Persons With Mental Retardation and Related Conditions in Mental Retardation Facilities: Selected Findings from the 1987 National Medical Expenditure Survey. Part 3: Alternative Population Estimation

Part I of this report noted a number of significant limitations in the sample frame for the Institutional Population Component of NMES. It was noted that there is strong evidence that this caused substantial underrepresentation of smaller community-based residential facilities and their residents in the National Medical Expenditure Survey. The general direction of underrepresentation of small facility residents is clear: "facilities" with 1 or 2 residents were completely eliminated from the study when it became apparent that the sample frame contained only a small proportion of all such facilities nationwide, and there was also considerable underrepresentation of other small facilities.

Corroboration regarding underestimation of small facilities and their residents comes from state reports on the number of facilities that they have under licensure or contract or that they directly operate, and the number of people with mental retardation and related conditions living in them (Lakin, et al., 1989). In addition, the estimation from NMES that the number of small community facilities and residents in 1987 was essentially unchanged from the 1982 NCRF (Lakin, Hill, & Bruininks, 1985), while state institution populations decreased by 25,000 people (most of whom were released into community-based group settings) seems implausible and also suggests underrepresentation of small facilities in the NMES population estimates. As noted earlier, estimates of persons in large facilities were reasonably comparable to expected values.

Given the problems with the estimates in the Institutional Population Component, consideration of ways to adjust this data source to permit more accurate population estimates seems warranted. In the following pages the simplest available method of compensating for the underrepresentation of small facilities is explored. However, it is important to note at the outset that data to assess the precision of any alternative estimates are not available.

The simplest alternative procedure for using the Institutional Population Component statistics to obtain more realistic estimates of the populations of residential facilities for people with mental retardation, particularly the smaller ones, is to in effect reweight its sample to reflect known populations of different types of facilities. This can be done by using the proportional estimates from the 1987 NMES sample, those reported in Part 2 and applying them to more accurate total population statistics on people in mental retardation facilities by size and type as are known and reported by the individual states. Such statistics, based on June 30, 1987 reports of all state mental retardation/developmental disabilities agencies, are available (Lakin et al., 1989). These statistics generally coincide with the dates of the NMES interviews. These statistics indicate that small (15 or fewer resident) mental retardation facilities did not house 65,000 people in 1987, they housed on the order of 118,500 people. Use of the data obtained in NMES to respond to the practical questions asked about mental retardation facilities and their residents in most instances need to reflect that reality.

Table 37 presents selected characteristics of the NMES sample which have been adjusted to the nationally aggregated reports of individual states regarding the populations of people with mental retardation and related conditions in mental retardation facilities in June 30, 1987. In Table 37, the statistics presented outside of parentheses are the proportions of all residents within facility categories reported to exhibit the selected behaviors/conditions as obtained from the analyses reported in Part 2. In parentheses are population estimates obtained when these proportions were applied to national population statistics reported by the states.

These alternative population estimates are briefly discussed in the following pages. These comments focus primarily on the differences of significance between the estimated populations of small mental retardation facilities, which the Institutional Population Component estimated to be 64,936 people with mental retardation and related conditions in 1987, but which states reported to be 118,570 people on June 30 of that same year. Following this presentation is a discussion of the extent to which evidence exists to support such alternative estimates.

Level of Mental Retardation

According to the NMES population estimates, in 1987 there were an estimated 8,834 people with profound mental retardation in small mental retardation facilities. This represented 13.6% of the population estimated to be in small facilities (64,939). If the 13.6% of all residents were applied to the state reported population of facilities with 15 or fewer residents, an estimated small facility population of 16,126 persons with profound mental retardation would be obtained. Similarly, the NMES estimated 15,258 persons with severe mental retardation in the smaller community based facilities. Application of NMES proportional estimates to the known population of the smaller facilities would yield an estimate of 27,864 people with severe mental retardation in community facilities.

Adjusting the NMES statistics may have importance beyond that of improved accuracy of estimate. Considerable debate continues at the federal and state levels with respect to the continuing need for institutional care. Many questions raised in this debate revolve around whether appropriate services can be provided for people with the most severe handicaps with small community settings. Clearly the extent to which community-based living is already being provided to persons with severe impairments is important evidence of the viability of community living settings for all, or virtually all, persons with mental retardation and related conditions. Regarding this issue an estimate that 16,126 persons with profound mental retardation (or 20% of persons with profound mental retardation in mental retardation facilities) are currently living in community facilities suggests significantly different placement practices and community residential services viability than an estimate of 8,834 (or 12% of persons with profound mental retardation in mental retardation facilities).

Disturbing Behavior

Estimated proportions of residents with mental retardation and related conditions exhibiting on occasion various types of disturbing behavior indicates the general prevalence of such behavior to be relatively similar in large and small facilities (again, with the important caveat that the Institutional Population Component did not include data on frequency, duration or intensity of these types of behavior). Simple reweighting of the NMES proportions to the known populations of small and large facilities increases population estimates of persons with behavior problems in small facilities by 82.6%. Again, the adjustments have the effect of suggesting that community-based settings are currently providing residential services to many more thousands of people with problem behavior than would be suggested by the original NMES estimates.

Functional Skills

Proportions of persons with mental retardation and related conditions in small and large mental retardation facilities show small facility residents to much more often relatively independent in functioning. However, if proportional statistics are adjusted to known populations, it is notable that there are as many people estimated to be able to dress without assistance in large facilities as in the smaller community-based facilities. There are nearly as many people able to use the toilet independently in large facilities as in smaller community-based facilities (81% of the small facility estimate) and there are nearly as marry people able to walk across the room independently in the large facilities as in the small facilities (91% of the small facility estimate). Without adjustments to known populations, estimates from NMES would suggest a much greater proportion of people with significant functional limitations living in large facilities than is actually the case. For example with respect to independent toilet use, NMES population estimates indicate that 86.9% of all residents of mental retardation facilities cannot independently use the toilet live in large facilities. Use of alternative population statistics produces an estimate of 76.4% of all residents not independent in toilet use are living in large facilities. As a result, not being able to independently toilet oneself appears less accepted (and perhaps less acceptable) as a criterion for limiting community living opportunities.

Use of Special Equipment

Wheelchair use is considerably more prevalent in large facilities than in small facilities. Adjusting the NMES estimates to reflect the considerably greater number of persons of smaller community facilities and the somewhat smaller number of persons in larger facilities than estimated in NMES would nearly double the estimated wheelchair users in smaller facilities from an estimated 3,237 persons to an estimated 5,929 persons. Reported use of urinary catheters and communication boards/symbol systems as primary means of communication was so limited in the NMES sample that it was affected by reweighting.

Medical Conditions

Applying proportional estimates from NMES to the known populations of facilities as reported by the states has various effects on estimates of medical conditions. For circulatory conditions, reweighting would provide an estimate that slightly more people with mental retardation and related conditions and circulatory system conditions are living in smaller, community based residential facilities than in larger facilities, while in the original NMES estimates only 33.3% of residents with circulatory conditions are indicated to live in the smaller facilities. With proportional adjustments the estimated numbers of people with diabetes in small mental retardation facilities increased by 2,250 (or 82%), or an estimated 46.4% of all mental retardation facility residents with diabetes living in community-based settings (an increase from 29.9% in the original NMES estimates). Reweighting of the NMES sample also makes considerable difference in estimated proportion of residents with frequent constipation living in community settings. In the original NMES estimates, 16.4% of all residents in mental retardation facilities who suffered frequent constipation were in small facilities with reweighting that proportion would be 28.6%.

The proportion of deaths among smaller, community-based facility residents was estimated to be .9%. This was similar to the .8% death rates obtained for small facility residents in the 1977 and 1982 NCRF surveys (Lakin, Hill, & Bruininks, 1985). But the total number of deaths among residents in the smaller residential facilities is probably better estimated by reweighting the NMES sample to known number of residents. The adjusted estimate would be 1,067 deaths in smaller, community facilities as compared with the original NMES estimate of 615 deaths in these facilities.

There are at least three conditions that would have to prevail to make this or any similar alternative estimation procedure adequate and preferable to using original NMES statistics for estimating persons with mental retardation and related conditions in different sizes and types of residential facilities. These include the following:

• States must more accurately report the total number of people making up the populations of facilities meeting the definitions employed in NMES than did the ILTCP, which served as the sample frame for NMES and is the basis for its population estimates. Based on the discussion in Part I of this report, this does appear to be the case.

• The sample size of NMES must be sufficient to yield accurate estimates of facility and population characteristics for settings with 15 or fewer residents. The NMES sample of 326 facilities and over 1,000 residents in facilities with 3-15 "beds" appears more than sufficient to obtain reasonably accurate estimates of the proportional distributions reported for these facilities. Although, for certain data elements, the relatively low occurrence within the sample produced estimates of low reliability, in general relatively adequate sample size was maintained among the smaller facilities.

• Sample members representing the approximately 65,000 community facility residents in the facilities contained in the sample frame must also be reasonably representative of those who were excluded from the sample frame and, thereby, the sample (both 1 and 2 resident places and the facility "types" underrepresented). It is extremely difficult to test whether this condition can be met satisfactorily. In general the residents of small facilities sampled in the 1987 NMES tend to be somewhat more impaired than residents of facilities participating in the 1982 NCRF. As such the NMES estimates of small community facility population characteristics tend to reflect what is known to have taken place within residential services nationwide since the 1982 NCRF. Since 1982 smaller facilities are known to have come to serve considerably more severely impaired people, as thousands of persons with severe and profound levels of mental retardation were released from public and private institutions to community facilities or have entered community facilities directly from their own homes. Table 38 compares estimated populations and proportions by level of mental retardation in the 1982 NCRF and the 1987 NMES, original and adjusted estimates.

The estimated proportional changes in small facility residents by level of mental retardation suggested are generally supported by census statistics gathered on the populations of state institutions. Between June 30, 1982 and 1987 state mental retardation institution populations decreased from 121,479 to 94,696 (White et al., 1989). During that period states reported a total of 43,189 discharges from state institutions. Based on the only available statistics on the placement of state institution discharges for FY 1982, FY 1985 and FY 1987 (the only years in which data were gathered), 50.67% of discharges went to community-based living arrangements other than a natural or adoptive home. In other words, an estimated 21,880 people were discharged to state institutions to community living arrangements between June 30, 1982 and June 30, 1987. If data on the level of mental retardation of FY 1987 releases (the only year available) are applied to these data, the estimated number and proportions of persons with different level of mental retardation entering community facilities would be as shown in Table 39.

Of course, not all releases to community facilities have resulted in successful tenure. In the only two years during the 1982-1987 period in which data were obtained on readmissions, FY 1985 and FY 1987, 33.9% and 32.5% of readmissions, respectively, were from people living in community facilities other than a natural or adoptive home. Adjusting data on total readmissions for FY 1982, 1985 and 1987 with the statistics on readmissions from community settings, and using data on the level of retardation of readmissions from FY 1987 (the only year available), the estimated number and proportion of persons with different levels of retardation leaving community facilities to return to institutions would be as shown in Table 39.

Of course, a limitation of these data is that they assume that people with more severe mental retardation released from institutions are as likely to be among the 51% going to community facilities as are released residents with less severe levels of mental retardation. Put another way, one might question whether it is possible that people with severe or profound mental retardation would be more likely to be among the 49% of institution discharges who did not go to community residential facilities. Unfortunately, the most recent data on this topic (1978) are too dated for contemporary analysis. However, among the estimated 14% of all institution discharges returning to a natural or adoptive home in 1978, no statistically significant differences were noted by degree of mental retardation. While available statistics do not prove absolutely that the populations of community mental retardation facilities have necessary changed in the absolute size and distribution as suggested by related movement statistics, it seems reasonable to estimate that the depopulation of state institutions alone has added over 18,000 people to community residential settings between 1982 and 1987, an estimated 6,000 of whom are profoundly retarded.

As important as these additions are to the number and characteristics of residents of community mental retardation facilities, persons coming from large public institutions comprise a significant minority of persons entering smaller residential facilities. In the 1982 NCRF statistics were gathered on previous place of residence of persons newly admitted between July 1, 1981 and June 30, 1982 to 88% of all facilities operating on June 30,1982. These statistics showed an estimated 13,030 new admissions to smaller community facilities in FY 1982, 27% came from large public facilities. Another 8% came from large private mental retardation facilities and 6% from other types of institutions (nursing homes, mental health facilities, hospitals, etc.). About 28% of new residents came directly from home or independent living situations and 31% came from other community facilities or moved to a new home with their existing residential household (Lakin, Hill, & Bruininks, 1985). Unfortunately no data exist on the characteristics of these now admissions. In addition statistics reported by states on the number of people in large nonstate residential facilities indicated a decrease of above 10,600 residents between June 30, 1982 and June 30, 1987 (Lakin et al., 1989). Presumably most of this number was made up of persons moving to community facilities.

Clearly the group most systematically underrepresented in the mental retardation facilities sample are persons in family/foster care settings. With an average size of 2.6 residents per "facility" (Lakin, Hill, & Bruininks, 198S), the majority of such facilities were automatically excluded from NMES when it was decided to exclude facilities of 1 and 2 residents. A large (unknown) proportion of the others were left unidentified because of the factors discussed in Part I. Given the exclusion of most foster care facilities which served an estimated 22,353 people on June 30, 1987, there is particular interest regarding the extent to which their residents might be represented in data gathered in NMES. Statistics obtained on the 17,147 residents of foster care facilities in 1982 showed that with respect to level of mental retardation foster care residents were quite similar to residents of other small facilities as shown in Table 40. While minor differences are apparent in the level of retardation of specialized foster care and other small facility residents in 1982, these differences were not large. However, age differences were substantial. While 37.4% of foster care residents were 21 years or younger, only 18.0% of other small facility residents were 21 years and younger. Therefore, with respect to NMES statistics, it seems clear that the exclusion of foster care homes of 1 and 2 residents and underrepresentation of the remainder has caused significant underrepresentation of children and youth in the sample and resulting population estimates. This underestimation appears to be about 3% of all residents, about 6,000-8,000 persons 21 years or younger, or an estimated 12%-16% of the expected number of persons of that age.

In summary, there is no way to clearly demonstrate how best to use the NMES statistics to estimate populations of persons with mental retardation and related conditions in mental retardation facilities, especially the smaller facilities. While there is overwhelming evidence that NMES has substantially underrepresented the populations of persons in small facilities, it remains the richest and most comprehensive data base on residential services for persons with mental retardation available. There is much evidence that the general characteristics of small community facility populations are shifting proportionally in the directions suggested by NMES. Unfortunately data do not exist to clearly guide adjusting NMES estimates to known total small facility populations so as to improve the ability to estimate the characteristics of the population. Nevertheless, some "reweighting" is inevitable in the many instances where population estimates needed and where NMES represents the single best data source of estimating the characteristics of residents, costs of residential services, and other data needed about mental retardation facilities. Simple efforts to do so will probably improve the ability of the NMES statistics to describe the population characteristics and residential services of persons living in small, community based residential settings. However, data to establish or justify specific procedures for doing so are not readily available.

Table 37. Characteristics of People in Different Facility Types¹

Table 38. Comparison of 1982 NCRF and 1987 NMES Findings Regarding Small Facility Populations with Mental Retardation

Table 39. Estimated Additions to Community Facilities from State Institutions during the period from 1982 to 1987 by Level of Mental Retardation

Table 40. Comparison of 1982 Foster Care Residents with Small Facility Resident Characteristics as Obtained in the 1982 NCRF and 1987 NMES