Comprehensive Geriatric Assessment in Long-Term Care and Nursing Homes
5.3 The Resident Assessment Instruments (RAI)
The development of the resident assessment instrument (RAI) minimum data set (MDS) [4] in 1987 and its introduction in 1991 were prompted by LTC reforms endorsed by the United States (US) government, requiring that all LTC residents
F. Panza et al.
49
undergo a CGA on a regular basis, on admission to a facility. A network of research- ers and clinicians committed to (the interRAI network) was formed to promote and guide the use of the RAI-MDS instrument for comprehensive assessment of the older subjects. InterRAI instruments include a clinical data set, a training manual, and algorithms that generate clinical assessment protocols (CAPs), scales (includ- ing screeners and severity measures), case-mix measures, and quality indicators (QIs). In 1995, a revised version of the RAI-MDS, the RAI-MDS 2.0, was devel- oped, resulting in over 400 data elements, with improved reliability [5]. In 2005, the multinational consortium interRAI released the interRAI suite of instruments tai- lored to a specific healthcare setting (interRAI Home Care, interRAI Acute Care, interRAI Long-term Care, interRAI Palliative Care, etc.) [6]. A more recent version of the LTC assessment instrument, the interRAI Long-Term Care Facility (LTCF), and an adaption of the RAI-MDS 2.0, the MDS 3.0, have been released. At present, the interRAI LTCF instrument has not been widely implemented, and the MDS 3.0 has been implemented in the USA only.
Data collected from residents in LTC is aggregated to produce indicators of the quality of care provided. One study examined 38 chronic care QIs, of which strong evidence for the validity of 12 QIs was found [7]. A systematic review on observa- tional studies conducted in “real-world” conditions tested the validity and/or reli- ability of individual QIs (falls, depression, depression without treatment, urinary incontinence, urinary tract infections, weight loss, bedfast, restraint, pressure ulcer, and pain) with mixed results. Indeed, this systematic review revealed the potential for systematic bias in reporting, with underreporting of some QIs (pain, falls, and depression) and overreporting of others (urinary tract infections) [8]. In 30 urban Canadian nursing homes with a total of 94 care units, an observational study showed the necessity of facility-level and unit-level measurement when calculating QIs derived from RAI-MDS 2.0 data for pressure ulcer, antipsychotic with no diagnosis of psychosis, and pain [9]. Furthermore, RAI-MDS can be a valuable tool in target- ing residents for a transition program from LTC to community. Secondary data from RAI-MDS assessments for an annual cohort of first-time admissions to nursing homes suggested that at 90 days the majority of residents showed a preference or support for community discharge, and many had health and functional conditions predictive of community discharge or low-care requirements [10] (Table 5.1).
However, a validation study of the RAI-MDS conducted in four states in the USA suggested that the accuracy for identifying hospitalization events and payment sources in LTC of this CGA-based tool varied across the study states and should be evaluated carefully with regard to the intended uses of the data [11] (Table 5.1). In a longitudinal cohort study on newly admitted Icelandic nursing home residents among RAI-MDS 2.0 variables and scales, significant predictors of mortality were age, gender, place admitted from, functional status, health stability, and social engagement [12] (Table 5.1).
Recently, given the lack of findings on a psychometrically evaluated CGA for nursing home residents with palliative care needs, a protocol based on the Medical Research Council framework has been implemented to examine the effect of using the interRAI Palliative Care on the quality of palliative care in nursing homes [13].
Interestingly, given that oral health in nursing home residents is poorly addressed,
5 Comprehensive Geriatric Assessment in Long-Term Care and Nursing Homes
50
Table 5.1Principal observational studies and systematic reviews on comprehensive geriatric assessment (CGA) in long-term care, i.e., rehabilitation units and nursing homes Author, year, referenceSettingType of study Number of participants/ trials with general characteristicsRole of the CGA interventionComments Arling et al. (2010) [10]Post long-term care community dischargeObservationalMDS analysis file of 24,648 first-time nursing home admissions
Major MDS variables were discharge status, resident’s preference and support for community discharge, gender, age, and marital status, pay source. Major diagnoses, cognitive impairment or dementia, activities of daily living, and continence
At 90 days, 64% of residents showed a preference or support for community discharge, with health and functional conditions predictive of community discharge (40%) or low-care requirements (20%). A community discharge intervention could be targeted to residents at 90 days after nursing home admission when short-stay residents are at risk of becoming long-stay residents Cai et al. (2011) [11]Hospitalizations in long-term careObservationalThe 2003 MDS, MedPAR, Medicare denominator file, MAX long-term care file, and MAX personal summary file for four states in the USA
The accuracy of the MDS in identifying hospitalizations and payment sources varied across the study states and should be evaluated carefully with regard to the intended uses of the data The MDS alone did not seem to be an ideal source for identifying payer source or hospitalization events in nursing home residents
F. Panza et al.
51
Hjaltadóttir et al. (2011) [12]Nursing homesObservational2206 residents admitted to nursing homes in Iceland in 1996–2006
Age, gender, place admitted from, ADL functioning, health stability, and ability to engage in social activities were significant predictors of mortality
More than 50% died within 3 years, and almost a third of the residents may have needed palliative care within a year of admission Hermans et al. 2014 [27]Nursing homesSystematic reviewSeven studies includedThe interRAI PC covered all domains for a palliative approach in residential aged care, while the McMaster quality of life scale covered nine domains
The interRAI PC and the McMaster quality of life scale were the most comprehensive CGAs to evaluate the needs and preferences of nursing home residents receiving palliative care Abrahamsen et al. (2016) [28]Nursing homes after hospitalizationObservational961 community- dwelling patients aged ≥70 years, considered to have a rehabilitation potential and no major cognitive impairment or delirium, transferred from acute hospital departments
Slow or poor recovery was significantly associated with low scores on the Barthel index and orthopedic admission diagnosis
Different caring pathways for different patient groups in intermediate care unit in nursing homes should be considered MDS minimum data set, MedPAR Medicare provider analysis and review file, MAX Medicaid analytical extract, interRAI PC inter resident assessment instru- ment palliative care 5 Comprehensive Geriatric Assessment in Long-Term Care and Nursing Homes
52
an assessment tool such as the RAI-MDS 2.0 appeared to be useful to monitoring and improving quality of oral healthcare. However, using data on 13,118 residents collected in a stratified random sample of 30 urban nursing homes in Western Canada, RAI-MDS 2.0 oral/dental items likely underdetected oral/dental problems and were not associated with well-proven predictors for oral health, indicating poor validity [14]. At present, the potential effect on this issue of the interRAI LTCF with its modified oral/dental items and more frequent collection is unknown.