into R01-funded research in any scientific area. There are relatively few partnerships in collaborations with other NIH institutes (Hispanic Community Health Study is the only example seen), academia, industry, patient advocacy, and other Federal agencies with regard to developing health disparities research and moving the field forward.
There are two potential avenues for doing this:
c. Evaluation and accountability
There is little evidence of an ongoing procedure to track and evaluate the success of health disparities research efforts at NINDS. A structured approach to track and evaluate the success of health disparities research initiatives is very much needed to determine effective use of resources.
d. Training in health disparities research
In order to reduce the burden of health and healthcare disparities in the neurologic population, we first need to build the research capacity to address this issue. A critical component of this increased research capacity will be the training of high quality health disparities investigators. These investigators will require research support, a unique research skill set, a multi-disciplinary research environment and experienced mentorship to be successful.
Training opportunities specific to health disparities research do not currently exist within the NINDS. During the years 2004 to 2009, the NINDS funded a total of 1,493 career development awards (K01, K02, K08, K12, K23, K24, K99), 2,409 fellowship awards (F series) and 449 training grants (T series). However, only 13 of these awards (all K’s) involved health disparities research. This translates to less than 0.30% of all career and training awards. A targeted, more intense effort to train health disparities researchers is needed.
e. Recruitment, enrollment and retention in clinical trials
It is increasingly realized that the incidence, pathophysiology and complications of diseases are variable in different human populations. The effects of treatment are also variable in these populations. Hence studies performed in majority populations may not be generalizable to all populations. The recruitment of minority populations and underrepresented minorities pose unique challenges. Data suggest that while some improvements in recent years have occurred in recruitment of women and African American populations in NIH funded clinical trials, recruitment of Hispanics continues to be difficult, and comparable data on Native American and Asian populations is not available. The strategies for recruitment of these populations are different from those of majority populations and may require more effort to address concerns. Community based approaches and development of an element of trust are crucial. Hence it would be important to empower researchers with the information necessary to improve and maximize recruitment of these populations in clinical studies. Further, monitoring the recruitment success at each level of research may be necessary.
f. Community education and dissemination of health disparities research findings to healthcare professionals and to policy-makers
Is community outreach worth the effort and expense? The answer is a resounding yes. The need to reach at-risk populations with information garnered from NINDS-funded research is huge. With an ultimate goal of improving health outcomes in underserved minority communities, the potential exists to a) encourage healthier lifestyles, b) foster a prevention mentality, c) develop role models not only for at-risk adults but for younger generations, and d) empower individuals to gain greater control over their own health. In the process, the opportunity exists to increase public understanding of the role of NINDS and its activities. Additional benefits of a powerful community outreach program will also be felt in the areas of NINDS health disparities training and clinical trial promotion.
NINDS has had public education programs for minority populations in place for several years. So some of the following suggestions/recommendations, particularly those related to audience and communications research, have almost certainly been followed in the past. However, it is important to revisit these issues on a regular basis to ensure that programs/campaigns/strategies remain focused and on point.
Also, while it is practical to refer to particular racial/ethnic groups in umbrella terms (e.g., Hispanic or Latino), we should always remember that within those groups lay multiple sub-groups (e.g., Mexican Americans, Cubans, Puerto Ricans, South and Central Americans) with very different characteristics, backgrounds, preferences, attitudes and behaviors, etc. To be effective, community outreach programs must always keep these differences in mind.
Benzeval, M., K. Judge and M. Whitehead. (1995). Tackling Inequalities in Health: An Agenda for Action. London.
Cooper, L. A., M. N. Hill and N. R. Powe. (2002). "Designing and evaluating interventions to eliminate racial and ethnic disparities in health care." J Gen Intern Med 17(6): 477-86.
Institute of Medicine, Committee on Understanding and Eliminating Racial and Ethnic Disparities in Health Care (2003). Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, D.C., National Academies Press.
Institute of Medicine, Committee on Cancer Research Among Minorities and the Medically Underserved (1999). The Unequal Burden of Cancer: An Assessment of NIH Research and Programs for Ethnic Minorities and the Medically Underserved. Washington, D.C., National Academies Press.
King, G. and D. R. Williams (1995). Race and health: A multidimensional approach to African-American health. Society and Health. L. S. Amick BC, Tarlov AR, Walsh DC. New York, Oxford University Press.
National Institute of Neurological Disorders and Stroke (2003). Five-Year Strategic Plan on Minority Health Disparities.
National Institute of Neurological Disorders and Stroke (2002). Health Disparities in Epilepsy Planning Panel Summary.
National Institute of Neurological Disorders and Stroke (2009). Strategic Planning Disease Panel Final Report.
National Institutes of Health (2008), NIH Health Disparities Strategic Plan Fiscal Years 2004-2008.
Shapiro, M. F., S. C. Morton, et al. (1999). "Variations in the care of HIV-infected adults in the United States: results from the HIV Cost and Services Utilization Study." JAMA 281(24): 2305-15.
Shields, A. E., D. E. Levy, et al. (2008). "Primary care physicians' willingness to offer a new genetic test to tailor smoking treatment, according to test characteristics." Nicotine Tob Res 10(6): 1037-45.
Vickrey, B. G. and M. F. Shapiro (2009). "Disparities research in neurology: an urgent need." Nat Rev Neurol 5(4): 184-5.
Woolf, S. H., R. E. Johnson, et al. (2004). "The health impact of resolving racial disparities: an analysis of US mortality data." Am J Public Health 94(12): 2078-81.
Last Modified June 24, 2013