New Opportunities for Research in Cognitive Aging: The National Nursing Home Resident Assessment Instrument
Brant E. Fries
Institute of Gerontology
University of Michigan
In preparing this ìThe Cutting Edge, ì I had the pleasure to read through previous printed columns describing state-of-the-art analyses of psychological and cognitive function in older persons. I realized that seasoned readers will find something different in my words here. Rather than address a particular research topic, I want to introduce a new source of information for researchers. To do so, I will describe some of the work in which I and a group of US and international researchers are involved that includes, but is much broader than only the issues traditional examined here. My hope is that researchers in cognitive aging will find in these descriptions new directions and opportunities.
The roots of this work is the National Nursing Home Resident Assessment Instrument (RAI) for Nursing Homes. The RAI was a regulatory response by the Federal government to problems seen in nursing come care. Originally recommended by an Institute of Medicine report on assuring quality of care in nursing homes (Institute of Medicine, 1986), the Omnibus Budget Reconciliation Act of 1987 (OBRA-87) mandated a standardized assessment of all nursing home residents. The Health Care Financing Administration (HCFA), charged with this implementation, contracted with our consortium to develop and test an instrument. The consortium consisted of Catherine Hawes (Research Triangle Institute), John Morris (Hebrew Rehabilitation Center for Aged, Boston), Vince Mor (Brown University), Charles Phillips (now with Menorah Park Center for the Aging), and me. The charge was to produce as assessment system that would improve care provided to individual nursing home residents through improved, individualized, and targeted care plans. The emphasis on ìsystemî reflects our view that what is needed is more than only an assessment instrument. Thus, the RAI was designed with two inter-related components. The Minimum Data Set (MDS), contains the core items necessary for a comprehensive assessment of nursing facility residents. It also provides ìtriggersî (individual items or combinations of MDS elements) to identify residents for whom specific Resident Assessment Protocols (RAPs) - the second part of the system - will be considered. RAPs were developed for each of 18 major problem areas associated with nursing home residents, such as delirium, communication, falls, psychosocial well-being, and cognitive loss. Each RAP provides guidelines for the development of care plans, including suggestions for additional information needed and state-of-the art summary of options for care planning and service provision. By law, full assessments are performed on admission, at least annually thereafter, and upon significant change in the residentís status. Instruments, instructions, and training materials have all been developed and widely distributed by public and private sources. (Morris, et. al., 1990).
The core of the RAI is the MDS, a broad assessment instrument with over 300 individual items that not only describes the nursing needs of residents, but also incorporates measures of residentsí strengths and psychosocial needs. In the area of cognition, the MDS has items describing decision-making, short and long-term memory, communication, indicators of delirium, and diagnoses. The MDS development included extensive testing, dozens of major drafts, and broad input from hundreds of clinicians, administrators, regulators, industry representatives, and consumer advocates. Considerable attention was placed on the specification of time frames, exclusions or delimiters (e.g., score how a resident eats, regardless of skill), and examples. Multiple-state testing of the MDS showed very good reliability across the entire instrument (Hawes et al., 1995). The MDS data is gathered under the supervision of a registered nurse, using all sources of information, including the resident him/herself, family, facility staff, the medical record, etc. The RAI is mandated for use in all U.S. nursing homes that qualify for federal paymentsóvirtually all of the approximately 16,000 nursing homes nation-wideówith over 3 million assessments performed each year.
The RAI achieves good results as it is integral to the process of planning care for a nursing home resident. With this direct linkage, facility staff are dedicated to accumulating excellent data. An evaluation of the US system has shown that as soon as little as two years after its introduction, the RAI was linked to both improved process of care and improved outcomes (these results will appear in a set of five articles in a forthcoming issue of the Journal of American Geriatrics Society). In turn, after the clinical purpose has been satisfied, the RAI can provide information for a variety of other applications, including case-mix (intensity-based) payment to facilities, assessment of quality of care, and policy research (Fries, 1997). As we learn to use the RAI for multiple purposes, different applications offset incentives for particular item responses, such as ìup-codingî to receive more payment or ìdown-codingî to avoid identification of quality problems.
I will focus here on the work done by me and the other RAI authors to use MDS data for a variety of studies ranging from epidemiology to health services research. Much of this work has been enabled by an accumulation of a large data set of MDS records at the University of Michigan.
The University of Michigan Assessment Archives Project (UMAAP) anticipates the development of a nationwide data system of MDS assessments by the federal government and provides a prototype of such a data system for investigating its scope, technical feasibility, and potential uses. Currently, UMAAP has complete data from 11 states for up to five years (1992-1996). With over 4 million assessments performed on approximately 2.5 million individual residents. and given the number of resident characteristic contained in the MDS, UMAAP may be the largest health care data sets ever assembled. In addition to size, however, it contains multiple assessments for many residents, permitting examination of outcomes and risk factors.
Some of the research we are currently addressing includes: describe the characteristics of rarer or emerging nursing home populations, such as frail centenarians, subacute care patients, or those with psychiatric disorders; and predicting outcomes for special types of residents. For this work, UMAAP is truly unique. The number of individuals and facilities permits us to examine with substantial power even rare types of residents, for example, the cognitively intact very old or virtually any gender/ethnic combination desired. Exploratory analysis can be performed on subsamples of substantial size, with large independent samples available for validation. With multiple states, results are not specific to regional, financing, or practice patterns, yet across-state comparisons are also possible. While UMAAP is not generally available, other researchers may have access to MDS data from individual facilities, states that have computerized the MDS, or from the federal data base that HCFA is planning to assemble in the next 1-2 years.
A critical issue to the use of the MDS for understanding issues of dementia and cognitive function, has been to develop the MDS Cognitive Performance Scale (Morris et al., 1994). Such a scale has at least two purposes: a) it summarizes succinctly the multiple dimensions of cognitive performance included in the MDS and b) its derivation can validate the MDS by comparison with ìgold-standardî measures. The CPS is constructed using five selected MDS items. Two of the items come from the cognitive domain: short-term memory and cognitive skills for decision making. A third item represents the small proportion of residents with coma (or persistent vegetative state). The fourth item is from the communication domainómaking self understoodóand the last item represents an Activity of Daily Living performance indicator: full independence in eating. The CPS classifies all residents into seven categorical levels of cognitive performance, ranging from intact to very sever impairment.
The CPS scale corresponded closely with scores generated by the Mini-Mental State Examination (Folstein et al., 1975) and the Test for Severe Impairment (Albert and Cohen, 1992), directly measured nursing judgments of disorientation, and neurological diagnoses of Alzheimerís Disease and other dementias, all in a split-sample design. It also has been separately validated (Hartmaier et al., 1995). With the CPS, we are able to examine in extremely large populations the effects of cognitive performance, for example on decline in physical function or mortality.
The development and implementation of the RAI in the US provided opportunities for experimentation and adoption elsewhere in the world. Some of these applications are well progressed, while others are in their early stages. For example, the RAI is now the national instrument of Iceland, the provincial instrument for Ontario, Canada, and the city-wide instrument of Copenhagen. At the same time, there is only a single facility experimenting with the RAI in the Czech Republic and a group of a dozen homes recruited in The Netherlands, to date. Other nations involved include Japan, Sweden, Finland, Norway, Spain, Germany, France, Italy, and Switzerland. Overall, the MDS has been translated into 11 languages and, in most cases, a reverse-translation back to English has been accomplished to help insure accuracy of translation.
The differences seen between the nations in nursing homes confirms our earlier findings that cross-national differences in the definition of ìnursing home,î regardless of how the actual words are translated, is perilous (Clauser and Fries, 1992). Nations use their institutions for different types of residents, through differences in admission and discharge policies, and care patterns. We suggested that institutional comparison can be replaced by a more valid comparison of residents.
Cross-national studies represent considerable potential both for the host nation and for more global understanding of long-term institutional care. We find considerable differences among the residents in multiple nations, even after adjusting for ìcase mixî. Such differences provide ìnaturally occurring experimentsî in which nations can learn from each other what works and what doesnít.
The interRAI group was formed to take advantage of this opportunity to share data and to examine nursing homes cross-nationally. interRAI began as a group of researchers, clinicians, and regulators interested in such cross-national comparisons of nursing homes, but has expanded its interest beyond nursing homes to a variety of other care settings fro the elderly and chronically ill persons. In addition to publishing papers about nursing home comparisons (e.g., a special issue of Age and Aging is currently in press), we have developed a assessment system for home care, board and care/assisted living, and acute care. Of particular interest to readers, we have recently been funded by the Province of Ontario to develop an assessment system for psychiatric care, compatible with and with the same philosophy as the RAI.
Albert M., Cohen C. (1992). The Test for Severe Impairment: An instrument for the assessment of patients with severe cognitive dysfunction. Journal of the American Geriatrics Society, 40, 449-453.
Clauser S., Fries B.E. (1992). Resident Assessment and Case-Mix Classification for Nursing Homes: Cross-National Perspectives. Health Care Financing Review, 13, 133-153.
Folstein M.F., Folstein S.E., McHugh P.R. (1975). ìMini-mental stateî: A practical method for grading the cognitive state of patients for the clinician. J Psychiatric Res., 12, 189-98.
Fries B.E. (1997). Changing the Technology of Assessing the Elderly: The Example of the RAI. Generations XXI, 59-61.
Hartmaier S.L., Sloane P.D., Guess H.A., et al. (1995). Validation of the Minimum Data Set Cognitive Performance Scale: Agreement with the Mini-Mental State Examination. Journal of Gerontology: Medical Sciences, 50A, M128-M133.
Hawes C., Morris J.N., Phillips C.D., et al. (1995). Reliability Estimates for the Minimum Data Set for Nursing Home Resident Assessment and Care Screening (MDS), The Gerontologist, 35, 172-78.
Institute of Medicine (1986). Improving the quality of care in nursing homes. Washington D.C.: National Academy Press.
Morris J.N., Hawes C., Fries B.E., et al. (1990). Designing the National Resident Assessment Instrument for Nursing Facilities. The Gerontologist, 30, 293-307.
Morris J.N., Fries B.E., Mehr D.R., et al. (1994). MDS Cognitive
Performance Scale. Journal of Gerontology: Medical Sciences,
49, M174-M182.
To direct comments about the information contained in these pages, please write to marsiske@ufl.edu