Life-Span Design of Residential Environments for an Aging Population

Prosthesis Ware: Personal Computer Support for Independent Living

Douglas L. Chute, Ph.D.
Professor of Neuropsychology
Drexel University

Margaret E. Bliss
CEO
Mac Laboratory Inc.

Defining the role of the microcomputer in support of activities for older people living in the home environment entails rethinking some current practices and expectations. The personal computer is not yet a competent housekeeper, nurse, or valet, nor is it a "cure" for cognitive impairment or an effective means of ameliorative retraining. What the microcomputer can do, in the foreseeable future, is serve as a prosthetic aid for some problematic, everyday living tasks. Like any prosthesis, however, it must be customized to fit the individual; it requires maintenance; and it necessitates training for users, caregivers, and health care professionals. The fundamental problem for its implementation is not the availability or appropriateness of high-technology solutions, but the requisite service industries to support such solutions in the desired home environment.

There are two principal reasons why the following discussion emphasizes applications of personal computer technology for the neurologically impaired. First, it is our area of expertise for which we have received research and development funding, although funding is generally not available to support work with the well elderly population. The second reason can perhaps be best illustrated by the following case summary: We supplied a Macintosh computer to a couple in their eighties living in a suburban home outside Toronto. Both individuals were effective and well in all regards; both were highly educated and motivated to learn "something new." The intent was to have them use the machine to manage personal affairs and otherwise take advantage of the technology, primarily through the use of commercially available software. For example, the woman corresponded with a number of people, and the computer's word processor offered some advantage to her over writing by hand or using the typewriter.

Over a two-year period, however, the computer was seldom used, except for the initial exploration of an easy-to-learn and engaging novelty and for the grandchildren playing games! Among the subsidiary issues working against its use was the inconvenience of double data entry: financial statements arrived on paper and needed reentry in the computer. There was an ergonomic issue as well, since the tilt of the head that was necessitated by bifocals proved uncomfortable for any length of time. Thus, the reason it did not become an integrated part of their lives was, quite simply, that these people were perfectly comfortable with the way they already managed, and the machine did not offer them any significant productivity or other perceived benefits.

For the neurologically impaired individual, on the other hand, the computer may be the only way to accomplish some desired task, and thus it is used out of necessity. For example, the Hanspeter Albisser Foundation of Zurich is supporting a joint effort to continue development of a speech prosthesis for patients with amyotrophic lateral sclerosis (Lou Gehrig's disease). These people eventually lose motor control of speech, and the Macintosh permits them to communicate. Figure I shows a card from the SpeechWare program developed as a ProsthesisWare module for some stroke, head injury, or other persons with an expressive aphasia. Simply clicking a button will have the computer "speak" any sentence or phrase that has been entered. Other buttons connect to different modules, providing "speaking" reminders or access to other software. SpeechWare can be customized for a person's particular needs and requires no more advanced computing skills than are required by the average word processor or painting software.

For a change, the needs and wishes of elderly people and the developments in health care systems are following parallel evolutionary tracks. Older persons need to be "in control" of their environment and activities of daily living (ADLs) for their sense of well-being and continued intellectual stimulation. The health care delivery system cannot afford for them not to be. As a tool, the microcomputer can perform prosthetic functions individualized to the ecological requirements of the impaired elderly person. Such skills as shopping, banking, dressing, cooking, personal hygiene, locomotion, home management, and a host of other ADLs are second nature to most individuals but can be severely disrupted in the physically or cognitively impaired elderly person. Many devices and applications, such as "smart houses" (model homes created by a number of industries and controlled by microcomputer technology to manage energy, personal convenience, and related ADLs), offer benefits to the frail or physically handicapped, but few are designed to accommodate the cognitive difficulties sometimes seen in the aged population. Borrowing from rehabilitation medicine and adaptive living technologies, however, has provided a good beginning.

Which leads us to ProsthesisWare. Although usually couched in "rehabilitative" terminology because of its historical beginnings with neurotrauma patients, the ProsthesisWare concept is just as easily viewed as "habilitative" or preventative. Its objective should be to enable the user to manage the most suitable environment. Where ProsthesisWare has its greatest potential is in problems arising from cognitive and central nervous system limitations, an area of concern that lags in research, model systems, clinical trials, funding, recognition, and conceptual development. A ProsthesisWare module does not offer general training or presuppose that some rebuilding of the central nervous system is needed for rehabilitation; rather, Prosthesis Ware is specifically customized to offer aid appropriate to the individual environment and context for any particular ADL. Like a prosthetic limb, a module may not resemble the original, but it can be very functional nonetheless.

Current Technologies

Basically, the most difficult and expensive aspect of employing the computer as a prosthetic aid is customizing it for each specific patient and each specific ADL. Customization entails the need to be rapidly adaptable (in programming time) and to be readily editable for multiple iterations of the customization process. The former is required for cost effectiveness, the latter for clinical effectiveness. A consistent graphic interface is also a boon since less human memory or cognitive "overhead" is required for operation. In general, however, less effort has been invested in the cognitive requirements for using computing machinery than in the physical requirements. Most popular personal computers have a variety of input devices for the physically and sensory handicapped, but all are weak on output devices that permit interaction with the environment Thus far, one type of microcomputer has been preferred in experimental studies of the prosthetic use of this technology. That machine is the Macintosh from Apple Computer Inc. An easily operated machine like the Macintosh can serve as an external ADL prosthesis, enabling individuals to live at home or in minimally structured environments. Flexible control of fairly easily mastered programming tools like HyperCard, ColorCard, and Controller allow the health care professional to adapt modules to individual needs readily, without the burden of reinventing core software routines. At the risk of oversimplifying too much, the computer industry refers to the ability to manage and manipulate core modules and their properties as "object-oriented programming." In our opinion, development of object-oriented programming alone will be the necessary and sufficient condition by which clinically cost-effective micro computing is brought to the elderly population. The testing of the ProsthesisWare concept has been made possible research context by an easy-to-operate, graphically based, high-resolution machine with a large capacity the Macintosh and by a readily customized, object-oriented authoring package HyperCard.

Actual clinical implementations do not require that either hardware or software technologies advance beyond their current state, although they surely will. A good deal of prosthetic programming useful to the cognitively impaired elderly person is possible now. What is required for implementation is some development of the infrastructure to provide for training, neuropsychological monitoring, professional development, software exchange, hardware engineering, interface design, and other support issues. Business, government, and information services are moving toward machine-readable data, thus obviating the double-entry problem for the individual user, but networking and access are not yet sufficiently clear, cost-effective, or available for widespread use.

Costwise, even a comparatively expensive machine like the Macintosh can be cost-effective today if more expensive health care or custodial services can be avoided or delayed. In any case, while hardware is actually not as expensive as the true cost of software, both should increase in power but not decrease much more in cost in the near future. Object-oriented programming, however, has the financial benefit of greatly reducing software costs and the clinical benefit of putting software design into the hands of the health care professional. As an individual's needs change, program modules for that individual can be quickly altered or augmented to reflect those needs. A national software exchange could provide prototype modules that can be readily adapted for individual use.

Neuropsychology and Software Design

Microcomputers in rehabilitation will move away from the role of tutor and the ineffective strategy of providing drill and practice in general cognitive skills like attention or memory. Data in support of training in such general cognitive factors are scarce. Even though they are widely employed, there is little evidence that repetition, rehearsal, internal mnemonic strategies, or other forms of cognitive training will, in fact, generalize to functional ADLs.

Various memory and cognitive strategies that can be used in normally functioning subjects are often prone to failure for ADL maintenance because they are not "ecologically based" in a way that is relevant to the individual's functional environment. Strategies that may be trained in normally functioning subjects are difficult for the impaired elderly person to acquire, and they may not be spontaneously used by that person even if they can be acquired. Such strategies often do not recognize individual difficulties in self-generation of reasonable associations, rhymes, images, or acronyms. Even when the rehabilitation effort is clearly based on "training to task" for ecologically valid functions, the risk is that the practiced ADL skills will not transfer to the environment of the user.

Both computer and human caregivers may seem to represent an external entity of power and authority, a loss of control over an individual's day-to-day environment. It doesn't matter that the authority exercised is benign; it still detracts from the perception of independence. The problems for computers in this regard are not much different from those for humans, and both need to be addressed in implementation. With an appropriate interface, however, the microcomputer as a cognitive prosthesis can be viewed by elderly people in a non-threatening way, like an answering machine or the computer technology in the modern car engine. The machine becomes a supplement to abilities.

Consider home monitoring for a moment. In this situation, the computer can offer a reasonable and adaptable level of safety and security, and can afford respite opportunities for caregivers. Bar code technology can monitor diet, medication, and other health care factors that relate to maintenance of independent living. ProsthesisWare software already can supply monitoring, cues, reminders, schedules, and sequences that maintain a stable, functional, and ecologically appropriate environment.

Concept software like Prosthesis Ware provides a whole battery of active strategies to complement the goal of limited dependency in an independent living environment. Although design effectiveness and utility are difficult to measure, to maximize the adaptability of object-oriented programs these qualities need to be evaluated and modified through an iterative customization process.

Finding new neuropsychological assessment and treatment strategies is necessary to optimize costs and provide a theoretical framework for interventions. The need to tailor software to each individual environment will deemphasize traditional standardized assessments, replacing them with more ecological and valid multiple sequential monitoring of specific activities of everyday life. Many neuropsychologists recognize that the emphasis of neuropsychological assessment in rehabilitation is moving away from diagnosis, as that function being replaced by continuously improving the neurodiological techniques. Neuropsychological assessment continues, however, in the valued role of making the behavioral and cognitive prognostication that is used in rehabilitation, discharge, and vocational planning, and in the evaluation of ADLs.

For the proper implementation of Prosthesis Ware, new types and styles of neuropsychological assessments need to be added to provide software design and interactive feedback. Since software can now be created and adapted for each unique individual, case study and time-series assessments must be used to monitor interventions, document changes over time, and predict the outcome of various pathological or recovery processes. Fortunately, considerable effort has been expended by academe to refine traditional assessment devices and better accommodate the changing emphasis from diagnosis to prognosis.

Implementation Issues

Listed below are the currently preeminent questions surrounding implementation of ProsthesisWare:

Who will do the programming?
What are the medical/legal issues?
How much will it cost?

The health care professionals will do the programming. The rehabilitation specialist in the past typically has had neither the time, the reimbursement, nor the specialized and arcane programming talents needed to develop any software. The need for such specialization is decreasing as a result of object-oriented programming, but it is far from gone. Historically, it has been difficult to bill for software development, but recognition by third party payers of long-term cost savings is rapidly changing this situation and encouraging individuals and institutions to begin development. Many rehabilitation facilities have on staff various types of cognitive therapists, speech therapists, physical therapists, recreational therapists, occupational therapists, and psychologists, all of whom could work to create appropriate customized software. Although HyperCard requires some programming skill, its speed of development permits a multiple iterative process in which the treatment team professionals can make recommendations and modification suggestions that reappear as functional software within a day or two. Prosthetic aids may be introduced in the rehabilitation facility, but they are typically intended for home placement and support.

With inappropriate application, the microcomputer can increase, not decrease, risks for home usage. Consider the typical kinds of home uses that are being widely proposed. The computer can be the daily reminder, task organizer, boon companion, and prosthetic device to help the impaired elderly person remember and initiate the activities required for independent living. It can become a sequential timer, reminding the individual not only of dietary requirements and medication schedules but also of how to accomplish them. It can be a mechanism for banking or grocery buying, overseen by a core program that limits and recalls. It can provide access to friends or education via information and communication services like ElderNet. By monitoring behavior, it may serve as a neuropsychological testing device and early warning system, with daily input based on assessment by artificial inte11igence and diagnostic programs. The microcomputer can also be the failsafe mechanism, the external "brain" for household circuitry.

All these are potential sources of disaster and litigation. Software programming and direct personal contact by licensed professionals solve only part of this problem. The consequence is that there will have to be a protracted phase of research and model building. It will not help that research in scientific gerontology is essentially a collection of disciplines without a core paradigm.

The cost problem is multidimensional. For example, SpeechWare costs only $39.95, assuming a person already has a computer. If not, it costs about $4,039.95, so the hope is that the computer can be used for other things. In general, insurance companies pay because clear benefits and/or long-term savings are readily apparent. As we get beyond specifically targeted software solutions to more general capabilities like the smart house as a long-term care solution, however, costs and benefits become less clear and are intermingled with government policy.

Government policy and the current desires of the elderly population are not well matched, as evidenced by the repeal of the catastrophic coverage law. High-tech solutions of any scope will require complete participation from many segments of society. We live in a world of diminishing natural and population resources. The 1989 U.S. Bipartisan Commission on Comprehensive Health Care (the Pepper Commission) was charged with developing a strategy to provide the 31 million uninsured Americans with health care and shape a long-term health care plan. More pieces cut from the same pie, with less and less filling.

It is unclear if the cost of developing a microcomputer-supported living environment would be more than offset by the savings in man-hours to the health care professions and the costs of building chronic or custodial care facilities for those who can live independently or in nuclear clusters in a computer-controlled environment. Insurance companies and Medicare could realize savings, but it is difficult to imagine them paying to retrofit a dwelling without some proven expectation of effectiveness. Perhaps the utility and savings to be realized in the new types of living arrangements for older people will be recognized by the changing demographic, environmental, and cultural milieus. A computer and its attendant software is much cheaper than most types of attendant care in nursing facilities. And since the psychological and financial cost to family members who opt for home care is high, and respite is a central issue, it may be that computer assistance is cost-effective in these settings as well. We have arrived at a time when the technology enables us to form a partnership with health care professionals, new living communities, and the older population. Instead of the problems of the aged, we can begin dealing with their abilities.

Theoretical Position

The ProsthesisWare concept is essentially a pragmatic, empirical, and theoretical treatment approach to the problems of elderly people. Quite simply, that which is done is that which works. The flexibility offered by high technology permits the multiple iterations of a project that are required to refine that which experience has shown to work.

In fact, two sorts of quasi-theoretical underpinnings should emerge in the future. One derives from ecological psychology, the other from cognitive Neuropsychology. For ProsthesisWare to be effective, it must conform to the unique needs and circumstances of the user. This models the thinking of the ecological psychologists in rehabilitation. At a practical level, the "reality" that can be manipulated by the microcomputer is most effective if it is ecologically valid for the particular patient.

A Prosthesis Ware example to illustrate this point would be a little reminder and sequencing routine for the personal hygiene ADL of shaving. Here the person sees computer-managed photographs or video of himself in his own bathroom. Depending upon neuropsychological requirements, the computer can offer various techniques to ensure success with the necessary steps in shaving. Manipulating a reality as nearly as possible to a user's actual environment increases the likelihood of usefulness to that environment. Figure 2 shows a demonstration screen similar to one that would be used by a person requiring cues to sequence or monitor any behavior or reminders for performing specific tasks or skills.

Emerging theory and practice in neuropsychology will increasingly play a role in high-tech solutions for the elderly. The problem might be thought of in engineering terms as a special case of software interface design. This approach tends to ignore the unique and sometimes bizarre cognitive functions of the impaired brain. At best, it produces an "easy or intuitive" interface, but it can ignore dynamic and idiosyncratic neurological processes. Perhaps the most debilitating cognitive incapacity that affects Macintosh operation is the loss of excessive functions and not, as many suppose, the loss of memory. Memory and language overheads are much less a problem with the Macintosh's iconic and graphic interface.

The neuropsychologist as ProsthesisWare designer works in partnership with software engineering to create the appropriate customizations. For example, the Macintosh has a much vaunted and, in fact, quite intuitive interface. But for the neurologically impaired, the usual is not sufficient. Consider "pull down" menus in which actions are accomplished: they provide few cues as to their content and function, and if "out of sight, out of mind" is a problem, they can be quite dysfunctional. Other interface conventions that appear quite simple can pose serious design considerations. Programming construction kits like HyperCard place a degree of control over the interface, which is a benefit; it is also a potential drawback, however, in that a consistent and standardized interface is a prime feature that keeps memory overhead for computer operation low. The success of any program ultimately is whether it is used. Utility and interface are by far the largest determinants of use by the elderly person.

Design Solutions

How "special" should design solutions be?

Design solutions should be unique for each person. The great advantage that high-tech computer-based solutions offer is their flexibility and customizability. Cognitively, it does not make sense to design for a general group like Alzheimer's patients, for example, when object-oriented programming capabilities make it possible to design for each unique individual. A paucity of data concerning ProsthesisWare, however, leaves the justification for the development expense, and maintenance of high-tech solutions to just the "face validity" that individualization is ecologically sensible and offers the highest likelihood of achieving the desired effect. Quite simply, the alternatives to this face-valid approachthe general training or rebuilding of the central nervous system, the use of the computer as a tutor in drill and practice in general cognitive skills, or the application of various memory and cognitive strategies used in normal subjects have been tried in the past, and the literature can be fairly summarized as indicating no consistent or outstanding success. The failure is due to a lack of specificity for the individual and a lack of generalization to real world, ecologically relevant circumstances.

What design solutions will people find acceptable or useful?

The smart house approach can be an ideal aid to the elderly person looking for an independent life, but only if the architects of such environments are willing to look clearly at the compensatory strategies required by the impaired aged. For the well elderly person, doesn't such technology amount to gratuitous gadgeteering? Can current designs be used by the cognitively impaired? What good is a wallmounted touch screen if the resident can't understand its function? Again, clinical experience in the cognitive sciences would seem an important part of the design process.

There will be, however, many useful smart house concepts in the near future. The voice-controlled computer that responds in kind begins to take on the form of a companion, follows rational commands, and reminds the resident of everyday living activities. The optical kitchen scanner currently used to identify recyclable plastics for the trash compactor will be incorporated into the environment to keep track of dietary intake and prepare replacement lists via the bar codes embedded in packaging. The same could hold true for the dispensing of medication. Household circuit controllers offer safety as well as efficiency, and integrated security systems offer medical and other emergency assistance. All this will be done unobtrusively and in a way that compensates for any particular human fallibility.

The median age of the population is steadily rising, and few of the baby boomers are sanguine about their own futures as far as their retirement and health care during those years are concerned. They recognize that the current health care crisis will carry over to their future, even if they are oddly misinformed or ignorant as to how the Medicare system works. Add to the baby boomers the number of currently retired individuals, and we get a large voting block that can dictate to Congress the need for a revised system of health care for the elderly population. What are we going to offer that august body if not a wiser, more cost-effective, more humane system based on the proper use of the technology available and under development? We are making our own beds and, presumably, would like the opportunity to sleep in a comfortable one. The integrated computer-supported environment would seem to offer a high probability of success. Again, its flexibility is both its strength and its weakness. ProsthesisWare does not constitute a fixed product but rather conforms to information-age concepts of rapid adaptation and dynamic change. The weakness, however, is that these are the very characteristics that are limited in elderly people and are sources of anxiety, frustration, and resistance.

How do other design solutions relate?

The main application of the concept is expected to be in the primary living environment since complete portability for many desirable functions is still some way off. Thus, designers of living environments need to accommodate the electronic technologies. The question is, which technologies? The architectural challenge seems to be to permit modularity and the "swapping" in and out of components. The main issues, then, are in the realms of access and connectivity. Although cables may change from twisted pairs to coaxial to fiber optic, they will be used for some time to ensure reliability and freedom from interference. Therefore, easy conduit routing would be a design advantage. Infrared control is likely to expand. For room-to-room access and system integration, light conduits and mirrors need to be planned. For a ProsthesisWare module like Controller, the ability to access and monitor individual household outlets needs to be made simple and commodious. Telephone connectivity is extremely important and somewhat complex. It can be summarized as requiring digital, high-speed, voice, and data components. Current data switch technology is quite good, but new construction needs to establish the necessary "backbone," hardware, and interconnectivity. Figure 3 shows the computer interface for the ProsthesisWare research module Controller that can be used to create a smart house.

Are particular types of design solutions more desirable?

Generally, high-tech solutions are not as desirable as other approaches for various reasons that usually include cost, reliability, effectiveness, practicality, and sturdiness. It is the current nature of high technology, however, that all these factors are in flux. At this time it is possible to say that in some situations, especially for persons with particular cognitive deficiencies, high-tech customized solutions may be expensive but offer long-term savings and/or greater clinical effectiveness. ProsthesisWare applications have been pioneered with the traumatically brain injured, in part because these people have a longer life expectancy, better insurance, more supportive infrastructure, and more comprehensive rehabilitation services than persons with other types of cognitive deficiencies. Since this field is still in a pre-clinical research phase, however, it is not well received or understood by funding agencies relevant to gerontology or aging.

Interestingly, there seems to be considerable antipathy to technological solutions among many professionals currently in the field of gerontology. They often raise the kinds of objections first heard when computers began to be integrated into classrooms and teaching specifically, claims of social isolation, depersonalization, and impracticality, not to mention expense. In education, none of these fears has proven to be founded. In fact, as experienced students advance through the system, many early problems evaporate as the technology becomes a normal part of life skills and experience. A similar result has occurred in the workplace. It will not be that long before population cohorts without technophobia swell the ranks of the elderly population.

Computer solutions hold considerable promise, but not much has been delivered. This should not deter us from beginning, but it should require that we develop in a broad, responsive, open, and flexible manner. As yet unknown or unexpected benefits can emerge from a core household system that has the microcomputer as an integral component. For example, recent clinical observations with applications for the brain injured suggest that future designs might also take advantage of some arousing and some palliative effects observed with computer use. Such ideas cannot be effectively pursued without an open software, hardware, and environmental architecture.

What can be done to make existing housing more accommodating?

Although the fundamental concept of ProsthesisWare is customization, it is not usual for the expense of a comprehensive system to be justified and retrofitted to an individual home environment. Nevertheless, various components may be practical. To use SpeechWare as an example again, it and a speaker phone can empower the stroke patient or aphasic by providing a synthesized voice.

The computer, appliance, service, entertainment, and communications industries are embarking on a new growth spurt in applications of technology to the personal and home domains. Wise policy will ensure a general level of standards and standardization that will permit competition, creativity, and growth, and yet provide enough common ground to facilitate interconnectivity, ensure safety, and regulate development. The current telephone system provides a good analogy to illustrate the point without belaboring both the strengths and weaknesses of the system, which are well documented. No matter which company is used, the human machine interaction is basically the same, a result of standardization. Yet advances in underlying technology, new features, reductions in cost, and proliferation of ideas have occurred through competition. For ProsthesisWare to flourish, it requires both standardization and competition.

Are there some low-cost, high-benefit advanced technology solutions?

Cost is relative. If ProsthesisWare technology can delay or prevent other expenses such as custodial care, it may be that its high cost will be offset by, higher cost savings. Even on a fairly short term basis, the transportability and the transferability of computing machinery may make its use cost-effective. For the most part, there are insufficient data to determine clearly the limits of any microcomputer application or its benefits beyond the hoped-for potential.

How should neuropsychology professionals be trained in design solutions?

The typical doctoral-level training program neuropsychology has its roots in clinical psychology, and psychometrics. Consequently, its focus is on assessment and not rehabilitation. Within assessment, the emphasis is on test batteries of general cognitive function without regard to their ecological validity, and not on evaluation of ADLs or on the monitoring of interventions. It may not be possible within the discipline of psychology to redirect the focus of an emerging subdiscipline, as there is tremendous inertia institutionalized in curricula and licensing and in restrictive accreditation procedures. These forces have led to a split organized psychology and to the emergence of a large, scientifically oriented alternative national organization.

The tendency for clinical psychologists to eschew the basic psychological sciences and the generally inadequate training in computer sciences indicate that future developments in ProsthesisWare are not expected to emerge easily from that group. Probably the most likely effect will be analogous that seen in the behavioral neurosciences, which escaped restrictive clinical psychology by participating in the emergence of neuroscience as a discipline. While the computer offers the potential for an effective treatment for the neuropsychologist, who by training and skill is uniquely qualified to direct it, it may only be able to emerge if neuropsychology distances itself from clinical psychology. New fields like gerontology seem willing to provide growth and success to such professionals, but there are broad gulfs between the sociological and neuropsychological philosophies. Ultimately, such disciplinary shifts and concomitant training occur only if there are sufficient resources to prompt such movements. In rehabilitation, many neuropsychologists have found the impetus and resources to move to new paradigms. We suspect the emergence of gerontology programs will have a similar effect.

At Drexel University, the first free-standing doctoral program in neuropsychology itself was created. Recent faculty additions have brought three new gerontological staff members to augment that research and training area. Combined with the resources of a technological university and the facilities of the Philadelphia community, we hope to pioneer in new curriculum and training initiatives in the broad field of evaluation, treatment, and remediation of cognitive and brain dysfunctions.

Table of Contents