Assessment in Geriatrics

The position statement of the American Geriatrics Society on Comprehensive Geriatric Assessment (CGA) was published in the May issue of the Journal,1 and a relevant report by Rubenstein et al2 from a Task Force on Health Assessment of the Society of General Internal Medicine (SGIM) is in this issue. We concur in the four positions taken by the AGS, although we believe it is important to stress two caveats, as previously emphasized in the NIH Consensus Development Conference statement3 and in associated commentaries.4, 5 The first is that, thus far, CGA has been found effective only in certain subgroups of geriatric patients, those neither “too well” nor too irreversibly disabled; thus, there is general agreement that targeting of CGA to the patients most likely to benefit is critical to successful assessment programs. The second caveat is that CGA is primarily a diagnostic, rather than a therapeutic, process; therefore, for optimal effectiveness, the assessment must be inextricably linked to the ongoing care of the patient, with the objective of assuring that the comprehensive care plan is implemented and modified as events may dictate. We would like to seize on this time of great interest in CGA to make some suggestions regarding nomenclature. As the literature of clinical geriatrics has proliferated, inconsistent terminology has hindered communication among those interested in the care of elderly persons. The most confusing example has been the use of the term, functional assessment (FA), as synonymous with CGA by some authors, while others view it in a more restrictive vein. We wish to propose the adoption of a common language to describe the assessment of older persons. Although preservation or restoration of function has been a principal goal of all geriatric assessment, we would suggest, for clarity, that the term functional assessment be reserved for the measurement of a patient's ability to complete functional tasks and fulfill social roles. Functional tasks can range from the simplest self-care to executive-level occupational responsibilities. The first step in clarifying the description of these tasks would be to classify them into three levels of function, stratified according to difficulty and complexity into basic, intermediate and advanced activities of daily living (BADL, IADL, and AADL, respectively). BADLs include the elemental functions estimated by the Katz Index or similar scales of self-care items.6, 7 IADLs have been measured as the Instrumental Activities of Daily Living of Lawton and in the OARS instrument; they are symbolic of similar tasks essential to maintaining independence.8, 9 Often, independence or dependence in these IADLs determines whether an elderly person can continue to live alone. AADLs may be thought of as “luxury” items of function, well beyond what is needed to maintan independent living. They tend to be volitional, specific to the individual and influenced by cultural and motivational forces. They usually relate to recreational, occupational, altruistic, or community service functions. Although not essential for independence, the AADLs may become an important area for future study, since one's capacity to continue AADLs may be conducive to strong mental health and certainly contribute to maintaining excellent quality of life. Furthermore, one might hypothesize that decline in the ability to perform AADLs might be a valuable early predictor of more serious functional decline. Some illustrative patterns of the effects of age-related disease on function in this three-level model are diagrammed in Figure 1. . Three examples of aging and functional status affected by illness. A is an example of reasonably healthy aging manifested by minor changes in AADL status with acute minor illnesses (eg, upper respiratory infection) and personal stresses followed by a gradual progressive decline due to age-related conditions. B is an example of a progressive illness, such as Alzheimer's disease leading to relatively rapid functional decline. C is an example of the functional effect of a catastrophic illness, such as stroke, with improvement following rehabilitation. Social function includes social activities, relationships, community and religious services, and employment. At the highest levels of ADLs and social function, the distinction between these two spheres often disappears. For example, a tennis match represents both a complex AADL and a social activity. Function may be assessed by self-report (either by the patient or a proxy), interview (either by the patient or a proxy), or by direct observation. Each of these methods has its advantages and drawbacks, discussed most thoroughly by Kane and Kane10 and commented on further by Rubenstein, et al.2 Their value is dependent on the reliability of the source of information. We believe that it is premature to convey “gold standard” status to any of these methods until more extensive validation studies have been conducted. Our definition of functional assessment excludes many domains (medical, cognitive, affective, social sup-port/caregivers, economic, environmental, and quality of life/well-being) that surely are part of a comprehensive geriatric assessment. These domains affect the person's functional status but are not measured in functional terms as defined above. Thus, measurement of a domain may not predict the effect of a deficit in that domain on functional status. For example, relatively minor cognitive impairment in a corporate executive may limit severely his or her ability to fulfill an occupational role, whereas the same degree of cognitive impairment may have no discernible effect on the job performance of a manual laborer. Furthermore, impairment in one domain may have an impact on other domains, in addition to or as a result of its effect on function. Some of the complex relationships between domains and function are depicted in Figure 2, but the key generalization is that domains and function are interrelated but by no means synonymous. . Diagram illustrating the domains of comprehensive geriatric assessment and some of their interactions. Note that we assess many domains that affect functional status but are not themselves part of the functional domain. Rubenstein et al have made the same point when they say,”… a test that diagnoses depression or impaired cognition is a measure of disease, not function. A measure of mental functioning would include behaviors in daily life that are linked to mental ability, such as the ability to make change at the grocery.”2 Comprehensive geriatric assessment begins with a thorough medical evaluation and includes appraisal of functional status and the domains described above. Thus, functional assessment is only one component of CGA, albeit a very important one. To name the entire process “comprehensive functional assessment” would seriously undervalue the contributions of medical diagnosis and evaluations of other domains. While the comprehensive care plan that crowns the CGA has the overall aim of improving function, many of the specific recommendations are directed at appropriate medical therapy for specific clinical diagnoses, properly adjusted according to geriatric considerations. Similar arguments can be made for the importance of interventions in other domains in preserving or restoring function. Thus, we recommend CGA rather than CFA as the most accurate descriptor of the total process. THE SGIM Task Force2 uses the term health status assessment more or less synonymously with our use of CGA; both terms describe a process that includes the estimation of functional status at its core, but also assesses other domains of physical, mental and social health. Parenthetically, we have it by personal communication that the Task Force has no objection to the term CGA, but could not accurately use that title for its paper because the Task Force was suggesting guidelines for the busy practicing internist who does not have available the time or personnel to perform CGA, an extensive process as defined in the NIH Consensus statement.3 Accordingly, the Task Force's recommendations are for selected screening tests to be carried out for certain classes of older patients. Hence, the use of the less explicit term, health status assessment. In the interests of clarity, we must mention that this term will itself be viewed by some as ambiguous, because scales measuring life satisfaction, quality of life, or well-being have often been grouped under the rubric of health status estimates or global health ratings. Hence, we return to CGA as the descriptor of choice; perhaps the adjective, comprehensive, could be dropped to describe less wide-ranging assessment routines as simply, geriatric assessment.11 Each domain that falls within the rubric of comprehensive geriatric assessment should be characterized specifically, and the measurement instruments should be clearly stated. In addition to a medical problem list, the elements most often measured by quantitative scales are cognition, affect, gait, balance, BADL, and IADL. Some instruments (eg, OARS and MAI) assess many domains.10, 12 In every instance, when reporting study data, it is important to be explicit about the domains assessed and the instruments used. In another facet of our plea for consistency, it would be very helpful if all of the instruments commonly used in a CGA were scored in the same direction. Thus, the scores on the Katz,6 Lawton,8 Folstein,13 and the two Tinetti scales14 are all reported in such a way that the higher the score, the better is the health status of the subject. The one commonly used outlier is the Geriatric Depression Scale (GDS) of Brink and Yesavage, in which the higher the score, the more depressed is the subject.15 This scoring pattern has been traditional for scales measuring depression, so it would be impossible to reverse the direction of scoring for those scales used for subjects of all ages. However, it would be feasible to reverse the direction of scoring of the GDS, and we would recommend that this change be considered. Further, creators of new instruments should design them so that higher scores reflect better performance. We hope researchers and clinicians in geriatrics will agree on a common language for the assessment of older persons. In an era of rapidly expanding knowledge in our discipline, we cannot afford to be impeded by imprecise communication or the inability to agree on an acceptable terminology.