MCBS Highlights: Characteristics of Medicare Persons in Long-Term Care Facilities

Introduction

The Medicare Current Beneficiary Survey (MCBS) includes persons living in the community and persons living in long-term care (LTC) facilities. In 1996, 91 percent of Medicare beneficiaries lived in the community all year. Approximately 2 percent of the Medicare population spent some time in an LTC facility but then returned home. The remaining 7 percent of the Medicare population either spent the entire year in an LTC facility or entered the facility during the year and remained a resident at year's end.

Using the 1996 MCBS Cost and Use file, we examine personal characteristics, income, health insurance, and health status of the 7 percent of the Medicare population who are continuing residents in LTC facilities. This group includes many of the sickest and most dependent Medicare persons. We compare their characteristics with those of the relatively healthier group of persons living in the community for the entire year. For a reference point between these extremes, we also include the 2 percent of beneficiaries who spent time in an LTC facility but only for short-term treatment. The objective is to compare differences in characteristics across the entire health spectrum.

One important point about the sample of LTC persons in this article is that it is broader than the sampling frame used in other LTC facility studies. The MCBS is a sample that follows individuals as they enter into any and all LTC facilities for treatment. Other surveys, such as the National Nursing Home Survey, are based on a list of existing certified facilities. This sampling approach generally includes larger and more stable State-licensed nursing facilities, but it tends to understate use of specialized and smaller LTC facilities (Bishop, 1999). More important, as we show later, using a facility-based sampling frame produces a different picture of the characteristics of the Medicare LTC population than using the person-based MCBS sample.

Characteristics of Persons in LTC Facilities

Table 1 shows several variables that display contrasts between individuals residing in LTC facilities, those that had short stays in skilled nursing facilities (SNFs), and persons living in the community. These variables include health insurance, sex, age, marital status, income, and health.

Table 1. Characteristics of Medicare Beneficiaries Living in the Community and in LTC Facilities: 1996.

	Beneficiary Characteristics
	All		Living in the Community		With Short SNF Stays		Living in LTC Facilities
	Number in Thousands	Percent	Number in Thousands	Percent	Number in Thousands	Percent	Number in Thousands	Percent
Total	39,639	100.0	35,932	90.6	803	2.0	2,904	7.3
Sex
Male	17,225	43.5	15,975	44.5	332	41.3	918	31.6
Female	22,414	56.5	19,957	55.5	471	58.7	1,986	68.4
Age
0 - 44 Years	1,643	4.1	1,451	4.0	13	1.6	179	6.2
45 - 64 Years	3,043	7.7	2,820	7.8	41	5.1	183	6.3
65 - 69 Years	9,616	24.3	9,377	26.1	61	7.6	177	6.1
70 - 74 Years	8,851	22.3	8,521	23.7	103	12.8	228	7.9
75 - 79 Years	7,058	17.8	6,534	18.2	210	26.2	314	10.8
80 - 84 Years	5,043	12.7	4,302	12.0	185	23.0	557	19.2
85 Years or Over	4,384	11.1	2,928	8.1	190	23.7	1,266	43.6
Marital Status
Married	20,580	51.9	19,743	54.9	311	38.7	526	18.1
Widowed	12,695	32.0	10,791	30.0	386	48.1	1,518	52.3
Divorced	2,836	7.2	2,587	7.2	46	5.7	203	7.0
Seperated	570	1.4	529	1.5	9	1.1	32	1.1
Never Married	2,908	7.3	2,261	6.3	51	6.4	597	20.6
Income
$15,000 or Less	19,007	48.0	16,269	45.3	428	53.3	2,310	79.5
More than $15,000	20,632	52.0	19,662	54.7	375	46.7	595	20.5
Health Status
Excellent	6,202	15.6	6,059	16.9	78	9.7	65	2.2
Very Good	10,187	25.7	9,750	27.1	113	14.1	324	11.2
Good	11,952	30.2	10,793	30.0	200	24.9	959	33.0
Fair	7,403	18.7	6,153	17.1	162	20.2	1,088	37.5
Poor	3,812	9.6	3,102	8.6	245	30.5	465	16.0
Functional Limitations
None	24,529	61.9	24,221	67.4	161	20.0	148	5.1
IADL Only	2,336	5.9	2,088	5.8	62	7.7	187	6.4
1 or 2 ADLs	7,046	17.8	6,349	17.7	247	30.8	450	15.5
3 or More ADLs	5,727	14.4	3,274	9.1	333	41.5	2,120	73.0
Health Insurance
None	18	0.0	12	0.0	0	0.0	6	0.2
FFS Only	5,279	13.3	4,550	12.7	41	5.1	688	23.7
Medicare Risk HMO	4,725	11.9	4,566	12.7	45	5.6	114	3.9
Medicaid	6,411	16.2	4,409	12.3	163	20.3	1,840	63.4
Employer-Sponsored	13,368	33.7	12,800	35.6	321	40.0	247	8.5
Self-Purchased	9,365	23.6	9,135	25.4	230	28.6	0	0.0
All Other	472	1.2	460	1.3	4	0.5	8	0.3

NOTES: LT C is long-term care. SNF is skilled nursing facility. IADL is instrumental activity of daily living. ADLs is activities of daily living. FFS is fee-for-service. HMO is health maintenance organization.

SOURCE: Medicare Current Beneficiary Cost and Use File, 1996.

One of the strongest trends that occurs as beneficiaries move from the community into LTC facilities is the shift in supplementary health care insurance from private plans to Medicaid. Of those individuals residing in the community, only 12.3 percent had Medicaid coverage, but 61.0 percent were enrolled in either an employer-sponsored or self-purchased private plan. These numbers are almost exactly reversed for those in facilities. Only 8.5 percent of the population had a private health insurance plan, while Medicaid covered 63.4 percent of them. One contributing factor could be income. People with annual incomes of $15,000 or less make up 45.3 percent of the community population but account for 79.5 percent of the facility population. Because of these low incomes, the facility population generally does not have the financial ability or need, if they are eligible for Medicaid, to acquire supplementary private insurance.

There is a noticeable shift in the gender ratio toward a population heavily dominated by females in facilities. Females make up a little more than one-half (55.5 percent) of those residing in a community setting, compared with more than two-thirds (68.4 percent) of those in a facility. (As we discuss later, LTC populations drawn from a narrower sample of LTC facilities show an even greater ratio of females to males.) The larger share of females in facilities can be attributed mostly to the fact that females live longer then males. Looking at people age 85 or over, it is apparent that the facility population is considerably older. This group makes up only 8.1 percent of the community population, 23.7 percent of those with SNF stays, and 43.6 percent of the facility population. The longer life expectancies of females could also explain why more than one-half (54.9 percent) of those in the community are married, but more than one-half (52.3 percent) of those in facilities are widowed.

The health status of these three populations illustrates another key difference. Declining health is obviously one of the determining factors as to whether a person remains in the community or requires short-term or long-term care in a facility. A clear pattern of declining health is evident moving from the individuals in the community to individuals with SNF stays and then to individuals in facilities. Those who reported their health as being either fair or poor were 25.7 percent of the community population, 50.7 percent of the SNF population, and 53.5 percent of the facility population. (The reason the facility share is not higher is because facility nurses make the health-status judgments, and they often compare an individual with others residing in the same facility, rather than with the healthier community population). Further, the percentage of individuals with no functional limitations dropped from 67.4 percent of those in the community to 20.0 percent for those with a SNF stay and then to 5.1 percent of those in facilities. These trends show that, as a person's health declines, they must seek health care outside of the home, with the sickest moving to LTC facilities. However, although declining health is a necessary condition for long-term facility care, it does not appear to be the sole factor. Long-term facility care is expensive, and a person in poor health also needs either private means or supplementary health insurance to finance their LTC in a facility.

To examine the importance of Medicaid to receiving long-term facility care, we looked only at the sickest portion of the Medicare population (Table 2). As our dividing line, we used persons with limitations in three or more of their activities of daily living (ADLs) (eating, dressing, bathing, walking, transferring in and out of a chair, and using the toilet). This is generally considered a reasonable dividing line to identify persons whose needs are very difficult to attend to in the home setting. More than one-half (57.2 percent) of these persons still remain in the community. However, breaking this group down by income and Medicaid coverage, the ratio of community to facility residence changes dramatically. Among people with annual incomes of $15,000 or less, persons with Medicaid were more than twice as likely (60.1 percent) to be in a facility than those without Medicaid (23.7 percent). An even greater disparity exists for people with incomes of more than $15,000. Nearly three-quarters (74.9 percent) of persons with Medicaid were in a facility, as opposed to only 18.9 percent of persons without Medicaid. By controlling for income and health status, a high correlation is seen between Medicaid enrollment and whether a person in poor health receives long-term facility care.

Table 2. Medicare Beneficiaries with Functional Limitations in Three or More Activities of Daily Living, by Income, Medicaid Status, and Residence: 1996.

Total Income Level	Medicaid Status	Number in Thousands	Living in the Community		Living in LTC Facilities
			Number in Thousands	Percent	Thousands	Percent
$15,000 or Less	No	1,726	1,203	69.7	409	23.7
	Yes	2,075	755	36.4	1,247	60.1
More than $15,000	No	1,746	1,282	73.4	330	18.9
	Yes	179	35	19.6	134	74.9

NOTE: LTC is long-term care.

SOURCE: Medicare Current Beneficiary Cost and Use File, 1996.

Where Do People Receive LTC?

The distribution of Medicare persons by type of LTC facility is shown in Figure 1. About 73 percent of persons receiving long-term facility care are located in nursing facilities. The largest share (58 percent) are in Medicare-certified nursing facilities, 12 percent are in nursing facilities certified for Medicaid, and 2 percent are in other licensed nursing facilities that do not participate in either Medicare or Medicaid. The remaining 27 percent of long-term facility care recipients are found in the following settings: 7 percent in facilities that care for mentally retarded/developmentally disabled (MR/DD) persons, 4 percent in facilities for persons with mental illness, and 16 percent in other LTC facilities. This last group consists primarily of group homes that offer board and care but are not nursing care or specialized treatment facilities. Figure 1 suggests that limiting LTC to nursing facilities excludes more than one-quarter of Medicare persons confined to other types of LTC settings.

Figure 1. Percent of Medicare Long-Term Care Beneficiaries, by Type of Facility: 1996.

Population Characteristics by Type of Facility

As previously mentioned, the person-based survey design of the MCBS allows for the collection of data from a broader spectrum of LTC institutions than surveys that sample from facility lists. By breaking down the facility population into types of LTC facilities, it becomes clear that the populations served by each facility type are not homogeneous (Table 3).

Table 3. Characteristics of Medicare Beneficiaries in Long-Term Care Facilities, by Type of Facility: 1996.

Characteristic	Medicare-Certified Nursing Homes		Medicaid-Certified Nursing Homes		Other Nursing Homes		Mental Health Facilities		MR/DD Facilities		All Other Facilities
	Number in Thousands	Percent	Number in Thousands	Percent	Number in Thousands	Percent	Number in Thousands	Percent	Number in Thousands	Percent	Number in Thousands	Percent
Total	1,410	58.2	302	12.5	51	2.1	88	3.6	179	7.4	393	16.2
Sex
Male	329	23.3	120	39.7	27	52.9	46	52.3	121	67.6	123	31.3
Female	1,081	76.7	182	60.3	24	47.1	43	48.9	57	31.8	270	68.7
Average Age	83	—	83	—	80	—	57	—	49	—	82	—
Health Status
Excellent	20	1.4	10	3.3	0	0.0	8	9.1	14	7.8	9	2.3
Very Good	118	8.4	21	7.0	4	7.8	10	11.4	52	29.1	63	16.0
Good	485	34.4	88	29.1	26	51.0	45	51.1	78	43.6	105	26.7
Fair	558	39.6	139	46.0	12	23.5	15	17.0	30	16.8	187	47.6
Poor	228	16.2	45	14.9	8	15.7	10	11.4	4	2.2	28	7.1
Functional Limitations
None	33	2.3	5	1.7	0	0.0	20	22.7	16	8.9	30	7.6
IADL Only	25	1.8	14	4.6	5	9.8	39	44.3	56	31.3	20	5.1
1 or 2 ADLs	128	9.1	61	20.2	7	13.7	4	4.5	48	26.8	100	25.4
3 or More ADLs	1225	86.9	222	73.5	39	76.5	25	28.4	59	33.0	243	61.8
Health Insurance
None	0	0.0	0	0.0	0	0.0	0	0.0	0	0.0	0	0.0
FFS Only	218	15.5	98	32.5	42	82.4	34	38.6	4	2.2	174	44.3
Medicare Risk HMO	55	3.9	2	0.7	0	0.0	0	0.0	0	0.0	29	7.4
Medicaid	1,025	72.7	190	62.9	5	9.8	42	47.7	174	97.2	139	35.4
Employer-Sponsored	111	7.9	12	4.0	4	7.8	12	13.6	0	0.0	50	12.7
Self-Purchased	0	0.0	0	0.0	0	0.0	0	0.0	0	0.0	0	0.0
All Other	2	0.1	0	0.0	0	0.0	0	0.0	0	0.0	1	0.3

NOTES: MR/DD is mentally retarded/developmentally disabled. IADL is instrumental activity of daily living. ADLs is activities of daily living. FFS is fee-for-service. HMO is health maintenance organization.

SOURCE: Medicare Current Beneficiary Survey Cost and Use File, 1996.

Studies based on a sample from a master LTC facility list show the ratio of females to males in LTC facilities to be approximately 3 to 1 (Bishop, 1999). Using the MCBS person sample, that ratio holds true—but only for Medicare-certified nursing facilities (Table 3). Other facilities such as those that specialize in care for the mentally ill or MR/DD have a higher concentration of males. These two specialized types of institutions also differ from other types of LTC facilities by having much younger and physically healthier populations. Another important difference by type of facility is in the level of Medicaid coverage. Only about one-third (35.4 percent) of the persons in the “Other” category (consisting mainly of group homes) have Medicaid coverage, compared with almost all (97.2 percent) of those in MR/DD facilities. These differences between facility types suggest that Medicare's long-term facility population is more heterogeneous than commonly believed based on profiles of Medicare beneficiaries in certified nursing facilities.