A call to action to include disability in intersectional health equity research and policy

Summary

Disability status is rarely included in health research and policy, including intersectional research, perpetuating health inequities for this population. This paper calls on researchers and policymakers to take concrete steps to advance health equity for disabled people, including those at the intersections of disability, race, ethnicity, poverty, and other marginalized identities. We propose four strategies with recommendations to promote: a) meaningful engagement of disabled and multiply marginalized people in research and policy planning; b) cohesive, systemic disability data collection and analyses; c) use of intersectional approaches to examine structural drivers of health inequities; and d) leveraging of administrative data to improve disability healthcare policies and practices.

Public health researchers and policy makers increasingly understand health inequities from an intersectional framework. Intersectionality is rooted in the premise that human experience is shaped by one’s multiple social positions (e.g., race, gender, disability), and cannot be adequately understood by considering social positions in isolation.¹ Globally, 16% of the world’s population lives with a disability,² yet this population is largely invisible in global public health,³ including in intersectional research.⁴ The United Nations (UN) flagship report on Disability and Sustainable Goals 2018⁵ documents that persons with disabilities experience large and preventable disparities in Sustainable Development Goal indicators, including increased likelihood of reporting “poor health,” barriers to quality health care, and unmet healthcare needs. For instance, among children with special healthcare needs in the United States of America (USA), White children were reported to have better health and fewer barriers to health care compared to Black and Latino children.⁶ These differences were largely explained by Social Drivers of Health (SDoH), including family income, family structure, and adverse childhood experiences.⁶ In the February 2024 issue of this journal, Lundberg and Chen defined structural ableism and introduced a conceptual framework for examining ableism in public health and health care.⁷ Structural ableism is “a system of historical and contemporary policies, institutions, and societal norms and practices that devalue and disadvantage people who are disabled … and privilege people who are … nondisabled” (p.1). We reiterate their call to address ableism and call on researchers and policymakers to take concrete steps to include people with disabilities in advancing health equity from an intersectional framework.

The authors of this report represent a coalition of US researchers and advocates, a majority identifying as individuals with disabilities, some with intersecting identities of LGBTQ+ and race. We reflect a range of expertise acquired through lived experience, policy analysis, program implementation, and public health research. We regard disability from an International Classification of Functioning, Disability, and Health (ICF) perspective⁸ whereby disability results from the interaction of a person’s limitations with environmental demands. Disability is often measured in surveys by self-report of functional limitations in seeing, hearing, walking/climbing steps, communicating, self-care, and concentrating/remembering.⁹ Consistent with Stangl and colleagues,¹⁰ we define prejudice and bias as negative attitudes toward a group of people (e.g., a racial group) that is presumed to have undesirable qualities and has less power than the dominant or majority group. Stigma refers to negative attributions to persons, often with atypical conditions (e.g., disability), by others and includes the internalization of such stigma by persons with those traits (self-stigma). Discrimination can be manifested interpersonally (behaviour, treatment) and systemically (policies).¹¹

In accordance with the preferences of our authors with disabilities, we use “disabled people” and “people with disabilities” interchangeably. While our views are informed by experiences in the USA, we intend that our recommendations could be applicable to other countries. We describe three premises that undergird our work and then call for four actions to improve health equity in the Americas for people with disabilities and other intersectional identities (refer to Table 1 for a summary).

Table 1.

Proposed strategies and recommendations to promote health equity for disabled and multiply marginalized people.

Thesis
Address health inequities for disabled and multiply marginalized people by including them in public health equity and intersectional research and policy.

Strategies for Action	Recommendations
I.Systematically and meaningfully engage disabled and multiply marginalized people in all aspects of research and policy planning. Target audiences: federal and state government agencies, funding agencies, institutions of higher education, researchers, and disabled, multiply marginalized communities	1.1Increase the number and diversity of researchers with disabilities. 1.2Include people with disabilities and intersecting identities in all research routinely and proactively. 1.3Expand opportunities for people with disabilities to engage in public health program and policy development. 1.4Partner with disability organizations in all-cause emergency planning and response.
II.Improve disability health data. Target audiences: federal and state government agencies, funding agencies, institutions of higher education, researchers, and disabled, multiply marginalized communities	2.1Include disability status in all national surveys and collaborate with disabled people to develop a disability data framework. 2.2Increase public health funding to collect information on Social Drivers of Health (SDoH). 2.3Make disaggregated health data widely available to researchers and the public.
III.Understand health inequities through intersectional research. Target audiences: researchers, federal and state government and other policymakers, healthcare providers, disability organizations, and disabled, multiply marginalized communities	3.1.Apply a data equity lens that includes disability in health research. 3.2.Strategically address small samples in intersectional research. 3.3.Delineate causal relationships between identity, social processes, and health inequities. 3.4.Analyze policy and enact inclusive policies and programs.
IV.Leverage administrative data to improve disability healthcare equity and quality. Target audiences: federal government agencies such as the US Centers for Medicare and Medicaid Services and US Department of Health and Human Services (DHHS) Office for Civil Rights, state policymakers, Big Data and healthcare organizations, disability rights organizations, and disabled, multiply marginalized communities	4.1.Recognize people with disabilities as medically underserved. 4.2.In the US, amend Medicaid reimbursement rates to fairly compensate healthcare providers for additional time and expertise needed to provide quality care to patients. 4.3.Determine methods to link Medicaid and Medicare data and facilitate ready access to the data. 4.4Develop anti-ableist ethical standards, policies, and oversight structures in healthcare plans and organizations to protect the rights of disabled people. 4.5Develop oversight structures to guide the use of algorithms, AI, and Big Data for use in health systems in the US and internationally.

Premises

First, disability and health are distinct though related constructs. This presumes a definition of disability based on functional limitations and not on medical diagnoses. Health disparities experienced by disabled and other marginalized groups do not stem primarily from biological differences or cultural practices; rather, they reflect SDoH. Health inequities for people with disabilities arise from exclusionary conditions, such as stigma and discrimination, in education and employment.² Many healthcare practices, including clinical preventive services, discriminate against patients with disabilities. For example, women with disabilities receive breast and cervical cancer screening at lower rates, are diagnosed at later stages, and receive differential treatment of detected cancer compared to nondisabled women.¹² These practices relegate disabled people to the margins of society where opportunities to be healthy are denied.

Second, as a group, disabled people are rarely included in health data, policy analyses, and program planning. Historically, many countries institutionalized people with significant disabilities, keeping them out of sight and out of mind. Today, people with disabilities remain invisible in many health surveys because they are explicitly excluded, not identified as disabled in data, or prefer not to disclose their disability because of stigma.¹³ For example, household surveys in the USA do not include institutionalized populations, excluding people with disabilities in group homes or larger institutions. Among the 23 countries in Latin America and the Caribbean that collected population surveys during the 2019 coronavirus pandemic, only four (Colombia, Costa Rica, Mexico, and Peru) included disability identification questions.¹⁴ Even when included in the survey sample, if participants are not identified as disabled, their data cannot be disaggregated, and they remain invisible. During the early months of the COVID-19 pandemic, the US Household Pulse survey did not identify people with disabilities. This omission delayed identification of—and actions to address—infection and death rates that became twice that of the general population.¹⁵ Disability identifiers were only added following community activism.

Third, disability often intersects with other marginalized identities. People with disabilities are more likely to identify as LGBTQ+, a racial minority, and low-income than the general population.¹⁶ Intersectionality is a concept for understanding how multiple forms of inequality compound to create overlapping systems of disadvantage.¹⁷ Policies and power structures differentially privilege and oppress certain groups. As examples, persons from marginalized communities typically have no or few legislators representing their views, have no or few healthcare providers who share their lived experience, and are underrepresented among researchers engaged in developing new knowledge that speaks to their lives. While intersectional research typically includes race, ethnicity, LGBTQ+ identity, and income level, it rarely includes disability status. A woman with mobility limitations who is Black and identifies as queer faces compounding types of stigma and discrimination. Achieving health equity for people with disabilities requires addressing ableism in the context of racism, sexism, heterosexism, and income inequality.

Strategies for action

Systematically and meaningfully engage disabled and multiply marginalized people in all aspects of research and policy planning

Research and policy planning often deliberately or incidentally exclude disabled participants.¹³ The result is a failure of policies to recognize and address the impact of intersecting systems of oppression on the health of disabled people.¹⁸ Engaging disabled and multiply marginalized people in all aspects of research and planning requires intentionality, effort, and community engagement. Such engagement ensures that: a) research questions address issues that matter to the disability community, b) measures are appropriate, and c) findings are interpreted correctly and communicated effectively to scientific and lay audiences.

We make four recommendations towards realizing systemic, meaningful inclusion. First, it is essential to increase both the number and diversity of researchers with disabilities, as emphasized in a 2022 advisory report to the US National Institutes of Health (NIH).¹⁹ Second, to encourage researchers to deliberately include disabled and multiply marginalized participants, we recommend that human subjects’ oversight committees set enrollment reporting requirements including disability alongside gender, race, and ethnicity data. Third, engaging the disability community in public health program planning and policy development requires proactive recruitment and accessible reports that are easy to read, easy to find, and free of charge. Finally, collaboration with local and national disability organizations is crucial for effective emergency planning. Disabled populations are disproportionately impacted by emergencies, natural disasters, and climate change, and can serve as vital partners in preparedness efforts.²⁰ We call for guidelines to manage, secure, and facilitate access to individual-level data on disability-related needs before, during, and after emergencies to improve response and resilience efforts.

Improve disability health data

In many countries, understanding and addressing health in disabled and multiply marginalized groups is hindered by inconsistent operational definitions of disability. The Washington Group disability identification questions were developed to allow comparisons between countries.⁹ Yet, a 2023 review across 188 countries found that 63 countries, including 13 in the Americas, still had no data sets with functional-difficulty questions.²¹ Further, researchers often develop ad-hoc disability definitions to suit their purposes. Finally, although USA policy mandates the collection of disability status in all national population health surveys, the required question set leaves many people with significant mental health or communication limitations, for example, invisible in the data.

In analyses, many federal datasets are neither disaggregated by disability¹³ nor by disability plus additional marginalized identities. Further, they fail to consider social and environmental contexts such as rurality, poverty, access to education, and healthcare. As a result, health inequities are incorrectly attributed to biological or behavioral factors associated with identity and perpetuate “blaming” individuals or groups for their poor health outcomes.

Firstly, to improve disability health data, we follow the US Healthy People 2030 in recommending the inclusion of disability status and questions to identify various types of functional limitations—such as mobility, sensory, developmental, intellectual, psychiatric, communication, and self-identified disabilities—in all national surveys. Following Australia’s example,²² governments are encouraged to collaborate with disabled people to develop robust disability data frameworks that catalogue available and needed information. These frameworks must extend beyond basic demographic variables to include minimal data elements like contextual factors, types of functional impairments, adequacy of programs in meeting health needs and disability-related expenses, and relevant outcome measures. Standardization and alignment of disability definitions across various agencies is essential to ensure data interoperability. Periodic monitoring of this process will be needed to identify who is and who is not counted within disability statistics. An example of progress in the USA is the State of Oregon Health Authority, which incorporates disability indicators, including types of functional limitations, and support needs in health data.²³ It is also vital to conduct health research that captures data from institutionalized populations living in group facilities and prisons,²⁴ many of whom have disabilities. Finally, the Robert Wood Johnson Foundation, a leader in addressing health inequities in the USA, is called upon to analyze data by disability status where available and advocate for its inclusion where it is not yet present.

Secondly, increasing public health funding to support the collection of SDoH data is essential, as recommended by the US National Academies of Sciences, Engineering, and Medicine.²⁵ National health ministries, in partnership with disability organizations, should develop and utilize common measures of SDoH, including valid indicators of ableism, racism, and other types of discrimination at both interpersonal and structural levels.

Lastly, making disaggregated health data widely accessible to researchers and the public is fundamental for advancing equity. In the USA, valuable examples include the Centers for Disease Control and Prevention’s Disability and Health Data System and the Maternal and Child Health Bureau’s National Survey of Children’s Health website, both of which provide detailed, accessible data essential for research and policy development.

Understand health inequities through intersectional research

Health inequities are well established for disabled and multiply marginalized people.²⁰ However, the mechanisms and opportunity structures that drive negative health outcomes are poorly understood.²⁶ Intersectional research can be especially vulnerable to small sample sizes that preclude population-based estimates of health. Omitting data for intersecting identity groups, or combining small groups into “other” categories, perpetuates ignorance and policy inaction. Researchers make dozens of choices when conducting a study, including theoretical models, study designs, participant recruitment, analyses, and interpreting and communicating findings. Unless examined thoughtfully, each choice is likely to reinforce researchers’ biases in ways that affect data outcomes and related actions. Researchers must practice reflexivity, systematically interrogating each inflexion point so that decisions are intentional and transparent about whose worldview is centered. Empowering people with disabilities and intersectional identities on research teams is an effective way to identify and address implicit bias.

Marginalized identities confer different barriers across life contexts of health, education, employment, and community participation. An intersectional framework allows health equity researchers to explore the complex ways that multiple social identities and SDoH intersect to perpetuate health inequities. Disentangling the contributions of disability relative to other identities requires understanding how people perceive their various identities and react to discrimination. Research is needed to understand the mechanisms of resilience in the context of multiple systems of oppression.

To better understand these health inequities, we recommend: (i) Applying a data equity lens that explicitly includes disability in all health research. Such a perspective encourages careful consideration of how disability and intersectionality data are collected, analyzed, interpreted, and communicated. (ii) Using strategic approaches to analyzing small samples such as combining multiple years of survey data, oversampling, and collecting supplemental data for populations of interest. When standard, large survey methodologies are not feasible, consider alternative methods such as case studies, descriptive statistics with appropriate caveats, stratified random sampling, or snowball sampling with weighting to increase representativeness.²⁷ (iii) Intersectional approaches—including interaction effects analysis, multilevel modeling,²⁸ latent class analysis,²⁹ structural equation modelling,³⁰ and decomposition analyses⁶—can delineate causal relationships between identity and social processes and the mechanisms driving health inequities. (iv) Intentional policy analyses and inclusive policies and programs are needed to address compounded health inequities. The Intersectionality-Based Policy Analysis framework offers a valuable approach for developing policies to change structural drivers impacting multiply marginalized populations.³¹

Leverage administrative data to improve disability healthcare equity and quality

Administrative data systems, initially developed to document care and payment, are increasingly used to benchmark care utilization and patient outcomes. Even when disabled and multiply marginalized people have financial access, long waiting times, discrimination, high costs, and transportation barriers lead to underutilization of services.³² Healthcare administrative data systems are repositories of eligibility data and longitudinal medical, pharmacy, and dental claims across public and private insurers.

Analyses of health services utilization can provide valuable information, but there are limitations in these systems for disabled members. Administrative health data, including electronic health records (EHR), rarely include self-reported disability or functional limitations.³³ Instead, disability is inferred from medical diagnoses, which miss people whose functional limitation is not indicated by a medical condition. The disability community has long advocated that EHR should record, at the patient’s discretion, patient-reported disability, functional limitations, and needed accommodations.

Lacking universal healthcare and a strong health data infrastructure, the USA could leverage health data from its two publicly funded safety net programs for low income (Medicaid) and older and disabled persons (Medicare) to understand health, healthcare access, and health equity over the life course. However, these data are underused because of the cost and complexity of linking data across these systems. Although the Centers for Medicare and Medicaid Services created a national data repository (T-MSIS) for Medicaid and Children’s Health Insurance Program data, states are not required to submit disability data, and Medicare data is not included. Linking data across federal and state programs is essential for monitoring health status and healthcare utilization for all people covered by public insurance.

Increasingly, health systems use algorithms and artificial intelligence (AI) to make coverage and clinical care decisions.³⁴ Misinformation and bias in the data used to train algorithmic tools leads to inaccurate conclusions about the health of subgroups of society. When healthcare costs lead to underutilization of care, healthcare algorithms underestimate the needs of people with disabilities and other groups with unmet healthcare needs.³⁵ When these algorithms are proprietary, it is virtually impossible to ‘look under the hood’ to ensure public accountability.³⁶ Flawed algorithms used to triage care during COVID-19 led to denial of care, including ventilators, and higher mortality rates for disabled and other marginalized groups.³⁷

Big Data captures data from sources like biospecimens, health records, satellite sensors, administrative records, environmental indicators, behavioral data, social media and mobile phone data, and public opinion. These datasets are used by AI systems and may reinforce the marginalization of some communities. Big Data and AI may infer disability status based on purchase history, social media postings, or search histories.³⁵ When these disability “flags” are flawed, analyses generate misleading disability health data. In the USA, healthcare entities are required by law to protect patient privacy but can purchase flawed Big Data to inform their operations, perpetuating errors and health inequities.³⁵

We recommend the recognition of disabled people as medically underserved, such as the Special Medically Underserved Population designation in the USA,³⁸ and that programs are designed to promote disability and health research and increase access to health care. In addition, in the USA, amending Medicaid reimbursement rates to fairly compensate healthcare providers would promote quality care for patients who require additional time and expertise. Establishing methods to link Medicare and Medicaid data will facilitate timely, affordable, and ready access for researchers, protect patient privacy, and provide longitudinal data. This includes requiring states to add disability to the Medicaid data supplied to the national repository. Furthermore, developing anti-ableist ethical standards, policies, and oversight within healthcare organizations is vital to recognizing the equal worth of disabled individuals and their right to equitable health care and effective communication. Healthcare entities should not only monitor indicators such as health coverage, healthcare utilization, unmet health needs, and forgone care, but also disclose their use of algorithms, AI, and Big Data—detailing how these tools inform decisions, which populations are affected, and how human judgment oversees outcomes. To ensure responsible use of advanced technologies, it is necessary to establish oversight structures that guide their application including auditing and eliminating biases in algorithms, enacting policies that promote lawful and ethical decision-making, and holding organizations accountable for harms resulting from the misuse of administrative data, AI, and Big Data.³⁹ Notably, the USA has passed an algorithmic accountability act aimed at increasing transparency and accountability.⁴⁰

Conclusions

This paper builds on the 2018 UN report on disability and the 2023 NIH designation of disability as a health disparity population. We celebrate these advances but note that health inequities for disabled people cannot be eliminated without a) considering people’s other intersecting and marginalized identities; b) bringing visibility to the prejudice, stigma, and discrimination that shape and perpetuate inequities; and c) addressing the structural barriers that cause health inequities. Not the least of these are the persistent mindsets of ableism and racism in healthcare policies and practices that need to be identified and eliminated across countries. We recommend specific actions toward equity. Now is the time to take action toward health equity for disabled and multiply marginalized people.

Contributors

Primary Authors: SMH (lead author), GLK, and AJM (corresponding author): Project administration, conceptualization, writing–original draft, writing–reviewing and editing.

Coauthors: IA, CDB, AB, MLB, LXZB, MC, RSD, DE, JPH, WHJ, DH, MM, NRM, EO, GP, SS, and SY: Conceptualization, writing–original draft, writing–reviewing and editing.

Please note that the roles identified here were taken from the http://credit.niso.org/taxonomy.

Declaration of interests

SMH Declares work on this manuscript was made possible through support of successive grants from the National Institute on Disability, Independent Living, and Rehabilitation Research awarded to The Ohio State University Nisonger Center (grants 90RTHF0002-01-00 and 90RTHF0006-01-00). These grants fund my employment at The Ohio State University. Payments were/are made directly to The Ohio State University.

GLK Declares consultant contract with The Ohio State University through grants funded by the National Institute on Disability, Independent Living, and Rehabilitation Research (NIDILRR grants 90RTHF0002-01-00 and 90RTHF0006-01-00) (same grants as AJM); consultant contract paid to individual for minimal writing time; GLK also declares library access through appointment with Oregon State University.

AJM Declares work on this manuscript was made possible through support of successive grants from the National Institute on Disability, Independent Living, and Rehabilitation Research awarded to The Ohio State University Nisonger Center (grants 90RTHF0002-01-00 and 90RTHF0006-01-00). These grants fund my employment at The Ohio State University. Payments were/are made directly to The Ohio State University.

LXZB Consulted for Center for Health Care Strategies on Medicaid services for LGBTQ+ people with disabilities; delivered one lecture for Johns Hopkins University Bloomberg School of Public Health; received paid travel expenses from Manatt, Phelps & Phillips, LLP to attend one meeting regarding LGBTQ+ inclusive health care; and served as member of board of directors for Alliance for Citizen Directed Supports, Disability Power Bloc, and the Disability Rights Bar Association. Also was appointed to serve in an unpaid role on the Maryland Stakeholder Advisory Group on Autism-Related needs by the State Commissioner of Autism Strategy.

MC National Institutes of Health; Substance Abuse and Mental Health Addiction Services; National Institute on Disability, Independent Living, and Rehabilitation Research; Connecticut Department of Mental Health and Addiction Services; Cambria County Department of Mental Health. All payments from these organizations were made to institution (Yale University). Received travel support from NIH to attend the NIMHD Health Disparities Research Institute and to serve as co-investigator on an NIH grant in Brazil and also received travel support from NIDILRR to attend the IPS International Conference 2024. Also serves in unpaid position as member of Keep the Promise Coalition, guiding council in Connecticut, USA.

DE Declares support from the following grants: CDC-RFA--DD21-2103 “Disability and Health State Programs” (CDC); “From Awareness to Action: Strengthening Community Capacity to Improve the Health of Individuals with Mobility Limitations” (NCHPAD, CDC); “The Direct Support Professional Workforce: Empowering Ohio DSPs” (Developmental Disabilities Basic Support and Advocacy Grants); “Addressing Needs of People with Disabilities in COVID-19 Local Preparedness Planning, Mitigation, and Recovery Efforts” (NACCHO, Franklin County Public Health). Payments made directly to The Ohio State University.

SMH, IA, CDB, AB, MLB, RSD, JPH, WHJ, DH, MM, NRM, EO, GP, SS and SY Declares no competing interests.