Taking the Measure of Health Care Disparities

Despite growing public and private efforts to reduce disparities in health care, these disparities persist in the United States. Soon after the Institute of Medicine (IOM) Unequal Treatment report systematically documented disparities in multiple components of the health care system (Smedley, Stith and Nelson 2003), Congress mandated the Agency for Healthcare Research and Quality to monitor national trends in health care disparities through the annual National Healthcare Disparities Report (Agency for Healthcare Research and Quality 2011). For health services researchers and policy makers, numerous methodological challenges have arisen in identifying disparities in health care, tracking them over time, understanding their underlying mechanisms, and evaluating the effects of practices and policies to reduce them.

This special issue of Health Services Research is devoted to highlighting challenges in the measurement of disparities in health care and exploring innovative approaches to address these challenges. The articles in this issue arose from a conference on “Methodological and Conceptual Issues in the Measurement of Healthcare Disparities” convened at Harvard University in February 2010. This conference was jointly organized by the Center for Multicultural Mental Health Research of the Cambridge Health Alliance through a Disparity Center funded by the National Institute of Mental Health and by the Health Disparities Research Program of Harvard Catalyst. The conference featured presentations and multidisciplinary dialog among researchers who assess health care disparities from multiple perspectives, including health services research, economics, sociology, biostatistics, medicine, and public health. The conference also elucidated conceptual and methodological challenges that the field needs to address. Financial support for the conference and publication of this special issue was provided by the Office of Behavioral and Social Sciences Research of the National Institutes of Health (NIH), National Institute of Mental Health, and Harvard Catalyst, the NIH-funded Clinical and Translational Sciences Center of Harvard University.

The articles in this special issue have three primary themes: (1) measurement challenges when examining and evaluating mechanisms of health care disparities from diverse disciplinary perspectives; (2) recent innovations in disparities measurement and analytic methods; and (3) recommendations to health services researchers for measuring and tracking health care disparities in diverse contexts.

In the opening article, Cook, McGuire, and Zaslavsky (2012) build on the strong foundation of the IOM Unequal Treatment report as a rallying point for health services researchers studying disparities in health care. They remind readers of how the IOM originally articulated the disparities research agenda that has evolved over the ensuing decade. The authors support the IOM definition of health care disparities as a standard for the field and discuss how this definition can be used to assess the various underlying mechanisms of disparities. The development of the IOM definition itself offers valuable lessons which, in turn, may be important to future studies of inequity and fairness in health care. The authors assert that a consistent definition of disparities in health care is needed, concordant with suitable analytic methods, to assess the use and quality of health care, track disparities, and evaluate policy and practice interventions.

As a reminder for researchers not to be constrained or limited by the IOM definition, articles by Williams et al. (2012) and White, Haas, and Williams (2012) in this issue encourage broader perspectives on health care disparities that include social determinants of health care and area-level factors related to neighborhoods and communities. Both articles illustrate how multiple risk factors and resources combine to affect the distribution of disease and its management.

Williams et al. (2012) introduce the concept of intersectionality theory from the field of sociology and apply this approach to disparities in lung cancer as a model. This article underscores the need to incorporate gender, race/ethnicity, and socioeconomic factors when analyzing disparities in lung cancer risk, treatment, and outcomes. The intersectional lens illustrated in this review helps unpack the factors that heighten risk, and how this information might help target interventions by noting the “distinct patterns of risk and resilience [that] emerge at these intersections.” This article supports a more nuanced approach to racial/ethnic categories for identifying potential epigenetic effects of cancer risk, given the differences in residential segregation and occupational environments by race/ethnicity, socioeconomic status, and gender.

With an in-depth consideration of residential segregation, White, Haas, and Williams (2012) highlight the role of place in health care disparities. They emphasize the influence of segregation operating in neighborhoods and health care systems to shape access to health care and the quality of preventive, diagnostic, therapeutic, and end-of-life services across the life course. The authors make a compelling case for how segregation continues to have lasting implications and serious consequences for the well-being of both individuals and communities.

The next two articles by Zaslavsky, Ayanian, and Zaborski (2012) and Alegría et al. (2012) present empirical analyses of key topics related to disparities measurement. Over the past two decades, data from the Medicare program have been widely used to assess racial disparities in health care for elderly Americans. Using self-reported data on race and ethnicity from Medicare beneficiaries in the Consumer Assessment of Healthcare Providers and Systems (CAHPS) survey during 2010, Zaslavsky, Ayanian, and Zaborski (2012) compared these self-reports with administrative records of race/ethnicity in the Medicare enrollment database. They found that designations of white or black race were generally accurate in Medicare enrollment data, but Hispanic, Asian, Pacific Islander, and American Indian beneficiaries were underidentified. A novel finding was that beneficiaries identified by Medicare enrollment data in these latter four groups were poorer and reported worse health and health care experiences than others in these self-reported groups who were not identified by Medicare data. In 2009 the IOM strongly recommended that health care organizations collect data on patients’ race, ethnicity, and language with more consistent and standardized methods using both direct and indirect measures to address disparities in quality of care (Ulmer, McFadden, and Nerenz 2009). The study by Zaslavsky and colleagues provides a reminder, however, that the national infrastructure to ascertain reliable data on race and ethnicity is still lacking.

The study by Alegría et al. (2012) is a timely contribution as the Patient Protection and Affordable Care Act is implemented to expand health insurance coverage in the United States. This study focuses on the role of insurance coverage in reducing racial and ethnic disparities in use of mental health services. The study demonstrates that broad policy interventions, such as providing insurance coverage, will have varying impacts on different racial and ethnic groups—thus suggesting that more tailored and multifaceted interventions are needed to reduce disparities for some groups. Looking forward, the authors remind us that future studies need to uncover the causal pathways that generate disparities to support appropriate interventions for different groups.

The next two articles offer new approaches and methods for examining disparities. Lapatin et al. (2012) present an innovative approach for using clinical vignettes to elicit the stories of patients, families, and clinicians about what is important to them in health care. Although vignettes have been used to some extent in health services research, the authors explicitly outline a step-by-step approach to creating vignettes in a study of health care disparities to explore complex questions and account for multiple perspectives. Health services researchers can combine vignettes with other methods in robust mixed-methods studies of disparities.

De Las Nueces and colleagues (2012) present a systematic review of clinical trials that utilized methods of community-based participatory research (CBPR) to encourage minority participation. Their review demonstrates the value of CBPR to achieve high rates of minority participation, and consequently, greater generalizability of findings. However, their review also found significant gaps in the role of CBPR in the interpretation of results or in the dissemination of effective interventions. These findings suggest new venues in which CBPR might help inform minority communities of pertinent research as well as gaps when translating research into practice in communities.

The final article in this special issue considers why and how genomic data should be incorporated into studies of health care disparities as a new frontier in disparities research. Shields and Crown (2012) highlight advances in genomics research that are becoming critical to predict disease risk and treatment outcomes more effectively. Noting that growing insights from genomics research are not yet well integrated into disparities research, these authors describe how such data can “help to differentiate geographical ancestry from self-identified understandings of race; provide additional information about health status, appropriate care, and expected clinical outcomes; and shed new light on the importance of certain socioeconomic factors.” They emphasize the growing role of electronic health records as a source of genomic data to be used by health services researchers to study health care disparities. The authors also demonstrate how including genetic variables in disparities research can diminish measurement error and reduce bias in estimates of treatment effects, thereby refining measurement of disparities in health care and health outcomes.

Together the articles in this special issue offer important lessons and identify challenges that remain in advancing an agenda for disparities research that can be applied in real-world settings to promote greater equity in health care. Some of the most salient lessons include:

• The value of a consistent definition of a health care disparity: The way in which a disparity is defined can have substantial influence on disparities measurement, research findings, and policy determinations. Implementing a more consistent definition with corresponding analytic methods is vital to assess rates of health care access and quality, monitor disparities, and evaluate policy and systems interventions.

• The centrality of using valid measures of race and ethnicity: More effort should be invested to evaluate racial and ethnic designations in health care databases. Given changing social understandings of race and ethnicity, past designations based on observations of health care staff are often inconsistent with the self-identification of individuals. Investing in the infrastructure to collect accurate race, ethnicity, and language data as well as geographic (e.g. residential segregation, neighborhood resources) and socioeconomic indicators (e.g. education, income) can help to disentangle factors that put groups at risk for disparities and illuminate the intersection of different social identities that can lead to health care disparities.

• Avoiding oversimplified explanations of disparities that only address race and ethnicity: Several articles reveal the importance of going beyond race and ethnicity to understand racial/ethnic disparities in health and health care. Gender, socioeconomic status, residential segregation, and genomic factors may often be mediators of observed disparities by race or ethnicity.

• The important voice of families and of communities in disparities research: The cultural norms of families and communities are crucial to understand the need for care and potential underlying mechanisms of health care disparities. Involving communities and families in disparities research can facilitate participation in clinical trials and intervention studies, clarify expectations of care, and help researchers to design and disseminate interventions that are acceptable to communities and families that experience disparities.

• Lack of single policy interventions to eliminate health care disparities: Given the complexity and multilevel factors that play a role in health care disparities, designing a single intervention that could eradicate these disparities consistently for all racial and ethnic groups is unrealistic. Tailored and multifaceted interventions will have much better prospects for success.

• The usefulness of multidisciplinary collaborations: This set of articles illustrates the importance of integrating diverse disciplinary perspectives (e.g. environmental health, sociology, economics, and others) on research teams to investigate clinical, social, economic, and environmental causes of disparities and to generate effective solutions aimed at reducing them.

The articles in this special issue reveal the persistence of health care disparities and the complexity involved in studying these disparities. Long-term investments are needed to infuse the field of disparities research with more innovative and bolder studies that can help to eliminate these disparities. We hope that the articles presented in this special issue stimulate such innovative ideas with new methods and approaches for investigating disparities in health care, thereby providing a stronger foundation for cutting-edge health services research.