| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||
a Duke University Center for Aging and Human Development, Durham, NC
Correspondence: Laurence G. Branch, PhD, Research Professor, Duke University Center for Aging and Human Development, Box 3003, Durham, NC 27710. E-mail: lgbranch{at}geri.duke.edu.
Decision Editor: Vernon L. Greene, PhD
The gerontological social sciences have a hellacious record in distinguishing between need and demand in the context of chronic care services for older people. In this article I review and propose distinctions among the overlapping concepts of need, demand, and use of gerontological chronic care services. I offer a suggested measurement protocol as well.
 |
Distinctions Between Need and Demand |
|---|
 TOP Distinctions Between Need and... Suggested Measurement ProtocolSummaryAppendixReferences |
|---|
 |
The concept of induced demand (operationally defined for this discussion as a process leading to expressed demand in the absence of felt unmet need) allows for an exception to the temporal sequencing in this simple model, which has need preceding demand. Induced demand can originate from sources as diverse as consummate salespersons whose skill would enable them to sell ice cubes to Eskimos, to unscrupulous surgeons who recommend unnecessary surgery to patients who have no basis or resources to contradict a recommendation from their doctor, to well-intentioned social service gatekeepers who see needs in people that the people do not see. The concept of induced demand can change the sequence from simple Model 1 to the following:
 |
Bear in mind that induced demand can be positive in some contexts and negative in others. Examples of positive contexts would include situations in which an individual has a symptomless disease such as high blood pressure, and hence a felt need as a precursor to expressed demand is not possible. In this context, induced demand from a health professional is appropriate. (In this example, it would also be appropriate for the individual to institute other alternatives to felt need as well, such as regular screenings.) Examples of negative induced demand are implied in the previous references to ice cube salesmen, unscrupulous surgeons, and well-intentioned gatekeepers.
Gerontological chronic care adds another layer of complexity to the two models discussed previously. The "no need" state posited as the starting point for the two models actually changes in some gerontological service contexts. Consider the context of the basic activities of daily living (BADL). In our culture at the present time, everyone has a need to bathe, dress, eat, toilet, and so forth on a daily basis. So the paradigm is not
|
|
Rather, the paradigm shifts to
 |
The need is always there for the individual; the issue is whether the need is met or unmet. As before, an induced demand factor can be substituted for the felt unmet need factor, and the induced demand factor can be either positive or negative depending upon the particular circumstances.
Next, consider the examples of instrumental activities of daily living (IADLs). For activities such as grocery shopping, food preparation, housekeeping, managing money, and taking medications, as was the case with the BADLs, it is legitimate to presume that the need always exists for the individual to have these activities performed on an appropriately scheduled basis. However, with the BADLs it can be assumed further that the individual would prefer to perform each of the tasks himself or herself and that therefore the involvement of another person in performing any of the BADLs is prima facie evidence of a limitation or dysfunction in the activity, which can be expressed in the following sequence:
 |
The mere presence or observation of external assistance with a BADL implies Model 3 above, complete with a presumption of a prior state of unmet need. Further, the notion of induced demand seems nonexistent in a culture such as ours that places such an emphasis on independence in each of the BADL areas.
One cannot make the same chain of assumptions with the IADLs, however. With the IADLs, one cannot legitimately assume that each person would rather perform each of the activities himself or herself. Indeed, there are some individuals of my acquaintance (who shall remain nameless) who have endeavored to and successfully avoided the actual performance of most traditional IADLs. They do not shop for groceries, they do not cook, they do not clean the house, and so on. By certain division of labor schemes, they have someone else (usually a spouse) perform all their IADLs. In one study several years ago, Allen, Mor, Raveis, and Houts 1993
reported that 80% of male cancer patients and 30% of female cancer patients attributed help received with IADLs to long-standing divisions of labor. Apparently in these situations the individuals start with the need met, and it never passes through an unmet stage but progresses directly to the need met by others stage. Other individuals indeed start with IADL needs met (either by themselves or by another) then progress to an unmet need stage (presumably because of a change in abilities or a change in availability of the one performing the IADL tasks), which leads to the felt unmet need and then to the expressed demand stage. The implicit models for the IADLs, then, are these:
|
|
 |
For the IADLs, then, the observation or mere presence of external assistance by another individual does not necessarily imply a prior state of unmet need as expressed in Model 4 but could be consistent with Model 5, which does not imply a prior state of unmet need. The presence of assistance from another person in an IADL may just as likely result from cultural expectancies, consistent with Model 5, as from unmet needs, consistent with Model 4. Hence, the chain of assumptions that allows an inference of limitation or dysfunction on the basis of the presence of external assistance that works so well for the BADLs does not apply to the IADLs. Furthermore, the notion of induced demand can be very influential with IADLs in direct distinction to its noninfluence with BADLs. Cultural values present no taboos in accepting assistance in IADLs as they do with BADLs. Accordingly, for the IADLs both Models 4 and 5 can be modified to incorporate an induced demand factor in lieu of the "change 
felt unmet need" sequence of Model 4 or the culturally precipitated "expressed demand" factor of Model 5. An external source can create or induce a demand for assistance with an IADL on the basis of the persuasiveness of the source when no unmet need exists or even in the presence of current culturally predisposed expressed demand (e.g., a person might already be receiving assistance with housekeeping from a spouse on the basis of culturally influenced divisions of labor, then change to an induced demand sequence in response to a well-intentioned service worker who convinces the individual that formal assistance is desirable).
These subtle differences among the need/demand constructs suggest that some of our previous approaches to the assessment of needs for chronic care services have been too simplistic. Collectively we have failed to take into account a substratum of complexity. All too often we infer incapacity or disability from the assistance of another person in performing IADLs. All too often we induce demand in the absence of unmet need. Just "how often" is "all too often"? In one article several years ago, Clarke, Clarke, and Jagger 1992 described their randomized social intervention among socially isolated older people in a medium-sized community in Great Britain and stated that "half the elderly in this sample declined several offers of help" defined as "individual packages of support that aimed at enhanced social contacts" for the socially isolated (p. 1517).
In my own experiences, I recall officials in a state nutritional program lamenting that their biggest obstacle in meeting the unmet nutritional needs of the older people in their catchment area was getting the people to open and eat the meals after they were delivered. In context, the issue was not arthritic fingers interfering with task completion but lack of concordance on the perceived need state between the older person and the case manager—these older individuals thought they were meeting their needs just fine, the case managers thought they had unmet needs, and the case managers neither convinced the older individuals that they really did have unmet needs nor induced demand.
Yet another example of ambiguity between need and demand comes from a careful reading of the evaluations of the various demonstrations of alternatives to institutionalization that have been offered during the last 20 years. When you read the fine print, you find that the Channeling demonstration, the Social/Health Maintenance Organization, the PACE (i.e., the Programs for All-inclusive Care of the Elderly) replication of the On Lok model, respite care programs, VA Adult Day Health Care programs, adult day treatment centers, and just about every other major alternative to institutionalization share a common characteristics—the professionals are enthusiastic about the concept and the implementation, but the intended recipients are not. The intended clients simply will not enroll. In nearly every case, these alternatives are without out-of-pocket cost to those dually eligible for Medicare and Medicaid, but we cannot give away these free (to the recipient) programs. In the previous examples, actual enrollments are often less than half of expected enrollments.
Does this pervasive phenomenon of underenrollment in our "better mousetrap" alternatives suggest fundamental confusion between need and demand and basic ambiguity of the underlying models that drive the process of meeting needs? I suspect it does.
If there is confusion between need and demand, our evaluation of the effectiveness of these alternatives to institutionalization would compare program recipients (who have both need and demand) with functionally comparable nonparticipants (who have need but no demand). Consider the case of the PACE replications. Preliminary information suggests that only approximately 10–35% of those who inquire about the program and who presumptively meet the functional criteria for nursing home eligibility in their state subsequently express their demand for PACE services by requesting admission into the program. In essence, the definitive characteristics of the PACE clients are that they live in the defined geographic area, are dually eligible for Medicare and Medicaid (some non-Medicaid individuals are enrolled on a fee-for-service basis, but these are exceptions), have chronic care needs that make them eligible for their state's nursing home program, and express a demand for services (i.e., request admission into the program). To serve as the counterfactuals in a case-comparison evaluation, the comparison individuals should share the definitive characteristics—including the characteristic of expressing a demand for services. In several of the large-scale evaluations of these innovative alternatives to institutional long-term care, the evaluators have defined the comparison group in terms of comparable need as measured by functional needs but with no requirement for demand and then expressed surprise that the use of long-term care services by members of the comparison group are very low—frequently in the 10–35% range. It is my contention that this kind of scenario suggests a confusion or ambiguity between need and demand that has hampered our ability to estimate the size of the population in need of long-term care services, hindered the development of efficient and effective services for those in need, and compromised evaluations of the innovations.
Another aspect of the assessment process, completely unrelated to the first, may also inhibit the accurate assessment of social service needs among frail elderly persons. This aspect is starting with a service (e.g., a transportation program or a home-delivered meals program) and then looking for appropriate clients, rather than assessing clients' needs first, and then organizing services to meet the clients' needs. The former case is analogous to the witticism that if the only tool you have is a hammer, then everything begins to look like a nail. On occasion our policies and programs constrain assessors to the point where they sometimes acknowledge that a potential client may require general homemaker assistance on a daily basis, but the only service they can authorize at the present time is a personal care attendant twice a week. The frail elderly person with unmet needs that can be met with homemaker assistance is then offered personal care assistance instead. Sad as these scenarios might be, they do occasionally exist.
The linkage between frailty as evidenced by professionally assessed unmet needs and external assistance designed to meet those needs is not as straightforward as it initially seems. The unmet need/external assistance linkage is heavily influenced by cultural values as interpreted by the individuals performing the assessments and recommending the solutions/treatment plans. Culture determines which activities or functions and what levels are necessary for minimally acceptable quality of life standards. The underlying logic is that individuals whose circumstances cause them to fall below the vaguely defined cultural norms of minimally acceptable levels are eligible for support or assistance from the larger community (usually defined as the public sector in the current climate). What dimensions or activities are currently considered the appropriate domain for social service needs assessments, that is, for the critical components of minimal quality of life that the public sector cannot ignore? First, the list varies by state, which typically is the unit responsible for making such value judgments. In simplistic terms, states with larger social service programs have defined the dimensions of minimally acceptable quality of life more generously than states with smaller programs. The size of state programs can be measured by both the breadth of services supported and the depth of coverage. The depth of coverage usually can be gauged by the income-eligibility levels (e.g., 100% of the federal poverty level is more stringent; 150% of the federal poverty level is more generous) and the functional limitations (e.g., limitations in three or more BADLs is more stringent; limitations in two or more BADLs or a severe cognitive limitation is more generous). The breadth of coverage can be gauged by the variety of service options available.
The common types of services offered by public agencies include personal care assistance for the BADLs, homemakers, home health aides, visiting nurses, other allied health professionals such as physical or occupational therapists, meal programs (home delivered and/or at congregate sites), social supports (home visitors and/or social programs at congregate sites), transportation assistance, respite care for caregivers, medical day treatment centers, social day health centers, medical equipment options, care management assistance, form completion assistance, and a variety of other options all designed to enhance an individual's quality of life. However, the cultural norms, the individual's norms, and the assessor's norms concerning the domains and the minimally acceptable levels of performance on each of the domains are not necessarily congruent. Some individuals are very content with minimal social interactions (e.g., hermits), whereas others desire near-constant conversation. Some desire only 1,000 calories per day from two meals (one cold); others value four hot meals per day and 4,000 calories. An assessment process that focuses solely on the behavior level, imputes normative values, and implies unmet needs is fundamentally flawed. Both the behavior and the values of the individual must be ascertained before any inference can be drawn.
|
Suggested Measurement Protocol |
|---|
|
TOPDistinctions Between Need and...Suggested Measurement ProtocolSummaryAppendixReferences |
|---|
in their classic article is still firm. In that article of 30 years ago, they stated that "the present state of the trade (of functional assessment) seems to be one in which each investigator or practitioner feels an inner compulsion to make his own scale and to cry that other existent scales cannot possible fit his own setting" (p. 179). As I have noted in other settings, it appears as if the reason we all spend time reinventing the wheel is that we hope to get a patent on it.
Nevertheless, although the foundation originally offered by Lawton and Brody 1969, and later advanced by the Duke Older Americans Resources and Services (OARS) assessment approach (Duke University Medical Center 1978) is still firm (and all 15 of the domains I offer in the following paragraphs are rooted in the Lawton/Brody and OARS traditions), I would suggest one significant departure. Both the Lawton/Brody approach and the OARS approach allowed the trained observer to render the final judgment. I would agree that a professional judgment is required in a clinical setting. However, population-based needs assessments for purposes of estimating the size of the population at risk may be better obtained with structured questionnaires and professionally generated algorithms for preliminary classification into major categories such as (a) need met and no apparent problems, (b) need met but potential problems, (c) uncertain, and (d) need unmet and current problem. The logic of this latter approach has been put forth elsewhere (Branch 1987).
The generic form of the structured needs assessment for the 15 IADLs might be the following:
The appendix presents actual items for the 15 domains in this approach.
The interpretation of the responses to this approach to IADL needs assessment is not difficult. Table 1 presents the various combinations of response patterns for each domain.
|
If the responder's pattern is that another person helps, the activity is done satisfactorily, and the individual could do it himself or herself (i.e., Profile 4), then the presumptive categorization would also be "need met and no apparent problem." However, if the response pattern is that help is received, the task is done satisfactorily, but the individual could not do the task independently (Profile 5), then there might be a potential problem because of the uncertain stability of the caregiver that may merit further discussion in a clinical context. But for population needs assessments, I would ascribe a presumptive category of "need met but potential problem."
If the response pattern is that another person helps but the task is not done satisfactorily, and the individual reports he or she could not do the task himself or herself (Profile 8), a clinician would want more information to verify that the individual did not have unrealistic expectations. But for purposes of population needs assessments, that individual would be presumptively categorized a "need unmet and current problem." It is worth emphasizing here that the mere presence of a caregiver does not automatically imply a met need.
If the individual reports he or she does the task completely by himself or herself but is not satisfied (Profile 2), the clinician would want more information on potential assistance from others, but the presumptive categorization would be "need unmet and current problem."
Response Profile 9 also presents an interesting pattern. The respondent reports that another does the task, it is not done to the target person's satisfaction, but the target person does not know if he or she could do the task independently. The presumptive categorization is uncertain because the underlying problem, of course, is that if one does not in fact do or try to do an activity, one cannot be confident about whether one can indeed perform it satisfactorily. This ambiguity is expressed by the target person as a "don't know" to the third item and by the assessment process as an "uncertain."
|
Summary |
|---|
|
TOPDistinctions Between Need and...Suggested Measurement ProtocolSummaryAppendixReferences |
|---|
change." In addition, underlying values—the individual's, the assessor's, the society's —all play a part in the judgment process. The assumption that the involvement of another in performing the task automatically implies a limitation on the part of the recipient is not warranted, nor is the assumption that the reliance on another by one in need automatically implies the need is met. For purposes of population needs assessments, it is possible to rely on presumptive categorizations as a function of simple response profiles. For individual care planning, however, additional information is essential.
Received for publication August 31, 1999. Accepted for publication March 20, 2000.
|
Appendix |
|---|
|
TOPDistinctions Between Need and...Suggested Measurement ProtocolSummaryAppendixReferences |
|---|
(c) DON'T KNOW
|
References |
|---|
|
TOPDistinctions Between Need and...Suggested Measurement ProtocolSummaryAppendixReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. E. Gaugler, K. A. Anderson, C. R. Leach, C. D. Smith, F. A. Schmitt, and M. Mendiondo The emotional ramifications of unmet need in dementia caregiving American Journal of Alzheimer's Disease and Other Dementias, November 1, 2004; 19(6): 369 - 380. [Abstract] [PDF]
|
|
|
|
 |
|
 J. R. Sharkey Risk and Presence of Food Insufficiency Are Associated with Low Nutrient Intakes and Multimorbidity among Homebound Older Women Who Receive Home-Delivered Meals J. Nutr., November 1, 2003; 133(11): 3485 - 3491. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Shea, A. Davey, E. E. Femia, S. H. Zarit, G. Sundstrom, S. Berg, and M. A. Smyer Exploring Assistance in Sweden and the United States Gerontologist, October 1, 2003; 43(5): 712 - 721. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
