Special Population Groups and Measures

** Special Population Groups and Measures

Other studies with disability components focus on particular age groups or
health conditions, and they often use definitions and measures unique to the
specific survey.  For example, NCHS conducts periodic special studies on
current health topics, such as severe and persistent mental illness (1989)
(Table 9), child health, including development, learning, emotional or
behavioral problems (1988) (Tables 7), and aging (1984) using individual
ADL/IADL measures (Figure 7).  Respondents in these NHIS supplements also
answer the activity limitation and other questions on the core survey
instrument, and for this reason multiple measures of disability are possible.
The 1984 SIPP also collected data using separate questions on child physical,
mental and emotional health (Table 8), and the International Center for the
Disabled, in conjunction with the National Council on the Handicapped,
sponsored a Louis Harris poll on the self-perception of persons with
disabilities (Tables 11 and 12).

** Chronic Conditions Causing Disability

Respondents reporting an activity limitation on the NHIS must also identify
the particular chronic condition, e.g., high blood pressure, which causes
this limitation.  Respondents may report more than one chronic condition, but
then they specify which one is the main cause.  Measures of disability,
therefore, can include not only the number of persons with a limitation but
also the associated numbers and types of chronic conditions causing that
limitation (see Table 13 for a detailed listing of chronic conditions).

NCHS defines a condition as chronic if (1) the respondent indicates it was
first noticed more than 3 months before the reference date of the interview,
or (2) it is a type of condition, such as heart disease or diabetes, that
NCHS considers chronic regardless of onset.

NCHS divides its NHIS sample of non-institutionalized persons into six
groups, each of which is randomly administered one of six checklists of
chronic conditions, regardless of disability, as follows:

   1. selected Skin and Musculoskeletal Conditions, such as skin
      cancer or arthritis;

   2. impairments, such as blindness;

   3. selective Digestive Conditions, such as an ulcer;

   4. selected Circulatory Conditions, such as heart disease;

   5. selected Respiratory Conditions, such as emphysema; and

   6. selective Miscellaneous Conditions (of the Genitourinary,
      Nervous, Endocrine, Metabolic, and Blood and Blood-Forming Systems), 
      such as diabetes.

NCHS links this condition information with the survey responses on chronic
activity limitation, restricted activity days and other measures of
disability for analytical purposes, and the Digest uses this linked checklist
information for presenting the disability rates, by chronic condition, in
Tables N, 13, 14, and 15.  These disability rates for each condition result
from dividing the number of disabling conditions reported by the total
reported condition count.  As previously mentioned, not all reported chronic
conditions cause an activity limitation, and many of the most prevalent ones
produce low rates in this regard.  Therefore, the existence of a chronic
health condition is not, by itself, an indication of disability.  Analysis of
multiple conditions within, but not across, the six checklists is possible
because an individual is administered only one condition list.  These
checklists do not include mental disorders which, therefore, do not appear in
these tables.

Chronic condition data also come from the responses to NHIS questions on the
causes of the major and outside activity limitation, irrespective of whether
these conditions pertained to the respondent's particular checklist.  In this
case, analysis across the six domains is possible and the data include
reported conditions not on the checklists, such as mental disorders, as Table
M shows.

** Disability, Health Status, and Health Care Utilization

Chronic activity limitation, as defined by NCHS, manifests itself in many
ways, and the Digest includes data which shows the relationship between
long-term disability and a variety of health indicators, including
respondent-assessed health status, restricted activity days, physician
visits, and short-stay hospital days.

The definition and measurement of health status come from respondents
describing health in general as 1) excellent, 2) very good, 3) good, 4) fair,
or 5) poor.  For presentation purposes, the tables combine data into two
groups, 1) good to excellent and 2) fair or poor.

Another term relating to disability and health is restricted activity days.
Except for persons who are bed-ridden, restricted activity days, as presented
in the Digest, measure the recurring short-term effects of long-term
disability (in some cases with accompanying acute conditions).  The
definition of restricted activity days is the average annual number of days a
person experienced at least one of the following:

   1) a bed day, during which a person stayed in bed more than half a
      day because of illness or injury;

   2) a work-loss day, on which a currently employed person 18 years
      of age or over missed more than half a day from a job or business;

   3) a school loss day, on which a student 5-17 years of age missed
      more than half a day from the school in which he or she was currently 
      enrolled; or

   4) a cut-down day, on which a person cuts down for more than half a
      day on things he usually does.

Depending on the particular table, the measures of restricted activity days
are either condition days (Tables 14 and 15), that is the number of days
during which a particular chronic condition caused a person to restrict his
or her activity, or person days (Table 17), that is the number of days a
person experienced at least one of the four types of activity restriction,
regardless of the number of separate conditions involved.  A person may have
more than one chronic condition, and for this reason the number of condition
days exceeds the number of person days.

Until now, the Digest has defined disability as a chronic (long-term)
condition, such as arthritis, causing an activity limitation (long-term
reduction in activity).  Chronic conditions also can cause restricted
activity (short-term reduction in activity) as just defined.  However, unlike
activity limitation, restricted activity may be caused by either an acute or
chronic condition, or both.  Nonetheless, researchers have found that the
effects of these acute conditions do not alter the strong relationship
between chronic disability and health status/health care utilization.

The tables present these data according to the major demographic
characteristics associated with disability to show how these patterns change
among the various population groups.  These health-related tables also
include AIDS information which is particularly significant given the
increasing prevalence, range of disabilities, and need for long- term care
involved.

** Work Disability

Another domain of analysis is work disability, which the Census Bureau
defines, primarily, as a health problem or disability which prevents persons
from working or which limits the kind or amount of work they can do.  From
its 1988 Current Population Survey (CPS), the Bureau estimates that 8.6
percent of the non-institutionalized population between 16 and 64 years of
age, or 13.4 million persons, had a work disability, including 4.8 percent
with a severe one defined as inability to work at all or on a regular basis.
Another 4.5 million elderly persons 65 to 74 years of age reported having a
work disability.  Work disability prevalence figures include persons in and
out of the labor force as well as those who are employed and unemployed.

Using similar definitions, the 1984 SIPP and the annual NHIS, among many
other sources, also provide data on work disability, and the corresponding
prevalence figures vary among the surveys.  However, each survey is quite
consistent regarding the major correlates of work disability such as labor
force participation and earnings, as well as age, ethnicity, gender,
education, income, and other factors.  Tables Q - V and 22 - 25 provide a
detailed treatment of this subject.

** Disability in Long-Term Care Facilities

Data on residents of institutions are extremely important because the
National Health Interview Survey, Survey of Income and Program Participation,
and the Current Population Survey, among others, are household studies and,
therefore, exclude persons with the most severe disabilities.  The Digest
includes data on three institutionalized population groups: nursing home
residents, mental health inpatients, and persons living in residential
facilities for the mentally retarded.

The major disability-related surveys which the Digest uses for this
information are the 1985 National Nursing Home Survey (Tables 26-29); the
1986 Client/Patient Survey of Inpatient, Outpatient and Partial Care Programs
conducted by the National Institute of Mental Health (Tables 30-32); and the
1987 National Medical Expenditure Survey Institutional Population Component
covering facilities for the mentally retarded (Table 33).  The 1986 Inventory
of Long-Term Care Places also collected data on a range of institutional
settings and served as a basis for sampling for the nursing home and mental
retardation facility studies.

While one can assume that all institutional residents have a disability by
virtue of their placement, for two of these groups, those living in nursing
homes and mental retardation facilities, the studies include ADL measures
and, for nursing home residents, IADL measures as well.

Among the 1.5 million persons in nursing homes, 88.4 percent of whom were 65
years of age and over, more than 90 percent were dependent in at least one
ADL (including continence as a separate activity, but excluding mobility
measures) and virtually all had an IADL dependency (Table 26).  Of the nearly
212,000 residents of facilities for the mentally retarded, 63.8 percent had
at least one ADL difficulty, including walking (Table 33).

When classifying the inpatient population of mental health facilities, NIMH
distinguishes between the long-term caseload, or patients under care,
numbering 160,862 as of April 1986, and the total admissions during the
course of that year, numbering 1.6 million.  The reason for these separate
groupings is that the overwhelming majority of inpatients had relatively
short stays, with a median of only 15 days.  This contrasts with the average
daily census at mental health inpatient facilities, 228,530, which includes a
combination of long and short term residents.  The patients under care
constitute a distinct component within the overall population, and for this
reason the tables present data separately for these two groups (Tables
30-32).

** Federal Benefit Programs for Persons with Disabilities

The Digest includes data from the major federal benefits programs for persons
with disabilities to complement the overall prevalence figures for the
population as a whole.  These programs include special education, vocational
rehabilitation, veterans affairs, Social Security Disability Insurance,
(SSDI), and Supplemental Security Income (SSI) programs for persons who are
blind or disabled.  Information on these programs comes from administrative
data bases on enrollees.  For this reason, the size and composition of the
participant population are strongly influenced by regulatory factors such as
eligibility criteria unrelated to ones functional capacity.  However, in
conjunction with the national survey figures, data on demographic
characteristics and chronic health conditions among beneficiaries provide
another profile of persons with disabilities in this country (Tables 34-45).

** State Level Data

Most data on disability show prevalence estimates at only the national level.
In some cases, this information appears according to the Bureau of the Census
Regions (Table 2).  The reason is that, with few exceptions, the sample sizes
from the various disability-related surveys are too small to allow
state-level estimates.  However, for several sources, disability data for
each state is available either because a full population base was used, e.g.,
all residential facilities for the mentally retarded, or the sample size was
sufficiently large to allow small geographic area breakdowns, e.g., the U.S.
Census.  To help overcome these data limitations, NIDRR is sponsoring a
computer modeling effort to produce state-level estimates from national data
and may be contacted for availability information.

** Interpretation and Limitations of the Data

As introduced, above, federal statistical agencies measure and report on the
prevalence of disability using many different definitions, including 1)
activity limitation in the annual NHIS, 2) functional limitation in the 1984
SIPP, 3) need for personal assistance in performing activities of daily
living (ADL) and instrumental activities of daily living (IADL) in the 1986
SIPP, among other studies, and 4) work disability in the 1988 CPS and the
additional surveys on this topic.  Many national studies which use these
measures also focus on particular health conditions, and they distinguish
between the household population and those living in institutions and among
demographic groups such as children and the elderly.

Activity limitation and functional limitation are the most general measures,
covering a wide spectrum of disability faced by the household population.  On
the other hand, measures of ADL/IADL limitation pertain to persons with
severe long-term personal assistance needs and describe disability in both
household and institutional settings.  Work disability focuses specifically
on the capacity to pursue gainful employment, with sustained earnings.

While the prevalence estimates legitimately vary depending on the particular
survey and aspect of disability, the major demographic correlates are quite
consistent across these data sources.  According to the Bureau of the Census,
those factors having the strongest association with disability are 1) age, 2)
education, 3) income, 4) race and ethnicity, 5) living arrangements, and 6)
gender.  It is important to note that the issue of causality between these
demographic factors and disability often cannot be determined at all from the
data.  For example, while disability can certainly cause poverty, the reverse
may be true as well; however, the extent to which each of these scenarios
occurs is unknown.

Using just some of the data from the Digest's detailed tables, the following
discussion provides a summary of prevalence information and shows the
relationship between disability and various characteristics associated with
it.  This information appears under the following categories, corresponding
to the major disability definitions and measures introduced, above: 1)
General Measure of Disability using the NHIS and 1984 SIPP, 2) ADL/ IADL
limitations using the 1986 SIPP, 3) Special Population Groups, Conditions,
and measures (Children, Elderly, Mental Health, Mental Retardation, and
Self-Perception) using various studies, 4) Chronic Health Conditions Causing
Disability from the NHIS, 5) Health Status and Health Care Utilization using
the 1984 SIPP, 6) Work Disability using the 1988 CPS, 7) Residents of
Long-term Care Facilities using multiple surveys, and 8) Federal Programs for
Persons with Disabilities using administrative data bases. The tables section
of the Digest provides detailed treatment of these topics using many sources
of data and accompanying text, and for this reason only a few examples appear
in this introduction.

To help understand the data, several explanations and interpretive guidelines
are germane.  First, the data presented, below, are normally prevalence
estimates, showing the current number of persons with a disability or number
of conditions existing at a particular reference point or period in time,
regardless of onset.  This contrasts with incidence figures which are the
number of new persons or conditions occurring during or since a specified
time period.

The tables and figures throughout this report present prevalence information
in various ways.  The most frequent measure is the percent of persons in a
particular demographic group, such as high school graduates, who have a
disability, compared to another demographic group, such as those without a
high school diploma.  In this case, the numbers of graduates or
non-graduates, regardless of disability, are the base figures, and the
percent of each with a disability is the measure (see Table 3 and Figure 6).
However, one also can identify the differences between those with and without
a disability relative to particular demographic characteristics, such as
living alone.  In this case, the base figures are the numbers of persons with
or without a disability, and the measure is the percent among each who live
alone (see Table 11 and Figure 7).

Depending on the particular table's structure, either of these rates may
pertain to a column (see Table 5) or row (see Table A) of numbers.  Tables
also may have individual percentages grouped under various subtotals (see
Table 25).  For some of the figures, such as those for different age groups
or those with or without a disability, the counts are mutually exclusive and,
therefore, may be added to a total, usually 100 percent.  Other numbers, such
as persons who have difficulty performing a particular activity, such as
eating, also may be included in the count for another activity, such as
dressing.  In this case the numbers are not mutually exclusive and normally
may not be summed.  To provide aggregate measures for these data, some of
tables present the numbers of persons limited in various combinations of
specific activities (see Table J).

Most tables are in the form produced by the originating federal statistical
agency, and for this reason variations do occur in the location and use of
these base numbers and the meaning of component figures.  The Introduction
and the Highlights and Explanatory Notes which accompany each table provide
interpretive guidance in this regard.

For the household population, the prevalence estimates represent a person's
(or proxy respondent's) own perception of a disability, as opposed to a
professional diagnosis.  However, studies show that respondent-reported
limitations represents an accurate assessment of a person's functional
capacity.

Most of the data in the Digest come from surveys of population samples and
are subject to sampling errors.  Overall, these errors are very small because
the sample sizes are quite large.  In general, when the sample size is too
small for reliable estimates, the figures are either omitted or marked with
an asterisk (*).  However, small differences between two otherwise reliable
numbers or percentages, especially for subgroups in a table, may be due to
chance, e.g., not statistically significant.  Differences highlighted in the
narrative portions of the Digest come from federal statistical agency sources
and are significant.  Other small differences among numbers and percentages
in the tables, or which a reader may recompute from the data, may not be
statistically significant.  The variety of data sources in the Digest
precludes providing detailed reliability documentation on each one; however,
for those interested in small differences or subtle patterns, the source
publications in the citations have complete information in this regard.

Finally, as with any data, statistical relationships for groups of persons,
such as the correlation between low educational attainment and work
disability, cannot be ascribed to one person or a few individuals, who may
not follow the group's pattern.