Maximizing the Minimum Data Set

In this issue of the Journal, Hartmaier and her colleagues have published what we hope will be the first in a series of articles that examine the validity of the Minimum Data Set (MDS).1The MDS was developed and initially tested by gerontologists and geriatricians under contract with the Health Care Financing Administration.2Its use is now mandated by federal regulations (originally contained in the Omnibus Reconciliation Act (OBRA), 1987) that require that all residents of Medicare- or Medicaid-certified nursing facilities undergo comprehensive, multidisciplinary assessments - including the items on the MDS.3The MDS covers 16 specific areas of assessment and must be completed within 14 days of admission, at the time of a major change in status, and annually. Eight of the 16 assessment areas are updated quarterly. Like all of the other OBRA, 1987 rules and regulations, the MDS seems to be viewed by nursing home providers as yet another example of unnecessary paperwork and governmental intrusion into nursing home care. My own casual and admittedly anecdotal observations of the implementation of the MDS (including in our own facility) reinforce my sense that there is still a great deal of misunderstanding about the rationale for the mandated assessment and its potential to improve care at individual facilities and nationally. For individual nursing home residents and the interdisciplinary staff caring for them, the MDS is intended to be the cornerstone of the assessment process. Individual items and combinations of items can “trigger” further assessments using one or more of the 18 Resident Assessment Protocols (RAPs).4The RAPs cover a broad array of conditions that are common in the nursing home population. Results of these further assessments are then incorporated into the resident's care plan. Thus, the MDS is more than just another form that requires completion and placement in the medical record. It can serve as the primary data source for documenting a resident's status, identifying important geriatric conditions that may require further assessment, generating management strategies for the resident's care plan, and determining the resident's progress over time. At the regional and national level, the MDS will be useful in at least two ways. First, it can assist in examining variations in selected care practices and quality indicators, such as the use of indwelling catheters and enteral feeding tubes, and changes in functional status outcomes over time. Targeted interventions to improve care might then be developed based on careful analyses of these data. Second, the MDS will be used as the basis for a prospective reimbursement system along the lines of the Resource Utilization Groups (RUGs).5Such a system could be based on the characteristics of the residents and the special services provided (as is the case in several states currently) or on the actual resident outcomes achieved in relation to expected outcomes.6A version of the RUGs that incorporates items from the MDS for reimbursement purposes is currently being tested in several states. Many questions need to be answered about the MDS. As is the case for a wide variety of geriatric assessment tools, an understanding of the reliability and validity of the MDS is critical to its optimal utilization. While some inter-rater and test-retest reliability testing has been carried out, more is needed. In this context it is important to note that in Hartmaier's study, MDS data were collected by a trained geriatric nurse who followed recommended guidelines. One wonders whether the same results would be achieved by more typical nursing home staff. Inter-rater reliability is obviously important when MDS data are compared among facilities, and in evaluating changes over time within a facility in which multiple staff are involved in completing the MDS assessments. Potential difficulties with definitions, training of staff, and inter-rater reliability have caused some to voice concern about the use of the MDS for research purposes.7,8 Test-retest reliability, as well as the sensitivity of selected MDS items for detecting meaningful change in response to interventions, are critical properties that require further study if the MDS is to be useful in measuring change over time and the effects of various intervention strategies. Our research group has, for example, recently found that MDS recordings about incontinence are valid in determining whether a resident is continent or incontinent (using independently performed physical checks for wetness as the gold standard), but that the MDS data are not accurate in identifying even marked changes in the frequency of incontinence in response to prompted voiding (unpublished data). Finally, testing the validity of MDS items compared with other well validated assessment strategies (as the Hartmaier study does) is essential in ensuring the MDS is measuring what it is intended to measure. In addition to determining its reliability, validity, and sensitivity to change, even more important questions remain about the MDS. Does the use of the MDS result in increased identification of conditions that require intervention? Are the RAPs triggered by the MDS being carried out? Most importantly, are interventions targeted at conditions identified by the MDS and assessed using the RAPs resulting in improved quality of care and better outcomes and quality of life for nursing home residents? While it is hoped that these questions will be answered over time, geriatricians and other professionals who work in nursing homes should recognize the potential of the MDS to improve the process and outcomes of care. Important challenges lie ahead in order to document this potential. TheJournalwould be very interested in considering more reports of studies that will help meet these challenges.