Rosemary Collaboratory: A Policy, Systems, and Environmental Approach to Understanding and Addressing Health Inequities for People With Intellectual and Developmental Disabilities, United States

Abstract

Individuals with intellectual and developmental disabilities (IDD) face significantly poorer health outcomes and experience greater health disparities compared with people without IDD, in part because of health system failures.

This article summarizes health system assessments completed in the United States at the national level and in 3 states (Pennsylvania, Washington, and Wisconsin) to identify existing barriers and gaps. To reflect the US health architecture, there is also a discussion of particularly relevant national or federal policy features that influence health equity for people with IDD.

Despite existing federal legislation, there are significant opportunities for improvement across all components of the health system, at both federal and state levels. One of the areas of most significant state-level innovation has to do with supported decision-making and alternatives to guardianship, demonstrating that there is precedent for state leadership. States should continue to treat federal legislation as the floor of protection, rather than the ceiling, and pursue further reforms that promote health equity for people with IDD. (Am J Public Health. 2026; 116(S2):S103–S111. https://doi.org/10.2105/AJPH.2026.308452)

People with intellectual and developmental disabilities (IDD) die 16 to 20 years sooner than the general population.¹ In life, they are disproportionately affected by obesity, diabetes, heart and respiratory diseases, and mental health conditions, and these preventable, treatable, and manageable health conditions contribute to premature mortality.^2,3

Like the general population, people with IDD are affected by social drivers of health inequity, but they also face unique and additional barriers compared with those with other disabilities or those without disabilities. Stigma, inaccessible communication, and complex systems are 3 examples of such barriers.

The role that remediable structural and process aspects of health systems play in perpetuating these inequities is poorly understood. This article presents findings from health system‒level assessments conducted at the federal and national level in the United States and state level in Pennsylvania, Washington, and Wisconsin, identifying policy, system, and environmental gaps and opportunities to strengthen the health system and reduce inequities for people with IDD.

METHODS

In 2022, the Missing Billion Initiative published a Disability-Inclusive Health System Framework, synthesizing scholarly and policy literature into a System Level Assessment tool with 48 indicators across 4 system-level and 5 service-delivery components, plus output and outcome measures.⁴

Building on several decades of mission-driven work related to people with IDD, Special Olympics International collaborated with the Missing Billion Initiative to adapt this tool to better reflect the needs and barriers faced by people with IDD.⁵ Without deviating from the framework, the IDD-specific tool includes 8 additional system or service delivery and 7 outcome indicators—for example, national law on capacity and decision-making and national guidance on reasonable accommodation for people with IDD—which are indicated with “IDD” in the IDD module toolkit. The tool was further tailored for the United States, reflecting features such as federalism and the significant role of nongovernmental actors.

Special Olympics International assembled, oversaw, and collaborated with teams to apply the IDD System Level Assessment tool at the federal and national level and the state level in Pennsylvania, Washington, and Wisconsin. These states represent a convenience sample and may not be representative of the entire United States. Beginning in 2024, teams reviewed primary policy documents and relevant secondary research, and engaged stakeholders, including people with IDD, caregivers, health workers, and public officials.

Because various indicators in the tool relate to health status, it was also necessary to complement policy analyses with epidemiological analyses. The latter were completed on data from both the National Health Interview Survey (NHIS) and the National Core Indicators Survey on Intellectual and Developmental Disabilities (NCI-IDD), which represent different subsets of the population with IDD.

The NHIS, a general population survey, asks individuals about difficulties with the following tasks:

• Cognition (i.e., remembering or concentrating),

• Self-care (i.e., washing all over or dressing),

• Communication (i.e., understanding or being understood),

• Seeing (even when using glasses or contacts),

• Hearing (even when using a hearing aid), and

• Mobility (i.e., walking or climbing steps).⁶

Respondents are asked to describe the level of difficulty they experience in performing each task as (1) no difficulty, (2) some difficulty, (3) a lot of difficulty, or (4) cannot do at all.⁶ Those who report a lot of difficulty or being unable to perform 1 or more of these 6 tasks are asked, “You said that you have a lot of difficulty or are unable to do [1 or more tasks]. Did this difficulty begin before age 22?”⁶

Meanwhile, the NCI-IDD is a smaller survey administered by the Human Services Research Institute, a nonprofit research and evaluation organization that works with state Medicaid, aging, and disability agencies to randomly sample beneficiaries receiving IDD-related services, typically through a waiver program (discussed subsequently).⁷ The NCI-IDD has no data on individuals with IDD who receive insurance through their employer or a family member’s employer, who purchase private insurance from a state marketplace, or who rely exclusively on Medicare benefits.

It is difficult to identify people with IDD using national survey questions, such as those included in the NHIS, while those in the NCI-IDD data set have been identified as having IDD through much more extensive assessments.⁸ This is one reason for analyzing both NHIS and NCI-IDD data and, therefore, presenting data on the different subsets of the population with IDD that each data source captures. Nevertheless, the analyses conducted still present limitations.

For the purposes of the NHIS analyses, individuals were counted as respondents with IDD if they were adults aged 22 to 85 years who responded having “a lot of difficulty” or “cannot do at all” for activities of cognition, self-care, or communication, and that this limitation originated before the age of 22 years. This will still undercount those with IDD—for example, by excluding adults who reported a lower level of difficulty with cognition, self-care, or communication, some of whom may receive or have received special education, vocational support, or other IDD services.

Unless otherwise indicated, the NCI-IDD analyses use unweighted national averages for 2021 to 2023.

Even using this operating definition for NHIS analyses, there are likely to be significant differences between the individuals in the NHIS group as compared with the NCI-IDD group. For instance, the latter are, by definition, receiving some kind of IDD-related service, and eligibility for these services is limited to individuals with low incomes and assets.

RESULTS

We present findings and analysis from the System Level Assessments conducted at the federal and national level in the United States, and the state level in Pennsylvania, Washington, and Wisconsin. While the assessments were conducted independently of one another, in some cases, similar data were gathered from both the state and federal points of view, as federal laws, funding flows, and regulations influence, but do not determine, state structures. This section pays particular attention to indicators added to the System Level Assessment for people with IDD, as these illustrate where this population’s needs diverge from those of the larger population with disabilities. We start with outputs and outcomes to frame the urgency of addressing system gaps uncovered through other System Level Assessment components and discussed later.

Outputs and Outcomes

The findings from analyses of the NHIS and the NCI-IDD reinforce findings reported in existing literature that indicate that people with IDD experience poorer health than those without disabilities. Selected results of analyses are presented in Table 1. Across almost all presented indicators, the data show poorer health outcomes for people with IDD, regardless of population subset, vis-à-vis NHIS respondents with no limitations, though the magnitude of the disparity is variable. For instance, diabetes prevalence is 18.3% among NCI-IDD respondents, which is nearly 3 times the rate for NHIS respondents with no limitations (6.3%), while the prevalence among NHIS respondents with IDD was closer to 75% higher than that for NHIS respondents without IDD.

TABLE 1—

Comparing Sample Data on US Adults With IDD From the 2021–2023 National Core Indicators Survey (NCI-IDD) National Sample to Adults With IDD and Without Limitations From the 2023 National Health Interview Survey (NHIS)

	NCI-IDD Respondents, % (n = 36 012)	NHIS Respondents With IDD, % (n = 296)	NHIS Respondents With No Limitations, % (n = 14 591)
Age, y
22–34	39.5	48.6	24.1
35–44	20.6	20.6	21.3
45–54	14.8	12.8	16.2
55–64	14.6	8.1	16.5
65–74	8.1	5.4	14.9
75–85	2.5	4.4	7.0
Health insurance coverage
Medicaid only	33.4	51.4	9.5
Medicare only	0.1	27.0	21.1
Medicaid and Medicare	37.5	14.9	1.4
BMI
Underweight	4.9	2.7	1.3
Healthy weight	29.3	28.7	31.8
Overweight	28.2	24.3	36.1
Obese	37.5	43.2	28.9
Health conditionsa
Diabetes	18.3	11.1	6.3
Anxiety disorder	22.2	60.8	11.4
Mood disorder	26.4	60.8	11.2
Asthma	11.0	32.1	11.8
Dementia	2.9	2.7	0.1
Cardiovascular disease (CHD, angina)	7.7	3.4	2.7
Hypertension	28.4	33.1	27.7
Health care utilization
Visited physician in past yearb	85.2	88.2	83.0
Had routine dental examination in past year	75.8	50.3	70.5
Went to the emergency department in past year	24.6	35.5	15.0

Note. BMI = body mass index (weight [in pounds] / height [in inches]² × 703); CHD = Coronary heart disease; IDD = intellectual and developmental disabilities.

Source. Adapted from Kennedy J, Dubois L. Rosemary Collaboratory USA: National and State-Level Analyses of IDD Health and Healthcare Disparities. Unpublished report. 2025.

^a NCI-IDD asks about “ever diagnosed with”; NHIS asks “noted in health records.”

^b NCI-IDD asks about “annual physical/examination with primary care provider” versus NHIS asks about “visited a doctor in the past 12 months.”

Other components and indicators in the System Level Assessment aim to identify some of the health system factors that contribute to these disparities.

System Component

The system component of the System Level Assessment includes indicators related to governance, leadership, health financing, and data and evidence.

Governance

Within the Disability-Inclusive Health System Framework, the indicators related to governance aim to capture laws, policies, and accountability mechanisms that “protect the right to health care for people with disabilities and outlaw discrimination based on disability.”⁴ The relatively robust federal legal and institutional framework in place when data were gathered—namely, the Americans with Disabilities Act (ADA), Section 504 of the Rehabilitation Act, the Patient Protection and Affordable Care Act, and the Developmental Disabilities Assistance and Bill of Rights Act of 2000 (DD Act)—represent a crucial foundation for disability inclusion in health in the United States. Together, they address critical considerations such as nondiscrimination, accessibility, cost protections, benefit and coverage adequacy, data gathering, and self-determination.

Although federal laws like the ADA and the DD Act apply directly in every state, they contain significant gaps of their own, which are then present at the state level. For example, federal agencies have provided little interpretive guidance on accommodations under the ADA for people with IDD. Communication is one area in which individuals with IDD often require accommodations; however, ADA guidelines on communication primarily focus on ensuring accessibility for individuals with hearing or vision impairment and on use of universal accessibility symbols in posted signage.^9,10 While many people with IDD can benefit from accommodations such as audio instructions and communication boards, many people with IDD require simple language in conjunction with other accommodations. Within the states examined, there has been little state action to supplement federal protections with state-level protections for people with IDD in this arena.

One of the indicators added to the System Level Assessment for people with IDD relates to protections of legal capacity and informed decision-making. Guardianship, “a state legal process where a court removes some or many of the legal and decision-making rights from an individual and transfers all or some of them to another person,” can be at odds with protections of legal capacity, yet it is permitted in all states and remains widespread.¹¹ One study identified through the assessment found that an average of 40% of people with IDD in a US sample of more than 25 000 people with IDD had a guardianship arrangement, while another study estimated that more than 1 million Americans with IDD are under guardianship.^12,13

Guardianship is a form of substitute decision-making, and the disability rights community has long raised concerns that it is in tension with core principles of the ADA and DD Act, such as nondiscrimination and self-determination.¹¹ Alternative models, called supported decision-making, are becoming more widespread at the state level: the vast majority of states require courts to consider less restrictive options for supported or shared decision-making before appointing a guardian.¹⁴ However, as of 2021, only in approximately one quarter of US states did laws specifically mention or recognize supported decision-making.¹⁴ Of the states examined, Pennsylvania law requires consideration of less restrictive alternatives to guardianship, without promoting supported decision-making, while both Washington and Wisconsin explicitly enable or promote supported decision-making.^15,16

Leadership

The indicators related to leadership assess the extent to which issues around disability or individuals with disabilities themselves are formally embedded in the governmental agencies, structures, and mechanisms responsible for health issues.^4,5 The Disability Rights Movement’s demand, “Nothing About Us Without Us,” proclaims that people with disabilities should be on board, if not at the helm, in decision-making that affects them. This shaped the legislative framework and its effect endures: encouragingly, at the national level as in the 3 states examined, formal avenues or mechanisms for people with IDD to participate in policy processes related to disability inclusion were present. This was largely attributable to the creation and financing of Developmental Disabilities Councils (DD Councils) in each state through the DD Act.¹⁷

DD Councils focus on the needs of people with developmental disabilities on a wide range of topics including, in some cases, health. For instance, health appears to be a minor part of the Wisconsin Board for People with Developmental Disabilities State Plan 2022–2026 but is very prominent in Washington State’s DD Council 2022–2026 State Plan, with Pennsylvania’s DD Council 2022–2026 State Plan falling somewhere in between.^18–20

At the federal level, the President’s Committee for People with Intellectual Disabilities included people with IDD and advised the president and the secretary of the Department of Health and Human Services on policy and programs. Similarly, Washington and Wisconsin both have an active Governor’s Committee for People with Disabilities, which includes people with disabilities and addresses health, among other issues, while Pennsylvania has had such a body and recent efforts (PA SB 592) to revive it. In a further step that stands to promote the involvement of people with IDD in policymaking, Washington passed the Nothing About Us Without Us Act (Washington State Legislature, HB 1541, 2024). It requires that all state task forces, work groups, and advisory committees working on issues affecting underrepresented populations include at least 3 individuals with direct lived experience with the issue.²¹

While the structures mentioned often include people with IDD alongside representatives of government agencies, it is worth noting that this institutional architecture resides largely outside the health system. There is, nevertheless, progress toward mentioning people with disabilities within population health policies. For example, in Healthy People 2030, 34 objectives refer to people with disabilities in general, and several national-level disease- and topic-specific plans mention people with disabilities as a disparity group.^22,23 The assessment identified 2 specific mentions of people with IDD in national population health policies: one in Healthy People 2030, related to deinstitutionalization, and the other in the National Plan to Address Alzheimer’s Disease, which, from the outset (2012), included objectives related to people with Down syndrome, and by 2015, expanded to people with IDD more generally.^24,25

Similarly, the Pennsylvania Department of Health State Health Improvement Plan, 2023–2028 and Wisconsin State Health Improvement Plan, 2023–2027 Implementation Guide both include some discussion of people with disabilities generally, but few targets related to people with disabilities, and none related to those with IDD specifically.^26,27

Health financing

Health financing indicators aim to assess the extent to which governments have adopted financing approaches that take health services for people with disabilities specifically into account. Before applying the assessment, anecdotal information from people with IDD, caregivers, and health workers suggested that the indicators related to health financing, including reimbursement adjustments to service providers for care provided to people with IDD, were of particular interest. However, it remains exceedingly difficult to make blanket statements about health-related financing for people with IDD because of the number of relevant variables, such as Medicaid program waivers that offer states “additional targeted flexibility to test new approaches to service delivery” and different billing approaches.²⁸

The assessments revealed that, according to fiscal year 2021 data, federal spending for services and supports for people with IDD in 2021 amounted to $80.6 billion, 87% of which was distributed through Medicaid-related services.²⁹ Pennsylvania and Wisconsin were among the states with relatively higher percentages of the state budget spent on services and support for people with IDD, while Washington’s proportion was lower but close to the national average.²⁹ Wisconsin also adopted (but subsequently did not fund nor implement) a unique measure to enhance Medicaid reimbursement rates for dental services rendered by providers that provide at least 90% of their dental services to individuals with cognitive and physical disabilities to equal 200% of the Medicaid reimbursement rate that would otherwise be paid for these dental services.³⁰

Pennsylvania and Washington both use waivers to offer Home and Community Based Services (HCBS) for people with IDD. However, people with IDD are not categorically eligible for IDD-specific HCBS in those or other states. Individuals with IDD may be eligible for waivers that are not specific to people with IDD, and there is no guarantee that eligible individuals will receive HCBS because of waitlists. The services covered by HCBS vary both between and within states because of differences between waivers.

Data and evidence

Data and evidence—namely, the collection of data on people with IDD using validated measures—comprised indicators in the system component of the System Level Assessment.⁴ There are relevant data at the national level and in all 3 states from the Behavioral Risk Factor Surveillance System, the NHIS, and the NCI-IDD. However, while the Behavioral Risk Factor Surveillance System and NHIS include validated sets of measures, the questions are both over- and underinclusive in capturing the population with IDD.³¹

Service Delivery Component

The service delivery component of the System Level Assessment includes indicators related to autonomy and awareness, affordability, human resources, health facilities, and rehabilitation services and assistive technology.

Autonomy and awareness

Within the Disability-Inclusive Health System Framework, the indicators related to autonomy and awareness aim to assess the extent to which people with disabilities are aware of their rights and options for making decisions about their own health care.⁴ Health information is an important element of autonomy and awareness. Section 508 of the Rehabilitation Act of 1973 and the Plain Writing Act of 2010 both promote accessible information. Furthermore, as of 2024, federal regulations for Title II of the ADA require states to adhere to Technical Standard Web Content Accessibility Guidelines (WCAG) 2.1, Level AA standard, at a minimum.³² Users with cognitive or learning disabilities—a group that includes many people with IDD—were 1 of 3 groups for whom WCAG 2.1 sought to improve accessibility guidance, vis-à-vis WCAG 2.0.³³

However, this indicator is one for which there is a gap between policy and practice, and assessments found only a fraction of health information available on government Web sites to be available in plain language. Caregiver-targeted information was more prevalent than materials for people with IDD themselves, limiting empowerment and engagement.

Affordability

Affordability indicators aim to assess the extent to which governments have tried to reduce or remove cost-related barriers to health care for people with disabilities.⁴

The federal Supplemental Security Income (SSI) program provides financial assistance to low-income individuals with disabilities and seniors. Although Pennsylvania, Washington, and Wisconsin are among the vast majority of states that provide a state supplement to the federal SSI payment, in 2025, the maximum monthly federal SSI payment for an individual was slightly under 75% of the Federal Poverty Guideline for a household size of 1.³⁴

Findings from the NHIS analyses corroborate existing literature showing that people with IDD are more likely than those without any limitations to be on a fixed, low income. The findings presented in Table 2 show that 19.9% of NHIS respondents with IDD had a family income below the federal poverty level—almost 3 times the rate of NHIS respondents without IDD. Similarly, benefiting from a federal nutrition program was much more common among NHIS respondents with IDD (33.4%) compared with those without IDD (9.1%).

TABLE 2—

Differences in Socioeconomic Status and Program Participation Between US Adults Aged 22–85 Years With IDD and Adults Without Functional Limitations

	NHIS Respondents With IDD, Estimated No., Millions (%)	NHIS Respondents With No Limitations, Estimated No., Millions (%)	Wald χ2	P
Total	2.9 (100)	131.4 (100)
Family income < federal poverty level34	0.6 (19.9)	9.7 (7.4)	55.2	< .001
Received SNAP or WIC	1.0 (33.4)	12.0 (9.1)	103.2	< .001

Note. IDD = intellectual and developmental disabilities; NHIS = National Health Interview Survey; SNAP = Supplemental Nutrition Assistance Program; WIC = Special Supplemental Nutrition Program for Women, Infants, and Children.

Source. Adapted from Kennedy J, Dubois L, Rosemary Collaboratory USA: National and State-Level Analyses of IDD Health and Healthcare Disparities. Unpublished report; 2025. Data Source. 2021, 2022, and 2023 National Health Interview Surveys, Sample Adult Files.

In most states, SSI beneficiaries are automatically eligible for Medicaid. The analyses of NHIS and NCI-IDD data described previously and summarized in Table 1 found that approximately one third to one half of people with IDD are insured through Medicaid. Medicaid provides access to some services that are unique; for example, people who are insured through Medicaid have access to Medicaid’s Non-Emergency Medical Transportation benefit, which is a mandatory plan benefit to ensure access to medically necessary care. Unfortunately, many challenges persist with respect to wait times, trip denials, missed pickups, and communication barriers.³⁵ Furthermore, individuals who may provide support to people with IDD are not necessarily eligible for subsidized transportation, which can make transport subsidies less effective at addressing transportation barriers to health care for people with IDD.

Issues with affordability are particularly prominent for services that are not covered by health insurance, which may be at least a partial explanation for the differences across population subsets in who had a routine dental examination in the past year (75.8% among NCI-IDD respondents, 50.3% among NHIS respondents with IDD, and 70.5% among NHIS respondents with no limitations).

Pennsylvania has implemented a good practice to streamline the process by which individuals could obtain dental care beyond the basic benefit limits.³⁶ As a result of this change, individuals with intellectual disability or 4 other qualifying conditions no longer need to present additional documentation to support a Benefit Limit Exception.³⁶

Human resources

Within the System Level Assessment, the indicators related to human resources largely probed the sufficiency of the health workforce, as well as the existence and extent of training requirements for the health workforce related to disability. Section 5307 of the 2010 Patient Protection and Affordable Care Act required the development, evaluation, and dissemination of disability curricula for inclusion in health professional training; however, this is not specific to IDD, and the status of implementation is uncertain.³⁷ The Alliance for Disability in Health Care Education published core competencies in the care of people with disability that were developed by a broad coalition of researchers and advocates, including persons with disabilities, and they have been validated across health professions and disability types.³⁸ A variety of curricula, including Special Olympics’ Inclusive Health Fundamentals, exist, but they are not systematically taught in medical or other health science schools.

Two particularly interesting policies are in place in Washington State. Washington’s established curriculum for community health workers previously included content related to disability.³⁹ Though unclear if it was sufficient to prepare workers to provide quality care to people with IDD, other core course learning objectives include plain language communication, using service coordination and navigation strategies, and understanding social determinants of health, all of which are crucial to deliver quality care to people with IDD.³⁹ Washington State also requires that health care professionals complete a minimum of 2 hours in health equity continuing education training at least once every 4 years.⁴⁰ Although the text of this act does not focus on people with disabilities in general or IDD in particular, it provides an interesting precedent, and the justifications for the law are applicable to people with IDD.

One of the indicators that was added into the System Level Assessment for people with IDD relates to the existence of a workforce to assist in care coordination. Care coordination is within the scope of a variety of roles throughout the US health care delivery system, with availability varying. Examples include early intervention under Part C of the Individuals with Disabilities Education Act, health homes created as an optional state Medicaid benefit, innovation models, and §1915(c) Medicaid waivers to offer HCBS. Consistent with this, in the state contexts, the assessments identified individuals with care coordination roles at community-based organizations, managed care organizations, and waiver programs. However, most of these roles do not have specific responsibilities for or are related to people with IDD, nor do all people with IDD have access to care coordination services.

Health facilities

The indicators of the System Level Assessment related to health facilities primarily focus on health facility accessibility. Here, the ADA was the predominant reference, but limited information was available about the extent of health facilities’ actual compliance with the ADA. ADA compliance or accessibility audits were not a widespread practice in any of the 3 states where the assessment was completed. Other relevant standards or assessments, such as those promulgated by the Joint Commission or the American Hospital Association, have limited content related to accessibility by those with disabilities.

Assistive technology and rehabilitation

The final indicators in the service delivery component of the Missing Billion Initiative disability-inclusive health system framework relates to assistive technology and rehabilitation. The objective of these indicators is to assess the availability, affordability, and quality of specialized services that are used by the general population but are in much wider use among people with disabilities.⁴

DISCUSSION

The System Level Assessment revealed numerous areas in which individual states and the United States can act to improve how the health system functions for people with IDD. While the outputs and outcomes data continue to be an imperfect representation of people with IDD as a population group, they capture individuals who receive some of the most robust health and related supports on offer (NCI-IDD population) as well as those with the highest functional limitations (NHIS respondents with IDD). These analyses represent a wide range of individuals within the population with IDD and add to the evidence base that documents significant unmet health needs and poorer health compared with the general population. The system and service delivery indicators are useful for isolating where states have taken action versus where they have relied on federal and national authority, such as the ADA. Recognizing the complexity of powers and responsibilities in the federal system, state action on the issue of guardianship shows promising state-level leadership and innovation that might be extended to other indicators. The initial round of assessments discussed in this article provide learnings not only about the states where the assessment was applied but also about the System Level Assessment tool itself, which can inform refinement of the tool. Applying this approach more broadly in the future could help identify health system barriers and gaps for people with IDD in other states and communities, ultimately supporting broader efforts to improve health equity for individuals with IDD.