Understanding the importance of rehabilitation science and engineering and the potential impact that it might have on improving the health of the nation first requires an understanding of the current status of the incidence, prevalence, costs, and potential savings associated with rehabilitation. This chapter describes the various types of disabling conditions and their frequencies of occurrence in the United States as measured by various surveys and other means. It also attempts to characterize the associated costs and savings that can be realized through effective rehabilitation.
The most recent estimates of the number of people with disabilities is 49 million noninstitutionalized Americans (McNeil, 1993). Almost 4 percent of the U.S. population have disabling conditions so severe that they are unable to carry out the major activities of their age group (playing, attending school, working, or attending to self-care) (Institute of Medicine, 1991). An additional 6 percent are restricted in their major activities, and another 4 percent are limited in other types of activities.
In addition to and partly as a result of the loss of human function, enormous economic costs are associated with disabling conditions. Estimates vary but seem to hover around an aggregate annual cost of approximately $300 billion, including the cost of the medical resources used for care, treatment, and rehabilitation; reduced or lost productivity; and premature death.
As described in several reports, including Disability in America (IOM, 1991), numerous federal programs exist for people with disabling conditions. Most recently, a report by the National Academy for State Health Policy identified 129 separate programs administered by 14 different federal agencies, with annual funding of $175 billion. Approximately 95 percent of this money is allocated for income support and medical coverage. The remainder is divided among research and a variety of service-related activities, especially in the areas of education, housing, and transportation.
The federal government's largest program in rehabilitation research is located in the National Institute on Disability and Rehabilitation Research (NIDRR) in the U.S. Department of Education. As mandated by the U.S. Congress, NIDRR also has primary responsibility for coordinating rehabilitation research among federal agencies. The NIDRR director is the chair of the Interagency Committee on Disability Research (ICDR), which is charged with promoting communication and joint research activities among the committee's member agencies.
Other agencies involved in conducting rehabilitation research include the National Institutes of Health (NIH) and the U.S. Department of Veterans Affairs (VA). In 1984, NIH described 688 rehabilitation-related research projects in addition to other basic studies that help to elucidate the biological underpinnings of impairment and disability. In 1990, a new center, the National Center for Medical Rehabilitation Research (NCMRR), was established at NIH to help coordinate and focus specifically on medical rehabilitation research. VA supports a rehabilitation-related research program that allocates approximately $22 million to fund more than 175 separate projects at 60 VA medical centers.
Major National Surveys
The main source of statistics on people with disabling conditions are the federal surveys based on nationally representative samples of the noninstitutionalized U.S. population.
National Health Interview Survey
The National Health Interview Survey (NHIS) is a household survey sponsored by the National Center for Health Statistics (NCHS) and is designed to assess the health status of Americans. In 1994, the survey consisted of interviews with 116,179 people in 45,705 households. It includes questions related to disability such as degree of activity limitation and provides information by demographic variables such as age, race, and gender.
Activity Limitations
In NHIS terminology, disability is defined as activity limitation. Activity limitation is defined at three levels: (1) inability to carry out a major activity, (2) limitation in the amount or kind of major activity that can be carried out, and (3) limitation in carrying out a nonmajor activity. Major activities considered usual for one's age group are defined as ordinary play for children under 5 years of age, attending school for children ages 5 to 17, working or keeping house for people ages 18 to 69, and capacity for independent living (ability to bathe, shop, eat, and care for oneself without the assistance of another person) for people ages 70 and older. Nonmajor activities include social, civic, or recreational pursuits. The 1994 NHIS estimate of the number of people limited in activity because of chronic conditions was 39 million, or 15 percent of the civilian noninstitutionalized population. Of these 39 million people, 18.2 million were male and 20.8 million were female; 32.4 million were white and 5.4 million were African American. Residents of the South (16.3 percent) and rural areas (17.6 percent) had a slightly higher prevalence of disability than did residents of other locations.
Table 2-1 presents disability rates by demographic characteristic for the 1992 NHIS. Table 2-2 indicates the prevalence of activity limitations, limitations in the self-reported ability to work among people 18 to 69 years, and limitation in activities of daily living (ADL) and instrumental activities of daily living (IADL) among people over age 5 years by the impairments and diseases or disorders causing the limitation. The data summarize information from LaPlante and Carlson (1996) and are derived from analyses of the 1992 National Health Interview Survey (NHIS) conducted by the Bureau of the Census for the National Center for Health Statistics. NHIS surveyed a stratified random sample of the noninstitutionalized population of the continental United States and had approximately 110,000 respondents. Sampling weights associated with each respondent allowed for the estimation of the total number of people in the continental United States with limitations associated with impairments and conditions.
TABLE 2-1
Crude and Age-Adjusted Rates of Limitation in Activity, by Selected Sociodemographic Characteristics, 1992.
TABLE 2-2
Prevalence of Activity Limitations, Work Limitation, and Need for Assistance in ADLs and IADL, by Impairments and Diseases and Diagnoses Causing the Limitation, Continental United States, 1992.
Total Prevalence Table 2-2 presents data on the prevalence of activity limitations associated with major classifications of impairments and diseases or disorders, including the number of people with the particular classification and limitation and the proportion of all activity limitations attributed to the classification. Overall, in excess of 61 million impairments or diseases and disorders contributed to activity limitations in 1992; of these, 16.3 million were impairments (26.7 percent of the total) and the remaining 44.7 million were diseases or disorders (73.3 percent). Among impairments, orthopedic impairments were the most common classification contributing to limitations, with a total of 8.6 million conditions accounting for 14.1 percent of all conditions that contribute to limitations. In excess of 1 million cases each of visual or hearing impairment, learning disability or mental retardation, and paralysis contributed to limitations, although none of these classifications individually accounted for more than 2.6 percent of all conditions contributing to an impairment.
The most common major classifications of disease and disorder contributing to activity limitations included musculoskeletal and connective tissue disorders (accounting for 17.2 percent of all conditions contributing to limitations), circulatory conditions (16.7 percent), respiratory conditions (7.8 percent), and nervous system and sensory organ conditions (7.2 percent).
Disability with a Primary Cause Table 2-2 also presents estimates of the major classifications of impairments and diseases or disorders reported by NHIS respondents as the main cause of their limitations; the number of people with each major classification as the main cause of their limitation and the proportion of all causes of limitation for which the particular classification is the main cause are presented. A total of 37.7 million people reported activity limitations in 1992. Of these, 10.9 million (roughly 2/3 of all 16.3 million individuals with an impairment) stated that any form of impairment was the main cause of their limitation. The probability that a condition will be reported as the main cause of limitation differs dramatically among impairments. Thus, only 43.1 percent of visual impairments were said to be the main cause of limitation, whereas 88.2 percent of the cases of learning disability or mental retardation were reported to be the main cause of limitation. In terms of prevalence, orthopedic impairments were the most common main cause of limitation.
More than 26.8 million people, or just under 60 percent of all 44.7 million people with diseases or disorders contributing to limitation, stated that a disease or disorder was the main cause of their limitation. Digestive, endocrine, nutritional, metabolic, immunologic, blood and blood-forming organ, and skin and subcutaneous conditions are least likely to be reported as the main cause of activity limitation, whereas respiratory, musculoskeletal, and connective tissue conditions, mental conditions, congenital anomalies, neoplasms, and infectious diseases are most likely to be reported as the main cause of activity limitation. In terms of prevalence, musculoskeletal and connective tissue and circulatory conditions are the most common diseases and disorders listed as the main cause of limitation.
Impairment and Work Limitation NHIS asked people 18 to 69 years of age questions about work limitations. Table 2-2 indicates the frequency of conditions contributing to work limitations and the proportion of all work limitations associated with each major classification of impairment or disease and disorder. In 1992, more than 31.3 million people reported having a condition that contributed to a work limitation. Of these, in excess of 8.5 million (27.3 percent of all people with conditions contributing to work limitations) had impairments that contributed to work limitations. Orthopedic impairments were again the most common form of impairment contributing to work limitations; more than 5.2 million cases of orthopedic impairment were cited, which represents 16.8 percent of all conditions affecting work capacity. Other impairments that were common causes of work limitation included visual impairments (mentioned 580,000 times), paralysis (552,000 times), and learning disabilities or mental retardation (546,000 times).
In excess of 22.7 million cases of disease or disorder were reported to contribute to work limitations, representing 72.5 percent of all conditions contributing to such limitations. The most common diseases and disorders contributing to work limitations include musculoskeletal and connective tissue disorders (5.4 million individuals), circulatory diseases (5.1 million individuals), and nervous system and sensory organ conditions (2.1 million individuals).
Impairment and Daily Life Finally, Table 2-2 indicates the number of conditions contributing to limitations in ADL or IADL and the proportion of all such limitations associated with each major classification of impairment or disease or disorder. The data concerning ADL or IADL limitations are limited to persons age 5 years or older. In 1992, in excess of 9.2 million individuals indicated that they had conditions that contributed to ADL or IADL limitations. Of these, 2.3 million (about 25 percent of all people with conditions contributing to ADL and IADL limitations) had impairments that contributed to ADL and IADL limitations. A total of 988,000 individuals had orthopedic impairments that contributed to ADL or IADL limitations; other impairments mentioned included learning disabilities or mental retardation (399,000 individuals), paralysis (278,000 individuals), and visual impairment (254,000 individuals). All forms of diseases or disorders contributed to ADL or IADL limitations in more than 22.7 million individuals, or roughly 3/4 of the total. The most common major disease classifications contributing to this form of limitation included musculoskeletal and connective tissue disorders (1.9 million individuals), circulatory conditions (1.6 million individuals), and nervous system and sensory organ conditions (884,000 individuals).
Prevalence of Activity Limitation in Children Among children under the age of 18 years, an estimated 4.0 million (6.1 percent of the U.S. population under the age of 18 years) have some type of disabling condition.* Disability in this age group is defined differently from disability in adults and includes any limitation in activity due to a chronic health condition or impairment. Work limitations for adults are translated to limitations in play (under age 5 years) and school-related (ages 6 to 17 years) activities, because these are major activities for children. Play is one of the most important ways that children learn about the world. If play and activity are absent during early life, an important part of the foundation on which the child's life is based will be missing. Engineering and technology, in association with rehabilitation science, can provide substitute play and activities to compensate for the typical play and activities that may be missed. Like those for adults, the findings presented in this section were derived from analysis of data from the 1992 NHIS. Data were collected from households of the noninstitutionalized U.S. population by asking questions of parents and guardians. Children were not interviewed or observed.
Play and School ActivitiesThe prevalence of children with disabling conditions is greatest in those attending school and represents 7.4 to 7.6 percent of all children ages 5 to 17 years (Wenger et al., 1996) (Table 2-3). The majority of these children with disabling conditions are unable to perform a major activity or are limited in the amount or kind of major activity that they can perform. More males than females are represented among children ages 5 to 17 years with disabilities (Wenger et al., 1996) (Table 2-4).
TABLE 2-3
Number and Percentage of Children Under Age 18 with Disabilities, by Degree of Limitation and Age, 1992.
TABLE 2-4
Number and Percentage of Children Aged 5 to 17 with School-Related Disabilities, by Degree of Limitation and Gender, 1992.
Distinct Childhood PatternThe data on impairments and diseases associated with all children with disabling conditions reveals a pattern distinct from that for adults. The major impairment associated with disabilities is mental retardation or Down's syndrome, occurring in 15.8 percent of all children with disabilities (Wenger et al., 1996) (Table 2-5). This is followed by speech impairments (6.7 percent), hearing impairments (3.8 percent), and learning disabilities (2.8 percent), whereas orthopedic impairments (2.9 percent) and deformities (e.g., spina bifida) (2.7 percent) are not as prevalent in children as they are in adults.
TABLE 2-5
Health Conditions and Impairments Causing Disability in Children Under 18, by Broad Condition Category, 1992.
Similarly, the major disease or disorder (i.e., pathology) associated with children with disabling conditions differs from that for adults because diseases of the respiratory system (23.6 percent) supplant cardiovascular disease in adults. Asthma is the leading respiratory disease associated with disability in children (19.8 percent); mental disorders (8.8 percent) and diseases of the nervous system and sense organs (7.5 percent) are the second and third most prevalent causes, respectively, of disabling conditions in children. Diseases of the musculoskeletal system and connective tissue are not as dominant in children with disabilities as in adults with activity or work limitations (Table 2-5).
In summary, the pathologies and impairments associated with child hood functional limitations and disabilities appear to be distinct from those for adults. This suggests that a research emphasis on pathology and impairment in rehabilitation science and engineering should reflect the fact that disabling conditions among children are distinct from those among adults. Also, the prevention or reversal of the most prevalent causes of activity limitations among adults might begin in childhood.
Functional Limitations
The questions in NHIS also address the need for personal assistance in activities of daily living, such as eating, bathing, dressing, and getting around the home, and instrumental activities of daily living, which are everyday household chores, necessary business, shopping, or getting around for other purposes.
Chronic Conditions Causing Disability
In the NHIS respondents identify chronic conditions that cause activity limitations. A condition is considered chronic if either (1) it was first noticed 3 months or more before the reference date of the interview or (2) it is a type of condition generally considered chronic by NCHS, regardless of the time of onset, such as diabetes. Most chronic conditions do not have high risks of disability. About 12 percent of conditions identified in NHIS cause activity limitations, the broadest measure of disability. Impairments have the highest risk of becoming a disabling condition. Of the conditions reported in NHIS to cause activity limitations, heart disease ranks first, followed by back disorders, arthritis, orthopedic impairments of the lower extremity, and asthma (LaPlante and Carlson, 1996) (Table 2-6).
TABLE 2-6
Conditions with Highest Prevalence, All Causes of Limitations, 1992.
Families and Disability
Prevalence estimates of disability have focused on the individual as the unit of analysis. A new study looks at disability prevalence with the family as the unit of analysis (LaPlante and Carlson, 1995). It examines the composition of families with members with disabling conditions in comparison with the composition of families without a member with a disabling condition, their demographic and socioeconomic characteristics, and their utilization of health services. The study found that an estimated 20.3 million families, or 29.2 percent of all 69.6 million U.S. families, have at least one member with a disabling condition. An estimated 2.3 million (4 percent) two-parent families contain one or more children with a disabling condition. The rate of disability is 29.1 percent for white families, 31.9 percent for African-American families, and 21.7 percent for other ethnicities. Among Hispanic families, 23.4 percent have members with disabling conditions. In general, the median family income is substantially lower if a head of household has a disabling condition, whereas income is affected much less by the presence of other members of the family with disabling conditions. The highest poverty rates by disability status are among families with single heads of households with two or more children with disabling conditions. More than half of such families are headed by women living at or below the poverty level.
NHIS Disability Supplement
NCHS fielded a disability supplement to NHIS that began in January 1994 and that continued through 1996. This survey represents a consensus reached by researchers and policy makers and will provide comprehensive information for estimates of prevalence and for program and policy development. In the first phase of the survey, conducted during 1994 and 1995, basic information on disability was obtained by personal interviews for a national representative sample of 225,000 people, about 45,000 of whom had some indication of a disabling condition. In the second phase of the survey, which began in late 1994 and which continued through 1996, the 45,000 people with an indication of a disabling condition were reinterviewed to obtain additional information.
A separate questionnaire was administered to children, including a control group of children without special health needs. In the first phase, questions about developmental milestones for children under age 5 and about performance of activities of daily living for children ages 6 to 17 were asked. The second phase included questions on (1) utilization and barriers to utilization of medical and mental health services, assistive devices, case managers, home care services, child care services, and educational and recreational services; (2) functional status, including measures of emotional and behavioral development; and (3) impact of the child's health problem on the family. The data were collected over a 2 year period to ensure an adequate sample size. The following were among the topics covered:
Physical health conditions
Childhood development
Mental health conditions
Functional assessment
Assistive technology devices
Income sources and amounts
Family impact of disability
Personal assistance services
Health insurance coverage
Self-perception of disability
Special education services
Supplemental Security Income and Social Security Disability Insurance participation
Transportation accommodations
Work site accommodations
Use of medical services
Vocational rehabilitation
Data will be released on electronic files for statistical analysis. The first release was in mid-1996 and included data collected for Phase I in 1994. NCHS plans to publish several descriptive reports based on the disability data, but NCHS does not have staff or funds to support fullscale analysis of the data. Verbrugge (1994) has outlined the research potential of this new source of data. Efforts are under way to identify sources of funding for data analysis.
Survey of Income and Program Participation
The Survey of Income and Program Participation (SIPP) is a panel survey designed to provide detailed information about income distribution and federal and state income transfer and services programs. A supplemental survey containing extensive questions about disability status was performed as part of the sixth wave of the 1990 panel and the third wave of the 1991 panel. SIPP contains information on economic and social variables on people with disabling conditions that are not usually included in health surveys that ask about disability. McNeil (1993) has provided disability data from SIPP.
In SIPP, functional limitation is defined as the ability of people ages 15 years and older to perform a set of sensory and physical activities. Limitations are ranked as 1 (with difficulty) or 2 (not at all or only with aid).
SIPP also uses need for assistance in activities of daily living and instrumental activities of daily living as a measure of disability. Mobility limitations are reported separately, because assistive devices such as wheelchairs and canes, rather than another person, are often used to overcome such limitations.
On the basis of interviews conducted during the period from October 1991 through January 1992, SIPP found that the number of people with a disability was 48.9 million, or 19.4 percent of the total population at the time of 251.8 million. Disability was defined as a limitation in a functional activity or in a socially defined role or task.
SIPP identified the number of people with a severe disability to be 24.1 million, or 9.6 percent of the population. The 24.1 million people identified as having a severe disability were identified as people who were unable to perform one or more activities, people who had one or more specific impairments, or people who used a wheelchair or who were long-term users of crutches, a cane, or a walker.
Current Population Survey
The Current Population Survey (CPS), a monthly survey conducted by the Bureau of the Census for the Bureau of Labor Statistics, is designed to collect information on labor force participation and income. In March, supplementary questions are asked about income and work disability, defined as a limitation in the kind of work that a person is able to perform because of a chronic condition or impairment.
Work Disability
Data from the 1995 CPS indicate that among people ages 16 to 64, 16.9 million people (10 percent) had a work disability (LaPlante and Carlson, 1995).
In addition, 11.4 million people (67.9 percent) among those with a work disability were not working and were not actively seeking employment.
The percentage of the population with a work disability increased with age. The group ages 16 to 24 had the lowest proportion (4 percent). This increased to 22 percent for those ages 55 to 64. The percentage of the population with a work disability decreased with the level of educational attainment, measured in years of school completed. People with fewer than 8 years of schooling had a work disability rate of 30 percent, compared with a rate of 4 percent for those with at least 16 years of education. This education-based disparity increased for people with a severe work disability. Those who had completed less than 8 years of school had a severe work disability rate of 23 percent, whereas the rate was 1 percent for those with at least 16 years of formal education. This means that the severe work disability rate among those with little schooling is greater than the among college graduates.
African Americans have a much higher rate of work disability (14 percent) than either whites (8 percent) or people of Hispanic origin (8 percent) (U.S. Bureau of the Census, 1989).
Back disorders rank as the most frequent cause of work disability (16.4 percent), followed by heart disease (13.1 percent) and arthritis (8.1 percent) (LaPlante and Carlson, 1996).
Labor Force Participation
According to 1995 CPS data, of those with a disabling condition, 27.8 percent have jobs, whereas 76.3 percent of people without disabling conditions have jobs (LaPlante et al., 1995).
Income
SIPP data indicate a negative association between earnings and disability status. For example, among people ages 35 to 54, those with no disabling condition had mean monthly earnings of $2,446, those with a disabling condition that was not severe had monthly earnings of $2,006 and those with a severe disabling condition had monthly earnings of $1,562. However, there was a strong negative association between education and disability status. Therefore, one of the ways that a disabling condition may affect earnings is through its effect on levels of education and training (McNeil, 1993).
Costs of Disability and Rehabilitation
Cope and O'Lear (1993) reported that research firmly establishes the clinical benefit and economic savings associated with early, aggressive, and highly expert application of rehabilitation technology to both brain injury and spinal injury patients. Just as estimates of disability prevalence vary depending on the definition of disability, however, so do estimates of the costs of disability. Estimates of cost of disability are done by surveying populations of interest and by secondary analysis of large databases. Most of the cost estimates available are for traumatic brain injury or spinal cord injury. Direct costs include medical treatment and rehabilitation. Indirect costs include loss of earnings resulting from the disabling condition. Also included are studies of the cost savings resulting from rehabilitation. Hill (1991) presented a comprehensive study of direct costs (cash transfer programs, medical care expenditures, and costs of direct services) and indirect costs (reduced earnings) and found that direct costs for fiscal year 1986 were $86.5 billion in cash transfers, $79.3 billion in medical care payments, and $3.5 billion in direct services. Indirect costs can range from 10 to 37 percent of preillness income.
Traumatic Brain Injury
The Brain Injury Association estimates that each year more than 2 million people sustain a brain injury, and 373,000 of these are severe enough to require hospitalization. Brain injury ranks as the leading cause of death and disability among children and young adults. An individual with severe brain injury typically faces 5 to 10 years of intensive rehabilitation, with cumulative costs of $48 billion annually (Brain Injury Association, 1995). In addition to the costs of hospitalization and rehabilitation, head injuries result in 14 million person-days of restricted activity each year (Max et al., 1991).
Lehmkuhl et al. (1993) examined data for 301 patients in model traumatic brain injury systems and found that total charges for acute care and inpatient rehabilitation, exclusive of physicians' fees, ranged from an average of $73,000 for mild traumatic brain injury to $154,000 for very severe traumatic brain injury. A population-based study of persons surviving traumatic brain injury (Brooks et al., 1995) found that costs for acute care and rehabilitation ranged from $17,015 for mild injuries to $133,467 for severe injuries. The study also examined costs for people with traumatic brain injuries 4 years after the initial injury and found that follow-up charges ranged from a mean of $2,323 for mild injury to $54,701 for severe injury. Follow-up costs included rehospitalization, visits to physicians, outpatient services, medication, equipment, supplies, attendant care, and other services. At an incidence rate of 69 per 100,000, the investigators projected that the total cost for new patients with traumatic brain injury requiring hospitalization will exceed $8 billion over the course of the first 4 years following injury.
Spinal Cord Injury
Harvey et al. (1992) used data from a survey of the U.S. population with spinal cord injuries to estimate the direct costs of traumatic spinal cord injuries. Direct costs were defined as the value (in 1988 dollars) of resources used specifically to treat or to adapt to the spinal cord injury. Estimates are $95,000 for initial hospitalization, $8,000 for modifications to the person's dwelling, $8,000 per year for medical services, supplies, and adaptive equipment, and $6,000 for personal assistance and institutional care. These are average costs and will vary depending on the severity of the injury, age, and patient motivation.
The National Spinal Cord Statistical Center at the University of Alabama at Birmingham estimates that there are between 7,600 and 10,000 new patients with spinal cord injuries each year. The lifetime costs directly attributable to spinal cord injuries vary greatly according to the severity of the injury. Average yearly health care and living expenses in 1992 dollars varied from $417,000 for the first year and $75,000 for each subsequent year for individuals who sustained severe injuries to $123,000 for the first year and $9,000 for each subsequent year for individuals who sustained less severe injuries. The average costs for all groups was $198,000 for the first year and $24,154 for each subsequent year.
These figures do not include any indirect costs such as losses in wages and productivity, which could average almost $38,000 per year but vary substantially on the basis of education, severity of the injury, and preinjury employment history (National Spinal Cord Statistical Center, March 1996).
Cost-of-Illness Framework
The impact of illness on society is frequently estimated by calculating the amount of medical care expenditures on behalf of people with disabling conditions (called direct costs in the cost-of-illness nomenclature) and the amount of wage losses or its equivalent in services provided by homemakers (called indirect costs) (Rice and Cooper, 1967). In an alternative formulation, some economists seek to price losses in all domains of activity, including work and housework but also encompassing leisure, family, and voluntary activities, by asking individuals how much they would be willing to pay to forego an illness (Thompson et al., 1982). However, the methods used to assess willingness to pay are primitive, and results have not differed substantially from those obtained by the more traditional methods of assessing costs of illness (Thompson, 1984).
Using the cost-of-illness framework, it becomes possible to estimate the economic impact of disability. Trupin and colleagues (1996) used the National Medical Care Expenditures Survey for 1987 to estimate the medical care expenditures of people with and without disabling conditions. They reported that the approximately 17 percent of the population with an activity limitation accounted for 47 percent of total medical care expenditures. These individuals incurred medical care costs four times as great as those for people without disabling conditions, accounting for 38 percent of hospital admissions but 57 percent of total hospital costs and 19 percent of costs for visiting physicians but 42 percent of total physician service expenditures. (See Chapter 7 for an examination of the implications of these figures on health services research.) Overall, people with disabling conditions had $157 billion in medical care expenditures in 1987. Expressed in 1994 terms, medical care expenditures for people with disabling conditions would amount to $205.7 billion, or 3.1 percent of the gross domestic product (U.S. Bureau of the Census, 1995).
Chirikos (1989) estimated both direct and indirect costs of disability using 1980 data. He reported an aggregate economic cost of disability of $176.8 billion, 51 percent due to medical care expenditures and the remainder due to the lost productivity of people with disabling conditions or family members who had to stop working to care for them. Expressed in 1994 terms, medical care expenditures would amount to $163.1 billion and indirect costs would total $155.0 billion.
Using the cost-of-illness framework, the committee estimated the magnitude of the indirect costs of disability with more recent data. First, the committee used the 1994 NHIS to compare the labor force participation rates of people with and without disabling condition of working ages, ages 18 to 64. In that survey, people with disabling conditions are those who report that they are unable to do the major activities for people their age, who report being limited in the amount or kinds of these activities, or who report being limited in nonmajor activities. Of 158.6 million working people in the United States, 22.5 million (14.1 percent) reported that they had disabling conditions. Of these individuals, 51.8 percent were in the labor force, whereas 83.0 percent of people without disabling conditions are in the labor force, a difference of 31.2 percent.
Assuming that people both with and without disabling conditions earned median hourly wages, wage losses for the percentage of people with disabling conditions who could be working but were not amounted to $158.7 billion, or 2.4 percent of the gross domestic product for 1994. The foregoing figure assumes that both groups bring an equal mix of skills and experience to the labor market. However, people with disabling conditions typically have other potential liabilities that affect their position in the labor market and, thus, in recent years earned only about 70 percent as much as people without disabling conditions (Yelin and Katz, 1994). Assuming that the earnings gap reflects differences in skill and experience—that is, objective characteristics—rather than discrimination, the indirect costs due to lost wages would amount to $111.1 billion.
Second, the committee used the 1994 CPS to estimate lost wages among people with disabling conditions. CPS is the source of the monthly unemployment statistics for the United States. Once a year, respondents are asked about their disability status. In the CPS people with disabling conditions are those who self-report the presence of a limitation that prevents work or limits the amount or kind of work that they can do. According to CPS, 11.8 million people have such a disabling condition. CPS also collects data on hours of work among those with disabling conditions who are working, allowing one to estimate wage losses for those who have stopped working altogether and for those who are working fewer hours. Owing to the more stringent definition of work disability in CPS, only 21.1 percent of people with disabling conditions were working in the week before the interview, whereas the proportion was 68.6 percent among people without disabling conditions, a difference of 47.5 percent. In addition, the people with disabling conditions who were working averaged 36.4 hours per week on the job, compared with 41.3 hours among people without disabling conditions, a difference of 4.9 hours. Summing the wage losses of those who stopped working altogether and those with reduced hours of employment, indirect costs due to disability amounted to $133.0 billion, or 2.0 percent of the gross domestic product in 1994. After taking differences in skill and experience into account, wage losses would still amount to $93.1 billion.
The estimates of indirect costs due to wage losses in CPS are lower than the estimates in NHIS because of the lower overall prevalence of individuals with disabling conditions in the former survey. Nevertheless, both estimates are in the same range as those of Chirikos (1989), suggesting that the definition of disability aside, the indirect cost of disability is well in excess of $100 billion annually.
Studies of the cost of illness emphasize wage losses, because such costs are relatively easy to measure. This methodology underestimates the impact among women, however, both because women earn less than men for jobs requiring similar levels of skill and because homemaking activities are poorly remunerated in the labor market, even though families value them highly (Lubeck and Yelin, 1988). Women have higher rates of disability than men, accentuating the problem of estimating the costs of disabling conditions.
However, the impact of disability extends far beyond work. Indeed, people with disabling conditions are less likely to be involved in all domains of human activity than people without disabling conditions (Lubeck and Yelin, 1988). In addition, because they devote so much time to personal care activities and to the time required to secure medical care services, even when they do participate in the same activities as people without disabling conditions, they are able to devote less time to many of them. These ''costs" of disability are not easily priced in the marketplace, but they are important to people with disabling conditions and to their families. Thus, the impact of disability is far larger than the costs accounted for by current economic methods.
Future Needs
The first National Disability Statistics and Policy Forum, organized by the Disability Statistics Rehabilitation Research and Training Center, was held in October 1994. The topic of the forum was the future of disability statistics. Participants at the forum identified the following needs for future data collection efforts: more emphasis on social participation of people with disabling conditions, designing data sets that respond to policy questions, more emphasis on mental illness data, the need for more state-level data, more sharing of data across agencies, and use of repeated cross-sectional data to identify trends.
The second National Disability Statistics and Policy Forum in June 1995 focused on employment statistics and policy. Participants emphasized the need for data on barriers to work; estimates of the number of people with disabling conditions who are working, looking for work, and out of the labor force; and earnings and benefits data by occupation, industry, impairment groups, and environmental barriers, all reported in a consistent time series. They agreed on the need to examine more closely the problems that make it difficult for people with disabilities to go to work. Surveys should be designed to collect data on health status separately from employment status and work limitations. Data on how the workplace accommodates impairment and on the costs of accommodations required under the Americans with Disabilities Act of 1990 need to be collected. These data are needed to respond to concerns that the American with Disabilities Act imposes costly burdens on employers.
The third National Disability Statistics and Policy Forum was held in May 1996 and was titled Housing and Disability: Data Needs, Statistics, and Policy. The conference examined ways in which the quantity and quality of statistical information on the housing situation of Americans with disabling conditions can be improved.
Conclusions and Recommendations
Estimates of the prevalence and economic impact of disability are dependent on the definition of disability used. Differences in methods aside, the prevalence of disability may be as high as 14 percent of the population and in certain age-gender-race groups may be significantly higher. Direct costs of disability would appear to be as high as $200 billion dollars a year and indirect costs may be as high as $155 billion. Thus, regardless of the definition of disability used, disability affects a substantial portion of the population and exacts a tremendous economic toll on the nation. In addition to diminishing or reducing disabilities due to paralysis or to visual or orthopedic impairments, rehabilitation science and engineering can contribute handsomely to enabling people to work by modifying the work environment, providing special equipment at work sites, enabling people to work at off-site locations, or providing the personal aids needed to carry out work tasks.
The foregoing data on the prevalence of and costs associated with disability were collected without an explicit conceptual model of disability. Accordingly, disability is defined in different terms for each of the major statistical series. More importantly, the design of each of these series predates the development of a more contemporary understanding of the process by which pathologies, impairments, and functional limitations give rise to disability, suggesting that the emerging definitions of disability are yet to be reflected in data collected in current surveys. In the chapter to follow, the committee reviews how the prevailing wisdom about the cause of disability has changed in the last several decades and then shows how the emergent model of disability might structure the research agenda for the foreseeable future, including the collection of data on the prevalence and impact of disability.
Recommendation 2.1 The Disability Statistics Subcommittee of the Interagency Committee on Disability should foster research to design and evaluate survey items to be used to ascertain the prevalence and impact of disability that accord with the contemporary model in which disability is jointly determined by characteristics of individuals and of their environments.
Recommendation 2.2 These survey items should be incorporated in on-going surveys, including the National Health Interview Survey, Current Population Survey, and Survey of Income and Program Participation.
- *
The committee believes that, given the potential for effective interventions that can enable people with disabling conditions, most of these conditions should be strictly defined as potentially disabling conditions. For the sake of readability, however, we use the term disabling condition throughout this report with the intent that ''potentially" is understood.