The Social Security Administration (SSA) administers two programs that provide benefits based on disability: the Social Security Disability Insurance (SSDI) program and the Supplemental Security Income (SSI) program. The SSDI program provides disability benefits to people who are under the full retirement age and are no longer able to work because of a disabling medical condition or because they have a terminal illness. The SSI program is a means-tested income-assistance program for disabled, blind, and aged people who have limited income and resources regardless of their prior participation in the labor force. The programs have a common definition of disability in adults—“the inability to engage in any substantial gainful activity by reason of any medically determinable physical or mental impairment which can be expected to result in death or which has lasted or can be expected to last for a continuous period of not less than 12 months”—and a common disability-determination process administered by SSA. Disabled workers might receive either SSDI benefits, SSI payments, or both, depending on their recent work history and current income and assets. They might also receive benefits from other public programs, such as worker’s compensation, which insures against work-related illness or injuries occurring on the job, but the other programs have their own definitions and eligibility criteria.1
In the present report, the Committee on Health Care Utilization and Adults with Disabilities discusses the SSDI and SSI programs because they both use the Listing of Impairments2 (also referred to as the Listings) in their determination process.3 The Listings includes hospitalization as a criterion for receiving disability insurance for health conditions of six body systems: respiratory, cardiovascular, digestive, genitourinary, hematologic, and immune. They require an applicant to have, for a particular medical condition or combination of conditions, “exacerbations or complications requiring three hospitalizations within a 12-month period and at least 30 days apart” of which “each hospitalization must last at least 48 hours, including hours in a hospital emergency department immediately before the hospitalization.”
The committee sought to understand the SSA programs and disability-determination process: SSA tasked the committee with providing an analysis of health-care utilizations as they relate to impairment severity and SSA’s definition of disability. The committee sought to identify types of utilizations that might be good proxies for listing-level severity, that is, what represents an impairment, or combination of impairments, severe enough to prevent a person
1 There is no federal role in state worker compensation. State compensation programs vary widely with regard to coverage, benefits, and administrative practices (Social Security Bulletin, Volume 65, No. 4, 2004).
2 See https://www.ssa.gov/disability/professionals/bluebook/AdultListings.htm, accessed February 1, 2018.
3 See 20 Code of Federal Regulations (CFR) Part 404, Subpart P, Appendix 1.
from doing any gainful activity regardless of age, education, or work experience. The committee’s full statement of task is provided later in this Summary.
THE LISTING OF IMPAIRMENTS
SSA has a five-step sequential process to determine whether someone is medically eligible for SSDI or SSI benefit. The third step of the sequential evaluation process relies on the Listing of Impairments4 to identify cases that can be allowed regardless of an applicant’s age, education, or work experience. The Listings are organized by 14 body systems for adults (see Table S-1) and includes for each system impairment that SSA considers severe enough to prevent an adult from performing any gainful activity. The SSA Office of the Inspector General noted in 2015 that “the Listings help ensure that disability determinations are medically sound, claimants receive equal treatment based on the specific criteria, and disabled individuals can be readily identified and awarded benefits, if appropriate.” Applicants whose impairments do not meet or medically equal a Listing can still be determined to be disabled at step 5 on the basis of the combination of their residual functional capacity, age, education, and work experience.
Although an established “listing of medical impairments” has existed since the disability program began in 1956, SSA did not publish the Listings in its disability regulations until 1968.5 Since then, it has revised the Listings periodically to reflect advances in medical knowledge. In 2003, SSA implemented a new process for revising the Listings, which was designed to ensure continuous updates and monitoring of the Listings about every 3–4 years.
TABLE S-1 Body Systems in Social Security Administration Listings for Adults
|1.0 Musculoskeletal system|
|2.0 Special senses and speech|
|3.0 Respiratory disorders|
|4.0 Cardiovascular system|
|5.0 Digestive system|
|6.0 Genitourinary disorders|
|7.0 Hematologic disorders|
|8.0 Skin disorders|
|9.0 Endocrine disorders|
|10.0 Congenital disorders that affect multiple body systems|
|11.0 Neurologic disorders|
|12.0 Mental disorders|
|13.0 Cancer (malignant neoplastic diseases)|
|14.0 Immune system disorders|
SOURCE: Disability Evaluation Under Social Security, SSA, September 2008.
4 Found in 20 CFR Part 404, Subpart P, Appendix I.
Seven body systems were last updated during 2009–2015. The SSA Office of the Inspector General recommended that by the end of fiscal year (FY) 2020 SSA should ensure that all Listings updates are less than 5 years old and that SSA continue to update them as needed to reflect current medical knowledge and advances in technology. Four more body systems (respiratory, neurologic, mental, and immune-system disorders) were updated in 2016–2017.
STATEMENT OF TASK
SSA asked the National Academies of Sciences, Engineering, and Medicine to conduct a study of health-care utilizations by adults being evaluated for disability determination. The statement of task follows:
An ad hoc committee will provide an analysis of health care utilizations as they relate to impairment severity and SSA’s definition of disability. The committee will identify types of utilizations that might be good proxies for “listing-level” severity; that is, what represents an impairment, or combination of impairments, that are severe enough to prevent a person from doing any gainful activity, regardless of age, education, or work experience.
The task order objectives for the ad hoc committee are to
- Provide a general description of the health care delivery system (for example, the Affordable Care Act, move to patient-centered care, medical homes, bundled payments, and other significant changes in the delivery of care);
- Identify factors that influence a person’s use of health care services;
- Identify health care utilizations that represent, and are a good indicator of impairment severity for the purposes of the disability program;
- Explain how intervals between utilizations and duration of utilizations affect whether health care utilization is a good indicator of impairment severity;
- Explain how types of utilizations are more or less probable for particular medical conditions or combinations of medical conditions;
- Describe how factors such as poverty and urbanization level affect health care utilizations; and
- Describe how the health care utilizations interfere with a person’s ability to work.
During the committee’s meetings with SSA representatives, it came to understand that the study’s focus should not be on the content of the current Listing of Impairments, nor should it be a critique of SSA’s evidentiary requirements. Furthermore, the committee was not tasked with examining functional outcomes or other objective performance measures. SSA does use such information when conducting assessments at steps 4 and 5 of the disability-determination process. This report, however, is focused on step 3 of the process, that is, “whether the impairment ‘meets’ or ‘medically equals’ one of the items on the Listing of Impairments.” In developing the statement of task, SSA provided numbered objectives (set forth above) in an effort to provide needed guidance. SSA noted that it did not require recommendations.
THE COMMITTEE’S APPROACH TO ITS TASK
A 16-member committee was formed to address the task. Members with diverse backgrounds and expertise were appointed to focus on the different aspects of the task. The committee met five times. It sponsored two open meetings, which enabled SSA representatives and the committee members to interact directly and discuss the motivation for the study. The committee also invited people to address health-care trends and different viewpoints on the measurement of disability. The staff conducted targeted literature searches and gathered information from relevant scientific, professional, and federal sources. After review of more than 60,000 titles and abstracts, about 700 studies were selected for further review and relevance to the committee’s task.
In an effort to approximate impairment severity, the committee reviewed the literature to consider to what extent and in which ways health-care utilization might reflect disease severity, disability, and ability to perform gainful activity. The committee recognized that varied access to health care potentially confounds any observed relationships, and it sought to identify predictors of health-care utilization to assist in understanding the context of the relationship between receipt of health-care services and ability to work. The committee also considered the many types of health-care utilization, the sites where they are delivered, and whether and how they might be associated with severity of impairments.
FACTORS THAT AFFECT HEALTH-CARE UTILIZATION
People use health-care services to diagnose, cure, or ameliorate disease or injury; to improve or maintain function; or even to obtain information about their health status and prognosis. In theory, health-care utilization should correlate highly with need, but some services are needed and not obtained, and others are utilized but not clearly indicated or are indicated only after other protocols are followed.
Health status and need for health care to improve or maintain health are major determinants of health-care utilization. Ideally, need should be the major determinant of healthcare utilization, but other factors clearly are involved, such as the ability to access care—including whether it is available, timely and convenient, and affordable.
There are many conceptualizations of access, and the term access is often used to describe factors or characteristics that influence the initial contact with or use of services. Access is also a continuum: even if care is available, many factors can affect ease of access to needed care, for example, the availability of providers who will accept a person’s insurance (including Medicaid), the ease in making an appointment with a given provider, the ability of a patient to pay for care (even if insured, because of cost-sharing copayments and deductibles), and how difficult it is to arrange transportation to and from health-care facilities.
Many factors affect health-care utilization independently of need. Some of the factors reflect biologic or environmental differences among population groups, such as disproportionate residence in polluted environments, access to healthful food and adequate housing, and education associated with more effective use of health care. Other differences in health-care utilization reflect differential access, such as access to health insurance coverage or to income needed to obtain services, ease of obtaining services, or discriminatory practices by providers. A few examples are discussed below.
Race and Ethnicity
Racial and ethnic disparities are found in many sectors of American life. For example, black people, people of Hispanic origin, American Indians, Pacific Islanders, and some Asian Americans are disproportionately represented in the lower socioeconomic ranks, in lower-quality schools, and in poorer-paying jobs. Those factors can result in some racial and ethnic minorities experiencing higher rates of chronic and disabling illnesses and infectious diseases and higher mortality than white Americans. Minorities have more difficulty than the majority population in locating a “usual source” of medical care, and black and Hispanic adults report greater difficulty than whites in obtaining medical care at a consistent location.
The 2016 annual National Healthcare Quality and Disparities Report (QDR) highlights continued disparities in care that are experienced by members of different racial and socioeconomic groups. The 2016 QDR states that disparities decreased from 2000 through 2014–2015 but that they persist, especially for poor and uninsured populations. Furthermore, although 20 percent of measures show disparities decreasing for black people and people of Hispanic origin, most disparities have not changed substantially for any racial and ethnic groups. More than half the measures show that poor and low-income households had worse care than high-income households; more than 40 percent of the measures show that middle-income households had worse care than high-income households. Nearly two-thirds of the measures show that uninsured people had worse care than privately insured people.
Many Americans in racial and ethnic minorities experience language barriers and have low or no proficiency in speaking, reading, or comprehending English. In a health-care setting, those barriers can present serious challenges to both patients and providers. A systematic review of studies of language barriers in health care of Latino populations showed that access to care, quality of care, and health status all suffer as a result of language barriers. If a patient does not speak the language of his or her health-care provider, multiple adverse effects on the patient’s health care might occur. For example, a patient’s inability to understand a provider’s diagnosis or treatment plan can lead to poor patient satisfaction, poor compliance, and underuse of services.
Patients who face language barriers are less likely than their English-speaking counterparts to have a usual source of medical care, receive fewer preventive services, and have a greater likelihood of nonadherence to medication. Of patients who have psychiatric conditions, those who encounter language barriers are more likely than others to receive a diagnosis of a severe psychopathologic condition—but are also more likely to leave the hospital against medical advice. Such patients are less likely than others to return for followup appointments after visits to an emergency department (ED), and they have higher rates of hospitalization and drug complications.
Income and Poverty
Income correlates highly with risk factors for chronic disease. People who have lower family income typically have higher rates of heart disease, stroke, diabetes, or hypertension. People in families that have income less than 200 percent of the federal poverty level are more likely to be obese and to smoke cigarettes than wealthier people. Adults who live in poverty are
also more likely to have self-reported serious psychologic distress, as measured by a series of questions about their perceived mental health.
In all racial and ethnic groups, people who are 18–64 years old and have family income below 200 percent of the poverty level were more likely than those who had higher family income to delay seeking or not to receive needed medical care because of cost in 2009–2010. During that period, the percentage who had unmet needs for medical care decreased as family income increased from below 200 percent of the poverty level to 400 percent or more of the poverty level. Needs for medical care that were unmet because of cost were equally likely in those whose family income was below the poverty level and in those whose family income was 100–199 percent of the poverty level. Adults who were living under the poverty level reported greater rates of not receiving or of delaying seeking medical care, obtaining prescription drugs, and receiving dental care because of cost than adults who were living at 400 percent of the poverty level.
Level of urbanization has been shown to be associated with health-care utilization in several ways, including correlation of residents’ sociodemographic characteristics with need and risk factors and with differential access to care. Communities at different urbanization levels differ in their demographic, environmental, economic, and social characteristics. Those characteristics influence the magnitude and types of health problems that communities face. In addition, more-urban counties tend to have a greater supply of health-care providers per capita, and residents of more-rural counties often live farther from health-care resources.
People with Disabilities
Disability is a multidimensional concept. Some health conditions associated with disability result in poor health and extensive health-care needs, but others do not. The population is clinically diverse. Some have multiple chronic conditions that are stable with treatment and will persist for years; others have extreme functional limitations. Some have mostly severe, persistent behavioral-health challenges; others have conditions that are greatly exacerbated by social factors, such as lack of housing, food, and supportive relationships. The term disability implies that there are limitations to a person’s ability to function in one or more ways, but types of impairment also differ substantially and might differ in the extent to which they limit a person’s ability to function. A person who has a disability might have
- An impairment in body structure or function or in mental functioning; examples of impairments include loss of a limb, loss of vision, or memory loss.
- An activity limitation, such as difficulty in seeing, hearing, walking, or problem-solving.
- Restrictions in participation in normal daily activities, such as working, engaging in social and recreational activities, and obtaining health-care and preventive services.
Disability can be caused by or related to, for example, genetic disorders, injuries, illness, or environmental exposures. It might be temporary, progressive, or intermittent. Thus, it is difficult to generalize about the relationship between disability and health-care utilization without targeting specific causes, diseases, or conditions. However, all people who have
disabilities have the same general health-care needs as everyone else and therefore need access to mainstream health-care services.
People with disabilities use more health-care services than people who do not, in part because many have greater need for medical and health services to treat their conditions. However, people with disabilities face a number of barriers to access to health care that are specific to their individual limitations in function. These include barriers to physical access and lack of working elevators or ramps, automatic doors, hallways and doors wide enough to accommodate wheelchairs, and accessible parking; policies that discriminate against people with disabilities as opposed to policies—in place and enforced—that are designed to accommodate people with disabilities; and lack of accommodation for communication with people with disabilities, including large-print materials, sign-language interpreters, and staff willing to try to communicate with impaired patients during scheduling or other interactions.
CHANGING PATTERNS OF HEALTH INSURANCE AND HEALTH-CARE DELIVERY
In addition to examining factors that influence access to health care, the committee was tasked with providing a general description of the health-care delivery system. Although the system is in flux, the committee did its best to summarize the current state of health care.
Health care in the United States is financed by a combination of public and private insurance, employers, and individuals paying out of pocket. In 2015, 37 percent of the US population paid for health care through a public insurance program at some point during the year. The major public insurance systems are Medicare and Medicaid. In 2016, Medicare benefit payments totaled $675 billion and accounted for 15 percent of the federal budget.
The US health-care delivery system consists of an array of clinicians, hospitals and other facilities, insurance plans, and purchasers of health-care services, all operating in various configurations of groups, networks, and independent practices. The health-care delivery system historically has been organized around the concept of fee-for-service medicine. In the fee-for-service payment model, patients (or their insurers) pay physicians and hospitals for any covered services delivered on a per-unit basis, without particular regard for price, patient outcome, or quality. Provider revenues increase as more services are provided—and insured (and some uninsured) patients do not bear the full cost of the additional services—so the fee-for-service model creates incentives to increase utilization of health-care services, and this leads in many cases to overutilization of physician and hospital visits.
The Patient Protection and Affordable Care Act (ACA) was the largest federal health policy initiative since the creation of Medicare and Medicaid. It brought about structural changes in the health-care system, which included efforts to improve access to health insurance through expansion of the Medicaid program and through subsidized and lower-cost health insurance plans made available through new health insurance marketplaces (exchanges), elimination of restrictions on coverage for pre-existing conditions, elimination of lifetime caps on health-care spending, and efforts to slow growth in health-care costs through innovative payment reforms.
A major goal of the ACA was to extend health insurance coverage to 32 million uninsured people in the United States. The plan had two major components: expansion of the Medicaid program and new structures to support the individual and small-group health insurance markets. By 2015, the uninsured rate for people who were 19–64 years old was 16 percent lower
than it had ever been, and around 30 million people who would otherwise have been uncovered had coverage from private insurance and Medicaid.
The ACA eliminated the concept of categorical eligibility for Medicaid and replaced it with a standard eligibility criterion of 138 percent of the federal poverty level. In 2012, the Supreme Court ruled that the federal government could not force the states to expand Medicaid coverage. As a result, only 32 states and the District of Columbia elected to expand Medicaid. One estimate shows that around 17.5 million people would not have been covered by Medicaid had it not been for the ACA.
For the individual and small-group markets, the ACA established health insurance exchanges in states to allow individuals and small groups to buy standard insurance policies with income-based subsidies from 138 percent to 400 percent of the federal poverty level. The ACA eliminated medical underwriting and imposed a legal mandate to purchase health insurance, with a penalty for those who did not comply. Before the ACA, insurance companies used medical underwriting to determine whether to offer a person coverage, at what price, and with what exclusions or limits based on health status; the purpose was to ensure a healthy risk pool by requiring people to pay premiums that reflected their expected medical costs. Because of medical underwriting in the individual and small-group markets, people who were sick often paid higher premiums or were denied coverage. The ACA’s individual mandate, in contrast, was intended to compel healthier people to purchase insurance and thereby to balance the risk pool and lower premiums for everyone.
The ACA included payment-reform provisions to incentivize the adoption of more effective care-delivery models. The new models involve some combination of risk-sharing among providers to enhance collaboration and coordination of care in an effort to reduce avoidable hospitalizations, ED visits, and other forms of expensive or unnecessary care. To protect against stinting, quality metrics are often used to evaluate provider performance. Beyond payment models, the ACA encouraged the narrowing of provider networks (perhaps unintentionally) and reshaped the delivery of long-term services and supports, all of which have implications for how people who have disabilities receive care and for the documentation of that care in the medical record.
The ACA has many provisions that are important for people who have disabilities. For example, denial of coverage because of pre-existing conditions is no longer allowed, and the expansion of health insurance coverage through the Medicaid program, the health insurance exchanges, and the dependent-coverage provision will allow many Americans who have disabilities to obtain health insurance coverage without having also to qualify for SSDI or SSI.
HEALTH-CARE UTILIZATIONS AS PROXIES FOR LISTING-LEVEL SEVERITY
The committee’s extensive literature review found no studies that addressed the usefulness of health-care utilizations in determining disability or impairment severity and few that addressed the association of health-care utilization with disability. The question of whether health-care utilizations can be proxies for disability or impairment severity according to SSA’s definition has not been extensively researched in the health sciences. In the absence of such data, the committee reviewed literature that links health-care utilizations to similar concepts, such as disease severity and ability to work.
The committee found no evidence that health-care utilizations alone can predict disability, impairment severity, or disease severity. For several medical conditions, including chronic obstructive pulmonary disease (COPD) and chronic kidney disease, there is some evidence that increased hospitalizations, ED visits, and outpatient physician visits might predict disease severity for some specific diagnoses. However, their relevance to the committee’s task is limited in that disease severity does not fit SSA’s definition of impairment severity and statistical modeling in the supporting papers involved more factors than health-care utilization, such as individual and societal factors that influence the use of health care.
Another intervening variable is the presence of comorbid conditions. In particular, psychiatric disorders were found to increase the likelihood of disability and of the use of healthcare services for medical conditions of several body systems.
With few exceptions, the health-care utilizations featured in the committee’s literature review were hospitalizations, ED visits, and outpatient physician visits. Given that annual data on hospitalizations have been collected in the United States since 1965, hospital data are easy to capture and are more likely than data on other utilizations to constitute a reliable measure of impairment severity associated with some diseases. The fragmented nature of the health-care delivery system, however, makes it difficult to capture all the different types and locations of utilizations for purposes of determining disease severity.
In addition, little research has attempted to account for the relationships among employment, health, and disability. Some studies have explored effects of health on disability or vice versa but rarely in a manner that acknowledges that health problems interfere with the ability to work.
A PROXY FOR DETERMINING LISTING-LEVEL SEVERITY
Inasmuch as there is scant literature on health-care utilizations that would be good proxies for listing-level disability, the committee considered “What would make a good proxy?” Listing-level severity—defined by SSA as the inability to perform any gainful activity regardless of age, education, or work experience—is not easy to measure in practice. It has been noted, in some studies, that the disability-determination process provides applicants with a strong disincentive to work inasmuch as any (substantial) gainful activity performed during or immediately before the application period can be used as evidence that an applicant can de facto work.
Development of a classifier to improve assessments of disability would have to find a balance between false positives and false negatives. From a societal point of view, a false positive might be viewed as a more expensive mistake than a false negative. A possible strategy might be to choose classifiers that err on the side of increasing false negatives rather than false positives, that is, classifiers that have higher specificity at the expense of lower sensitivity. However, the relative emphasis on positive predictive value versus negative predictive value depends on the goals of the classification. In any case, the decision to privilege false negatives over false positives, or vice versa, is a judgment that SSA must make. The committee examined the characteristics of classifiers of listing-level severity and discusses the elements of a study that might predict utilizations that would be a proxy for listing-level severity.
Many factors influence the use of health care (e.g., race and ethnicity, income, and language). Those factors combine to affect the health of individuals and communities, and it is often difficult to disentangle these issues. Numerous studies have demonstrated a strong association between race, low socioeconomic status, and the lack of potential access to care (e.g., no insurance coverage and no regular source of care) and poor receipt of health-care services. Social determinants (e.g., race and ethnicity, language, income, and poverty) have a substantial effect on health-care utilization and health outcomes. Different geographic regions of the country experience different degrees of availability of health care; urban regions have easier access than rural areas. Transportation to and from health-care sites also influence utilizations.
People with disabilities face a number of barriers to access to health care that are specific to their individual limitations in function, including physical access—lack of working elevators or ramps, automatic doors, hallways and doors wide enough to accommodate wheelchairs, and accessible parking—and lack of accommodation for barriers to communication, such as staff willing to try to communicate with impaired patients during scheduling or other interactions.
There have been many changes in the health-care system, for example, movement away from hospitalizations, movement toward outpatient settings or ambulatory care centers, and discouragement of rehospitalizations; thus, utilization might be a poor marker of disease severity and disability. Furthermore, patterns of utilization can be dramatically altered by changes in insurance coverage or by legislative, executive, or state actions. The ACA has expanded coverage, especially of low-income people and more so in states that have expanded Medicaid, and there has been a substantial increase in cost-sharing in the private insurance market (including the growth of high-deductible health plans not related to the ACA) that might decrease utilization.
The committee was unable to find an association between health-care utilizations (such as inpatient hospitalizations, ED use, and hospital readmissions) and disease severity as it relates to SSA’s definition of listing-level impairment severity (i.e., what represents an impairment or combination of impairments severe enough to prevent a person from performing any gainful activity regardless of age, education, or work experience). In some cases, health-care utilizations and repeat utilizations indicate disease severity, but are not necessarily under SSA’s definition of impairment severity.
In other cases, utilizations might indicate poor health care, poverty, lack of insurance, and so on, rather than impairment severity. The scientific and medical literature reviewed did, in some cases, link utilizations (and repeat utilizations) with disease severity, but it did not (and therefore the committee could not) make the leap of connecting health-care utilizations and disease severity with SSA’s definition of impairment severity.
Finally, given appropriate data—such as administrative records of past SSDI applications merged with data on health-care utilizations available at the time of determination and data on postdetermination work outcomes—there are models for quantifying the value of health-care utilizations for determining impairment severity.