Analyzing Nursing Home Characteristics: Issues In Comparing State And Federal Data Sources. by Joel B. Cowen' Coordinator Health Services Research College ofMedicine at Rockford University ofIllinois Introduction Aging of the population is increasing the importance of long-term care and nursing homes as a component of the health care system in the United States. Policy planners need to fully understand the nature of the industry, as well as emerging trends. Nursing homes, as we know them currently, are a relatively new entity on the health care landscape. The present long- term care system largely emerged in the early 1970s following the impact of Medicare. Earlier facilities hke board and care homes had sprouted up after the passage of social security, along with was activity by some churches and fraternal groups who had built retirement centers for their members. Since nursing homes are only a couple of decades old and changes in regulations and financing impacting them are being seen with increasing frequency, nursing homes could well change in form or format into the next century, when the elderly portion of the population will move toward one-fifth as baby boomers edge into their senior years. Planning for an essential service like long-term care requires accurate information. Conjecture cannot take the place of data vital for shaping the nature of future nursing homes. However, despite the importance of nursing homes as part of health care, no comprehensive information system is in place nor are basic definitions of data elements accepted in a widespread manner. The quality of planning suffers when information is limited. For instance, the impact on nursing homes of recent changes in Medicare reimbursement on hospitals needs to be known. In this paf)er, current data for nursing homes is analyzed. The two major federal data sets for nursing home facility and resident characteristics, plus similar informafion collected in one major state, Illinois, are reviewed. Data elements obtained and their definitions are compared and, finally, recommendations made as to how to improve the present situation. Nursing Home Data Sources National Nursing Home Survey The National Nursing Home Survey (NNHS) is a continuing series of national sample surveys of nursing homes, their residents and staffs by the National Center for Health Statistics (NCHS). Three surveys have taken place, in 1973-74, 1977 and 1985. No plans have yet been announced for a fourth survey. The surveys employ a stratified two-stage probability design, first the selection of facilities, then the selection of resi- dents and employees. Data has been collected using both personal interviews and forms for self-completion. In 1985, added information was collected from relatives of residents, the "next of kin" questionnaire. In 1977 and 1985, a sample was also taken of persons discharged from the home in the past year, whether alive or dead. The NNHS is the only national survey which includes variables for individuals which can be analyzed against each other. Other sources group data, which cannot be crosstabbed or correlated. All data is available on tape from the National Center for Health Statistics. Figure 1 shows the available variables. Tapes are also available for the states of California, Illinois, Massachusetts, New York and Texas. More cases were provided in these states so that reliable estimates could be made. Results appear in written form in Vital and Health Statistics . Series 13, Nos. 97, 98, 102, 103, and 1 15 plus Advance Dala, Nos. 131, 135, 142, 147, and 152 from NCHS. lASSIST Quarterly Figure 1 SUMMARY OF 1985 NNHS DATA TAPES BY TYPE OF HLE Facility file Resident file Facility number Ownership code Number of beds (1985 and 1984) Certification status Per diem rates by certification status Admissions (1984) Residents days (1984) Services offered to rsidents Services offered to non-residents Physician service arrangements Full- and part-time staff Part-time staff hours Nursing staff hours Volunteer staff (Geographic region recode DHHS administrative regions MSA recode Fadbty weight Facility number Age Sex Race Hispanic origin Marital status at admission and currently Living children Date of last admission Residence before admission Hospital stays while a resident Previous nursing home stays Diagnoses at admission and currently Mental disorders Therapy services received Vision and hearing status Activities of daily living Adapted instrumental activities daily living Behavioral problems Disorientabon or memory impairment Deprssion, anxiety, fearfulness, or worry Sources of payment at admission and last month Total monthly charge for care last month Amount paid by source last month Resident weight Record length Block size 665 19,950 1,078 Record length Block size. 873 17,460 5;238 Discharge file Nursing staff file Facility number Age at discharge or date of birth Sex Race Hispanic origin Marital status at admission and at discharge Date of admission and discharge Discharge status (alive/dead) Residence before admission Residence after discharge for live discharges Hospital stays while a resident Nursing home stays before and after sample stay Diagnoses at admission and at discharge Mobility staff Continence status Sources of payment at admission and at discharge Discharge weight Facility number Member of staff or other arrangement Type of position Length of work experience Hours worked Salary Services performed Employment conditions Sex and age Ethnicity Marital status Children living at home Education Staff weight Record length Block size Number of records 544 21,760 6,017 Record length Block size Number of records 307 21,490 2,760 Exfjense file Facility number Expenses and revenues Expense weight Record length Block size Number of records 366 18300 731 Fall 1992 17 Inventory ofLong-Term Care Places The federal Inventory of Long-Term Care Places (ILTCP), formerly part of the National Master Facility Inventory (NMFI), was conducted in 1986 by the Census Bureau for the National Center for Health Statistics. This differs from the NNHS in that the entire sample of "nursing and related care homes" is covered. Variables are limited to ownership, certification status, number of beds, residents, and race of residents. Information on residents is tabulated in grouped data for reporting. One way that the ILTCP differs from the NNHS is that special homes for the mentally ill, developmentally disabled, and other groups with special handicaps are included. Also available on tape from NCHS, the full set of variables is shown in Figure 2. Both the NNHS and ILTCP are also available through the Inter-University Consortium for Political and Social Research. Figure 2 SUMMARY OF NATIONAL MASTER FACILITY INVENTORY DATA TAPES BY TYPE OF FACILITY Hospitals Nursing homes and other health facilities Name Name of administrator Ownership Tyf>e of facility Number of beds Days of care Discharge Admissions Type of service Outpatient visits Employees facilities and service offered 1971 1972 1971 1974 1975 1976 Name Adress Number of beds Ownership Type of facility Ages served Sexes served Number of residents 1971 1973 1976 Record length 840 Block size 8,400 748 4,488 7,480 1 748 4,488 7,438 1 748 4,488 7,370 1 748 4,488 7,336 1 748 4,488 7,271 1 Record length Block size Number of records Number of reels 600 3,600 26,773 1 196 1,176 26,003 1 210 6,720 26,748 Number of reels 1 1 Illinois Annual Survey ofLong-Term Care Facilities Most states collect data on their nursing homes and their residents. This is usually made necessary for the state's licen- sure certification or planning functions. Illinois is typical. The Illinois Department of Public Health (IDPH) has con- ducted annual surveys of long-term care facilities since 1981. These surveys are carried out under the state's Certificate of Need Act which authorizes data collection for planning purposes. In addition to using the data to create an inventory of long-term care services and bed need plan, IDPH makes survey data available to others who have an interest in long-term care, both public and private. Other state agencies also constitute a major user. The lUinois survey format is relatively standardized year to year with supplemental "special" studies each year. Nearly a thousand licensed facilities (including mental heallh/DD) receive the questionnaires and participation is almost universal. UICOM-R Long-Term Care Report The Health Services Research office of the University of Illinois College of Medicine at Rockford has created reports characterizing nursing homes in northwest Illinois. These reports, completed in 1980, 1985 and 1990, utilize information from the IDPH Long-Term Care Facility Survey. This effort is notable because few local area studies are conducted which track changes in the industry on a local area basis. State reports generally do not evaluate long-term changes or focus on regional differences. lASSIST Quarterly Data Elements In this section, certain data elements have been selected to illustrate the availability of nursing home data, so as to reveal potential sources, their differing methods of gathering, commonalities and differences in definition. What Is a Nursing Home? Defining a nursing home is not a clear and simple task. The NNHS includes all types of "nursing homes" regardless of their level of care, participation in Medicare or Medicaid or licensure. No "Board and Care" homes were included or those providing residential care. They define nursing home as: Facilities with three or more beds that provide to adults who require either nursing care or personal care (such as help with bathing, correspondence, walking, eating, using the toilet, or dressing) and/or supervision over such activities as money management, ambulation, and shopping. Facihiies providing care solely to the mentally retarded and mentally ill are excluded. A nursing home may be either freestanding or a distinct unit of a larger facility. Illinois relies on licensure for its survey definition. In Illinois, a nursing home is defined as a: "private home, institution, building, residence, or any other place which provides personal care, sheltered care or nursing for three or more persons who are not related to the owner." Long-term care institutions in Illinois are further classified into two types of care, nursing and sheltered care. Nursing care includes the provision of diagnostic, therapeutic and rehabilitative care under a patient's plan of care as prescribed by a physician. In addition to the medically oriented care given by nurses and the living assistance given by aides, other services commonly provided by a facility that provides nursing care includes physical therapy, speech therapy, occupational therapy and social activities. There are two levels of nursing care, skilled and intermediate , with the levels differing in the amount of available nursing expertise and supervision provided to residents. Sheltered care includes the provision of personal care and support in daily activities with limited nursing consultation available, such as the taking of medications. The term "sheltered care" is somewhat unique to Illinois. Other states tend to utilize the terms "assisted living" or "personal care." The federal NNHS survey relies on Medicare/Medicaid definitions for designation of skilled or intermediate care levels. Other beds are shown as "not certified." Both units of government agree in that a facility must have three beds and provide nursing or personal care. Illinois does not consider personal care to determine a nursing home unless there is nursing consultation, such as the taking of medication. Terms common to definitions such as nursing care and personal care are not precisely defined. Another area for possible confusion is whether facilities such as those for the mentally ill or developmentally disabled are counted. The NNHS excludes these. The ILTCP includes them as does Illinois. The ILTCP, which begins with categories similar to the NNHS, goes beyond these categories to also include homes for unwed mothers, substance abuse, orphans and the terminally ill (hospice). In general, the definition of nursing home in this country is still imprecise, leaving board and care, congregate living and certain retirement centers and specialized facilities in a zone of uncertainty. Beds Beds are classified in various ways such as licensed or unlicensed, set-up and staffed, or occupied. The NNHS and ILTCP both primarily use set-up and staffed, while Illinois uses licensed beds. Licensed beds usually reflect capacity whether actually used or not. The creation of swing beds and distinct unit skilled nursing beds at hospitals has increased in recent years. Swing beds are those that can be used for either acute or extended care through reclassification of the patient who doesn't actually move from the bed. Swing beds are certified by the Medicare program. Distinct units provide only extended care. For the most part they are utilized for hospital patients no longer needing acute hospital nursing care but who are not yet Fall 1992 19 well enough to return home. Under the "DRG" reimbursement system, hospitals benefit financially from discharges to a lower level of service. The NNHS totally excludes hospital-based beds. On the other hand, the ILTCP includes long-term care units in hospi- tals. Presumably this includes both swing beds and distinct units. Illinois counts distinct units as a "nursing home" category, but does not consider swing beds in this category. As is probably obvious, counts of beds can differ greatly according to the counting method. Figure 3 shows some of these conflicting counts. Figure 3 COMPARISON OF BED COUNTS: NNHS, ILTCP, AND IDPH Nursing Homes in U.S. Total NNHS ILTCP Nursing Care Hospitals Not Certified Residential Nursing 14,400 4,700 Homes in Illinois 16,388 • 734 9,258 Total ILTCP IDPH Nursing Care Hospital Based Residenlia 744 25 48 898 734 Admission, Discharge and Slay Length of stay is important for policy decisions and planning of resources, not to mention actuarial calculations such as those for the long-term care insurance industry. Average length of stay cannot be calculated with ease as is done in hospitals: Patient Days = Average Length Discharges of Stay Nursing home stays more often cross multiple years and may involve intervening hospital stays. The ILTCP collects annual admissions and number of "residents last night." Average length can only be calculated if the assumption is made that "residents last night" times 365 = an estimate of patient days. IDPH collects patient days, admissions and discharges so that an estimate of average length of stay can be calculated. Again, the year to year crossover can be a problem in such calculations. The NNHS is far more precise in its treatment of stay length. This is important because a good source is needed which differentiates the characteristics of stay types, especially the nature of post-hospital short stays from longer term chronic type stays. The differing nature of different "types" of nursing home residents is important for policy. The NNHS provides for a question on duration of stay at discharge. Recently they have reconsidered this indicator of stay length in favor of the long-term care use history of individuals. An individual may be admitted and discharged several limes a year and, in fact, that pattern has been found to be common. Additionally, nursing homes treat stay definitions differently, especially with regard lo temporary transfers to hospitals. Some consider the movement to be a discharge with formal readmission on return, while others make no such change in ASSIST Quarterly status. Some facilities include the hospital stay as part of the stay, while others exclude them or calculate "bed-hold" days. Dlinois provides no instructions on how to deal with the reporting of "bed-hold" days. This is left to the institu- tion. Another element of admissions and discharges is the place of origin and discharge - Where are residents coming from? and Where are they going? The ILTCP does not collect this, but both the NNHS and Illinois do. The NNHS collects "Uving arrangement prior to admission and after discharge," while Illinois asks for "admissions from" and "discharges to." The categories are similar except that the NNHS is far more precise in terms of home residence. A comparison is shown below in Figure 4. Figure 4 LTC ADMISSIONS AND DISCHARGES COMPARISON OF NNHS AND IDPH NNHS IDPH Own Home or Apartment Private Residences Relative's Home or Apartment Other Private Home or Apartment Retirement Home Boarding House, Room Other Nursing Home Other Nursing Home General Hospital General Hospital Mental Hospital Psychiatric Hospital State Mental Health Facility Community Mental Health Residences Chronic Disease Hospice Other Other Resident Demographic Characteristics Recording the demographic characteristics of residents provides an essential component of any description or analysis of the nursing home industry. Age, race, and sex constitute the core of these descriptors. Birthdate of the resident is collected by the NNHS, allowing any age groupings. The ILTCP utilizes three groups: 0-21, 22-64 and 65-t-. IDPH provides for seven groups: 0-17, 18-44, 45-59, 60-64, 65-74, 75-84 and 85-t-. Sex (gender) is collected by the NNHS and IDPH, but not the ILTCP. Nursing homes tend to be dominantly female. Like many of these data elements, race is classified three different ways. The NNHS applies the usual census format in which race (white, black, American Indian, and Asian) is a separate variable from ethnic origin (Hispanic). Illinois, however, uses combined racial/ethnic groupings (white, non-Hispanic; black, non-Hispanic; Asian, non-Hispanic; and Hispanic). The ILTCP collects only the number of black and Hispanic residents, no other racial groups. Two resident elements in the NNHS only are marital status (current and at admission) and number of living children. Health and Activity Health of the individual is often expressed either through categorization of the major condition or disease resulting in the admission or indications of which activities of daily living (ADLs). The ILTCP does not record diagnosis at all. Illinois utilizes groupings of ICD-9 codes such as "Diseases of the Circula- tory System" or "Musculoskeletal Diseases." The NNHS Usts certain common reasons for nursing home placement such as Stroke, Hip Fracture and Alzheimer's Disease. The NNHS also collects indicators of health status prior to admission, including DRG if hospitalized. Fall 1992 Figure 5 DEMOGRAPHIC CHARACTERISTICS OF NURSING HOME RESIDENTS Indicator Age Sex Race/Ethnic Marital Status Living Children? NNHS Individual Years Collected Yes White, Black, Amer. Indian, Asian Hispanic Yes, current and at admission Yes ILTCP 0-21 22-64 65-t- Grouped Only No Black, Hispanic Only No No IDPH 0-17 18-44 45-59 60-64 75-84 85-H Grouped Only Yes White, non-Hisp. Black, non-Hisp. Asian, non-Hisp. Hispanic No No Figure 6 HEALTH AND ACTIVITY STATUS VARIABLES FOR NURSING HOME RESIDENTS IDPHIndicator NNHS ILTCP Activities Bathing Bathing Bathing of daily Dressing Dressing Dressing living Eating Eating Eating (ALDs) Walking Inside Walking Mobility Walking Outside Medication Toileting Toileting Shopping or Orientation Transfer Letter Writing Diagnosis DRG if Hospitalized No 15 Groupings Disease or Condition of ICD-9 Resulting in Admission 15 Catergories such as Hip Fractures; Prior Health Status 22 {ASSIST Quarterly Payment Source The ways that residents pay for care is an important issue in the delivery of care. The ILTCP does not collect payment source. All sources of payment are indicated on the NNHS, while IDPH obtains only major payment source. The NNHS includes insurance within private pay, while IDPH breaks it out. Nursing Home Trends Despite their drawbacks and lack of standardization, the studies reviewed in this report can be used to form a picture of trends in nursing homes and their residents. The focus in this section is on recent changes in the industry. Figure 7 PAYMENT SOURCE VARIABLES FOR NURSING HOME RESIDENTS IDPHNNHS Medicare Medicare Medicaid Medicaid Private (own) Pay Private (own) Pay VA VA State Agency State Agency Other Public Pay Other Public Pay Church, Foundation, Agency Insurance Life Care Funds ILTCP (1967-1986 unless otherwise noted) * The number of nursing homes nationally grew 18.2%. * The number of nursing home beds grew 105.0%. * The number of nursing home residents grew 106.5%. * Persons 65+ in nursing homes grew 45.5% from 1967-1976, but then declined 6.2% from 1976 to 1986. * Occupancy stayed around 92%. NNHS (1977-85 unless otherwise noted) * The number of elderly patients discharged from hospitals to nursing homes increased 36% from 1982 to 1985. * The proportion of nursing homes affiliated with a nursing home chain rose significantly 1977-85 from 28% to 41% of all facilities. * Discharges rose 9.5%. * Women in nursing homes aged 85-f rose from 34% to 4 1% of residents. * Race and ethnic status was obtained for the first time in 1985. Minorities were underrepresented and generally younger than the white population. * The proportion of individuals not dependent on help for mobility or continence dropped from 40.1% to 30.1%. * Medicare-covered days in skilled nursing facilities per 1,0(X) beneficiaries dropped from 370 in 1977 to 320 in Fall 1992 23 1984. * Residency in a nursing home for persons 85+ dropped from 257 per 1 ,000 in 1974 lo 220 in 1985. * The average length of stay rose from 2.7 to 2.9 years. * The proportion of nursing home residents with mental disorders rose, while circulatory disorders fell. UICOM-R gPPH Data for Northwest Illinois) * The ratio of beds per thousand population 65+ fell from 92.9 in 1980 to 85.9 in 1990. The elderly population is increasing more rapidly than the facilities for care. About 6.3% of persons aged 65 years and up reside in area long- term care facilities, down from 7.3% a decade ago. * The overall annual occupancy of general long-term care facilities in northwest Illinois is 87.3% based on beds licensed. Occupancy rates have increased since 1980 when the corresponding rate was 84.4%. * Just under half (47.7%) of northwest Illinois long-term care facilities are owned by for-profit concerns, down from 60.0% in 1981. The remainder are owned by churches (18.5%), government (13.9%) and not-for-profit corporations (20.0%). * Hospitals are increasingly entering the long-term care "business." Nine of the fifteen hospitals in northwest Illinois have swing beds or distinct units. * Most general long-term care residents are age 75 or older (81.0%), female (75.2%), and white (97.3%). Residents aged 85 and over made up 47.7% of residents in 1990 but only 39.2% in 1980. The median age is now estimated to be 84.3 years, up from 82.1 in 1981. * Leading admission diagnoses include the circulatory system (28.6%), nervous system (13.8%), and musculoskeletal (1 1.8%). One in ten residents is reported to have Alzheimer's disease. Mental illness, nervous system and musculoskeletal have risen, circulatory has declined. * Nursing home residents are highly dependent on others for performing certain activities of daily living (ADLs) including bathing (62.8%) and toileting (56.6%). Dependency rates have been increasing. * Medicaid was the source of payment for 46.4% of the long-term care residents in 1989. Private payers covered 48.9%. Some of the remainder were covered by Medicare (2.7%). Medicaid coverage has been rising. * The average daily charge for a skilled care double bed in 1990 was S73, up from $48 in 1985. Intermediate care averaged $56 for a double, up from $41 in 1985. Many institutions make additional charges for supplementary services as rehabihtation therapies, bandages, or assistance in bathing. * 65.2% of area beds are certified for Medicaid residents, but only 6.3% are certified for Medicare. Medicare certification has been stable or declining. Improving Nursing Home Data OBRA Data Requirements Beginning this year, all nursing homes certified to provide care under Medicare or Medicaid must use assessment instfuments required by the state and approved by HCFA. The insffument must include a uniform minimum data set (MDS) of care screening and assessment elements with common definitions. The MDS contains a comprehensive set of data elements which describe most nursing home residents nationally. If modified to add certain elements, the MDS could form the core of a standardized insu^ment which could be collected and analyzed on a national basis to analyze the characteristics of nursing home residents (see Figure 8). Other Sources For homes certified by Medicare or Medicaid, financially-centered reports are filed with government or a fiscal interme- diary. These "cost reports" generally include a great deal of information on the nature of the facility. (ASSIST Quarterly Conclusion As has been shown in this paper, data describing nursing homes nationally is haphazard at best, and lacking in standardi- zation of definitions. Sources usually cannot be compared to each other, because of differing coverage, variables, and definitions. Of the two federal sources, one is a f)eriodic sample which is relatively comprehensive, which employs several sub-component studies. The other is a periodic census with relatively sparse variables. Neither follows a regular schedule. One was last completed in 1985 and the other in 1986. No immediate plans are currently in place for repeat- ing these studies so as to yield more timely data. Another data source is the surveys performed by individual states. Most states have annua] surveys of the type exempli- fied by Illinois. Stales could form the framework for regular assessments of nursing homes as long as guidehnes for definition and collection are put forward by the federal government. Much like the vital statistics system operates, states could report to the National Center for Health Statistics, which could compile the information. Long-term care is too important a component of health care not to have systematic data reporting. Action is needed in this direction. 1 Presented at the lASSIST 92 Conference held in Madison, Wisconsin, U.S.A. May 26 - 29, 1992. Fall 1992