What's Wrong With ADLs?

he most common diseases afflicting older persons are T those that are chronic and incurable. Hence, measures other than those of disease cure must be employed to monitor a patient’s health status, both in the face of progressive disease and in response to treatment. Although traditional laboratory and radiological measures of disease burden are important, these measures may be less helpful because of age-related physiological changes and the frequent occurrence of multiple interacting diseases. Since the 1960s, functional status has been used as a research measure and, more recently, as a clinical measure, of the health of older persons and serves as a final common pathway in the transition from specific diseases to disability.’.2 The seminal early works on defining functional tasks essential for self-care (the activities of daily living or ADLs)~ and for maintaining an independent household (the instrumental activities of daily living or IADLS)~?’ provided a sense of clarity and simplicity that led to widespread acceptance and use of these measures. When presenting a case at our institution, the geriatrics fellows mention the “Katz level” with the same degree of confidence that they accord to reporting blood pressure. Medical students and residents are also becoming conversant with these concepts and the timehonored categorizations. In fact, the recognition of the importance of functional status may be regarded as a major success in the integration of “geriatrics thinking” into the health care of older persons. So, what’s wrong with the current measurement of functional status? Perhaps the very simplicity of the metric, which has been its great virtue, has also been its greatest weakness. First, the definition of disability has been problematic. Most of the scales have created response categories that indicate the degree of assistance required, which, albeit important, frame disability in economic terms (i.e., resource demand) rather than personal estimates of quality of life. More recently developed scales (e.g., the Functional Status Questionnaire6 and the Short-Form-36 10-item physical functioning scale’~*) have phrased responses in terms of degree of difficulty and amount of limitation. One can argue the pros and cons of each approach, but clearly none is comprehensive. Moreover, differences in how functional status questions and responses are phrased are not trivial. Wiener et al., have noted fairly substantial differences in disability rates based on the way the concepts are operationalized.’ For example, the National Medical Expenditure Survey estimates that there are 60 percent more older persons with ADL problems than does the National Health Interview Survey Supplement on Aging. Second, the quantification of functional status has been troublesome. Concern has been raised about traditional methods of quantifying function at the ADL or IADL levels because of their insensitivity to change.” The perception among clinicians and researchers is that patients may experience considerable decline in functional capacity before the transition from “independent” to “needing assistance” actually occurs; this change is not captured by conventional instruments. These instruments have also been cumbersome from an analytical standpoint because, although functional tasks are hierar~hical,”-’~ precise mathematical relationships between tasks have not been demonstrated. Even the hierarchical relationship originally established by Katz et al. appears to be only one of many that meet the assumptions of Guttman s~a1ing.I~ From a clinical perspective, traditional functional status instruments measure only needs rather than the matching of needs to resources. When caring for an individual patient, the degree of mismatch, rather than need per se, often guides clinical decisions.” For example, an older man’s dependency in meal preparation is of no clinical significance if his wife has always prepared meals and is perfectly capable of continuing to do so. A recently developed instrument begins to address the match of resources to needs16 and is promising as a clinical and research measure. In this issue of the Journal, Finch et al., have attempted to address one limitation of traditional measures of functional status by using magnitude estimation to combine ADL and IADL items into a single ratio scale.” The investigators relied on opinions of experts, mostly clinicians who provide care for older persons, to create weights for traditional functional status items. Not surprisingly, after weighing, the distances between functional tasks were not equal. For example, the weights assigned to needing complete assistance in bathing and dressing were almost identical (i.e., very little distance between the tasks), whereas the weights assigned to needing complete assistance in toileting and feeding differed by 123 units (i.e., considerable distance between the tasks). In creating these weights the authors were able to determine the importance of dependence in specific functional tasks relative to one another in a ratio rather than in an ordinal relationship, which will permit different analytical methods to be employed when examining functional status data. The research represents an important step as the field continues to refine its conceptualization and measurement of function. Nevertheless, some cautions must be recognized when using this new metric. First, it reflects the weighing by health professionals rather than patients themselves. Other approaches to weighing patient preferences using time-trade offs are in development” and may assign different values to functional independence. Second, though weighted, this metric still treats dysfunction in specific tasks as additive rather than synergistic. For example, needing “a lot of assistance” in two valued tasks would likely confer a weight higher than the strict sum. Conversely, the sum of needing “a little assistance” in performing several tasks may be numerically equivalent to complete dependency in one task but may be far less