| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||
Correspondence: Address correspondence to Jon A. Sanford, Atlanta VA Medical Center, Rehab R & D Center (151R), 1670 Clairmont Road, Decatur, GA 30033. E-mail: jon.sanford{at}med.va.gov
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: Home modifications • Aging in place • Home assessment
Home modification, which includes reconfiguration or addition of spaces, installation of new or adaptive hardware and products, removal of barriers, and rearrangement or changing of finishes and furnishings to meet the specific needs of older adults and their caregivers, can facilitate functional independence, improve safety, minimize the need for personal care services, and reduce health care costs (Connell & Sanford, 2001; Gitlin, 1998; Mann, Ottenbacher, Fraas, Tomita, & Granger, 1999; Pynoos, 1993). As a result, when remaining at home best serves the needs of the individual, home modification can be an effective strategy to help older adults age in place.
Despite their benefits, there are a large number of older individuals who need, but do not have, modifications to facilitate aging in place. Moreover, it is estimated that of the 2.1 million households with older adults who need home modifications, only 1.14 million of these households have them (Joint Center for Housing Studies, 2000). In other words, almost half of all elderly households (and many include more than one older individual) that need modifications to facilitate aging in place do not have them.
A number of factors contribute to the lack of home modifications. These include limited awareness of home modifications among older adults, institutional appearance associated with many modifications, a lack of affordability, limited funding for home-modification programs, and a fragmented, difficult-to-access service-delivery system that includes few providers with expertise in both the needs of older adults and environmental modifications to meet those needs (Pynoos, Liebig, Overton, & Calvert, 1996).
Currently, a number of efforts are being undertaken in various communities to address many of these issues, including consumer education through local home modifications actions coalitions, such as those in Pasadena and Santa Clarita, California (Overton, 1998); more consumer-acceptable designs featured through wider availability of universal design features as well as through local visitability ordinances requiring basic accessible features in all homes (National Resource Center on Supportive Housing and Home Modification [NRCSHHM], 2003); increased funding through reverse mortgages; and increased affordability through subsidized and volunteer housing programs (e.g., Philadelphia Corporation on Aging and Rebuilding Together, respectively). In contrast, there have been few efforts to effect changes in the service-delivery system that will increase access to home-modification specialists. The two major efforts that have been made—the Certified Aging in Place Specialist (CAPS) Program by the National Association of Homebuilders (NAHB, 2003) and the Executive Certificate Program in Home Modifications offered online by the NRCSHHM at Andrus Gerontology Center, University of Southern California (NRCSHHM, 2003)—have focused on increasing the number of knowledgeable providers. While potentially effective, it will take a considerable amount of time before there are a sufficient number of trained providers throughout the United States to have a substantive impact on service delivery.
An alternative strategy is to extend the capacity of specialists who already understand the needs of older adults and who can prescribe appropriate modifications for aging in place. Unfortunately, the disproportionate geographic distribution of these specialists limits their ability to provide in-home services to many individuals with unmet needs, particularly those who live in smaller communities or in remote areas. Most specialists, such as those at Extended Home Living Services, an accessibility remodeling firm outside of Chicago, are located in major metropolitan areas where there are large enough concentrations of older adults who need home modifications to support the businesses and agencies that provide these services. Even then, there are usually no more than a few providers in any one city. Although there is no way to identify how many public and private providers of home-modification services actually exist and where they are located, the sole national directory of home-modification resources provides some insights (NRCSHHM, 2003). For example, only 10 providers are listed in the city of Chicago, the third largest metropolitan area in the United States. Comparatively, there are only six, each in a different city, in the rest of the state of Illinois. There are even fewer providers in states with smaller, less urban populations. For example, there are only four providers listed for the entire state of Iowa.
In-home services are typically provided over small geographic areas due to restrictions associated with travel time. While it might be possible for current in-home service providers to travel farther to serve more geographically remote individuals, more time would be required for travel, thus effectively reducing the number of people that could be served. Moreover, increased travel distance drastically increases the cost of services. For major renovations or new construction, it might be cost effective to pay the added travel costs; for minor changes, like adding grab bars, added travel might not make the service cost effective. An important consideration is that older adults, in contrast to younger people with disabilities, usually only make minimal modifications to their homes.
One way to increase the capacity of specialists to serve more individuals without incurring the cost of travel is to provide remote delivery of home-modification services. While it is obvious that construction or installation of the modifications is not possible through remote techniques, this aspect of service delivery does not necessarily require a specialist to carry out. In contrast, assessment requires a specialist but could be accomplished through remote strategies that enable local practitioners to collect the on-site information needed by a specialist to identify problems and prescribe individualized modifications. The modifications could then subsequently be implemented by a local contractor or remodeler.
Whereas remote techniques can potentially eliminate the high cost of traveling long distances to perform in-home assessments, the success of this strategy is dependent on the ability to provide a specialist with the same information about individual competencies and environmental demands as would be obtained through a traditional in-home assessment. Such a strategy, if successful, would enable specialists to prescribe home modifications for anyone, anywhere, regardless of location. The purpose of this article is to examine whether or not remote assessment is feasible and whether or not it concurs largely with traditional in-home assessment based on expert judgment. Thus, we examined whether the paper-and-pencil and televideo home assessment and planning methods achieve the same results as traditional home-visit assessments and planning methods.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Instruments
We conducted a detailed review of existing assessment instruments as well as traditional in-home assessments by specialists develop the initial versions of the remote instruments. Both remote assessment tools were developed by multidisciplinary teams that included therapists, case managers, and home-modification specialists. The primary considerations were to develop protocols that captured the essential information required by a specialist as well as to promote sensitivity in interacting with and engaging older adults. In practice, the test protocols were developed to replicate the step-by-step process of photographing, measuring, and recording critical characteristics of the home environment that are used in a traditional in-home assessment.
Paper-and-Pencil Remote Protocol
The Comprehensive Assessment and Solutions Process for Aging Residents (CASPAR) is a consumer-directed assessment that enables an older adult, family, and/or caregivers to identify high-priority home problem areas that are in need of modification (Sanford, Pynoos, Tejral, & Browne, 2002). CASPAR permits a nonspecialist in home modifications, such as a local home-health therapist or other aging service provider, to collect the same critical information about older individuals and their homes that would be collected by a specialist. The information is then sent to a home-modification specialist to identify, design, and specify individualized solutions.
CASPAR documentation includes: client-reported information on functional abilities and types of problems experienced completing tasks in the home (see Figure 1); physical measures of the home by activity and problem areas (see Figure 2); and summary information supplied by the individual who administered the assessment. The completed forms and photographs are sent to a trained home-modification specialist to analyze, determine needs, and make recommendations for home modifications.
|
|
The televideo protocol entails a 3-step process of photography, preassessment, and televideo assessment. The photographs are used to complete a preassessment in which general characteristics of the home (e.g., are there steps between the parking area and exterior door?) are described (see Figure 3). From this information, potential problem areas in the home that need further analysis through a remote televideo assessment are identified. In this final step, more detailed information is collected (e.g., height of exterior steps) in each of the problem areas of the home to determine the specific modifications (e.g., length of ramp required to access the home) that will best fit the needs of the individual in his or her home (see Figure 4).
|
|
In this study, we used different methods of record keeping for the traditional in-home assessments. For the CASPAR comparisons, specialists used their own improvised strategies to record assessment data. In the televideo comparisons, the same record-keeping strategy was used for both the remote and in-home protocols.
Equipment
The video system used for televideo assessment consists of "off-the-shelf" technology to transmit real-time audio and video images using plain old telephone systems (POTS) located in virtually all homes. The images are transmitted from a remote videophone (StarView 500 TravelStar Pro XL) plugged into any phone jack in the home to a base station videophone (StarView 2000 Pro plugged into a large computer monitor) used by the specialist (Figure 5). The videophones are designed to replace standard phones without additional costs. The phone calls can be initiated at either end in the same manner as a typical telephone call, although for purposes of this study, calls were initiated by the base station to avoid long-distance charges to the clients. An initial auditory connection was established between the phones, after which the video button was depressed to establish the video transmission.
|
|
Sample
We recruited 70 participants who were consecutive clients of Extended Home Living Services (EHLS) and had requested and were receiving services provided by the company. As part of the company's typical assessment, clients were at home during the assessments, and they identified specific problems that they encountered in the home. This portion of the sample comprised the group that was used to test the remote paper-and-pencil protocol.
In addition, 3 participants were inpatients from Shepherd Center, a rehabilitation hospital in the Atlanta area, who had not functioned in their homes after the onset of their disabilities. Their homes were assessed to identify the types of problems that they might encounter based on their current and anticipated levels of functioning. This portion of the sample was used to test the remote televideo protocol.
The different sample sizes reflect the different purposes of the two sets of evaluations. The CASPAR approach was tested as part of a Phase 2 Small Business Innovative Research (SBIR) project to determine the efficacy of the remote process to assess its potential as a nationwide service. In contrast, the televideo approach was part of a Phase 1 development project intended to demonstrate that using teleconferencing technology to conduct home assessments was feasible.
Procedures
A pair of home-modification specialists, matched in background and experience, participated in each study. A remote assessment and a traditional in-home assessment were completed in all 73 homes. Data collection for the remote assessments was carried out by an individual inexperienced in home modifications, while analysis of the remote data to identify problems and determine solutions was performed by the specialist who was unfamiliar with the home or client (i.e., the specialist who did not perform the in-home assessment). Consequently, in both evaluations each specialist performed only one assessment per home, either a remote or a traditional in-home assessment. Traditional in-home assessments and data analysis were conducted by a specialist, blinded to the remote assessment data.
Collection of assessment data
In the EHLS sample, in-home assessments were conducted by one of two home-modification specialists from EHLS as part of the company's usual services. There was no standard method of conducting or documenting these assessments. Specialists were matched with clients by territory, the usual manner in which they were assigned. As part of the study, a second home visit was conducted in each home by an occupational therapist, inexperienced with home modification, who used the CASPAR forms to collect necessary data.
In the Shepherd Center sample, one occupational therapist conducted in-home assessments. A second occupational therapist independently conducted remote home assessments by directing a technician who was inexperienced with home modifications through the home. Both occupational therapists specialized in home modifications. Data for both the in-home and remote assessments were recorded on the same assessment forms. Based on the photographs and preassessment information, the occupational therapist instructed the technician where to locate the camera, what information to collect (e.g., what measurements to take), and how to collect it (e.g., how to measure a doorway). Because this was a feasibility study, remote and traditional comparisons were based on all problems identified in the preassessment process to compensate for the small number of homes. In this part of the study, the type of assessment was randomly assigned. One specialist performed two televideo assessments and one traditional assessment, and the other performed two traditional and one televideo assessment.
Identification of home problems
In both parts of the study, we analyzed remote and in-home assessment data to identify task-related problems and prescribe appropriate modifications to address those problems. Remote data from the EHLS sample were analyzed by a specialist blinded to the in-home assessment. Remote data from the Shepherd sample was analyzed by the occupational therapist who conducted the remote assessment. The in-home data in both studies were analyzed by the same specialist who had collected the data. It should be noted that because the 70 assessments in the remote paper-and-pencil protocol comparison focused on actual issues identified by clients prior to the assessment, those data reflect fewer problems and solutions per home than data from the 3 assessments in the televideo project, which focused on potential problems in the entire home.
Problem identification varied by type of assessment. In the remote paper-and-pencil protocol, problems examined were based on clients' experiences and self-report, and task performance problems were derived secondhand from observations made by the therapist conducting the assessment. In-home assessments incorporated observations of task performance. Similarly, problems identified from the remote assessment were based on self-report. In the remote televideo protocol, there were no observations of task performance. Thus, problems were based on specialists' expertise and expectations of needs given their knowledge of a client's level of ability and knowledge of environmental impact on those abilities.
To simplify comparisons between protocols, we grouped problems with task performance by type of activity. This resulted in seven categories of problems: getting in and out of the house, going up and down stairs, moving from room to room, using the bathroom, using the kitchen, doing laundry, and using the bedroom.
Identification of home solutions
Regardless of the type of assessment, solutions were determined by specialists using their knowledge and understanding of the problems and the situation. While there is no "how-to" book of home modifications, there is a fairly well-defined set of solutions for any particular problem that specialists understand. For example, if an individual is having difficulty getting up the front steps to the home, the solutions could include adding or fixing handrails, making the step risers lower, adding a ramp, installing a lift, or using another entrance. If the individual uses a wheelchair, only the latter three solutions will work. It takes an expert to know the widest range of possible solutions (or where to look) to any particular problem in order to provide the client(s) with sufficient information to make an informed decision about which solution is the best for them in their own situation.
Data analysis
To achieve greater objectivity and reduce potential for individual bias of assessors, we determined reliability of the remote assessments by level of agreement between each pair of traditional and remote assessments. This included sensitivity and specificity rates for problems identified with task performance (e.g., difficult to maneuver a wheelchair) and agreement in recommendations for modifications (e.g., widen doorway, move toilet).
To determine whether the remote assessment identified the same accessibility problems as traditional on-site assessments by home-modification specialists, we used a typical "Yes–No" response design to determine true response rates, including sensitivity (true positive rate) and specificity (true negative rate), and false response rates, including correct rejection (false positive rate) and errors (false negative rate). A true positive rate (yes–yes) is the identification of problem by the remote assessment that is present according to the in-home assessment. A true negative rate (no–no) is the proportion of problems not present of those not identified by either type of assessment. A false positive (yes–no) rate is the number of problems identified by the remote assessment that are absent in the in-home assessment. A false negative (no–yes) rate is the incorrect or lack of identification of a problem that is present in the in-home assessment. In this type of analysis, validity is dependent on the percentage of matching responses, that is, true positive and true negative, between the remote assessment and the gold standard in-home assessment.
Although false response rates are merely the total sample size less the true response rates, they are important to consider because false positives and false negatives have differential consequences. A false negative represents failure to identify or diagnose a potential problem. Such failure results in no modifications, which can have negative consequences for clients. In contrast, a false positive represents overidentification of problems—the diagnosis of problems not deemed problematic by the "gold-standard" in-home assessment. Thus, false positives err on the side of caution and safety. As home modifications are designed to enhance the safety and facilitate independence in clients, erring on the side of caution and safety through overidentification of problems is preferable to the converse.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
Specificity
Overall, the paper-and-pencil protocol had a true negative agreement rate of 97.8% (1,525 out of 1,559 problems). By category of problem, the specificity ranged from a low of 93.3% for using the bathroom to a high of 99.7% for using the kitchen. The televideo assessment had a slightly lower specificity at 88.2% (30 out of 34). Specificity data for the televideo study involved fewer problems (n 
12) per cell. Nonetheless, only two problem categories had true negative rates less than 1.00—moving from room to room, which had a specificity of 5 out of 6, and using the kitchen, which had a specificity of 6 out of 9.
Total true response rate
As reported in Table 1, overall agreement for between the remote paper-and-pencil (96.4%) and the remote televideo (87.1%) protocols were significant (p =.000) when compared to the in-home assessment. Using the remote paper-and-pencil protocol, response rates for all problem categories were significant (p =.000) with Kappas ranging from.524–.857. In contrast, only three problem categories had significant agreement rates using the remote televideo protocol. These included getting in and out of the home, going up and down stairs, and using the kitchen. The first two had 100% agreement; the latter had 79.2% agreement.
False response rates
An analysis of the problems that were incorrectly diagnosed suggests that 34 of the 59 errors (58%) made using the remote paper-and-pencil protocol were false positives. The frequency of false positives was fairly consistent across problem categories, accounting for 50% of the errors in going up and down stairs (4 of 8) and doing laundry (1 of 2); 44% of the errors getting in and out of the house (4 of 9), 46% of the errors moving room to room (6 of 13), and 64% of the errors using the kitchen (18 of 28). In contrast, false positives only accounted for about one third of the total errors (4 of 12) in the televideo comparison. However, the total number was so small that there were no more than five problems in any problem category. In fact, only two categories, moving from room to room (n = 1) and using the kitchen (n = 5), had false positives at all.
Agreement rates for problem solutions
Among problems that were correctly identified by the remote assessments, the agreement rate for both the remote paper-and-pencil (78.8%) and televideo (77.4%) protocols were significant (p =.000) when compared to the in-home assessment (see Table 2).
|
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
False Positive Data
False positive data have positive implications for the effectiveness of the remote instruments. In contrast to false negative errors due to missed diagnoses, false positive errors due to diagnoses that do not exist actually err on the side of caution and safety. This is particularly important in the absence of additional postassessment follow-up data that would likely be available in real-world situations. More than 60% of the errors attributed to CASPAR and 36% attributed to the televideo protocol were false positives. While falsely identifying a problem that does not exist might result in modifications being recommended when they are not needed, the consequences of such errors (not accounting for cost) are generally either benign and might be helpful, in contrast to consequences resulting from misdiagnosing a problem. For example, recommending a handrail on a stair because a problem was perceived when one did not exist, might actually be helpful in negotiating stairs or preventing a fall, whereas omitting a handrail when one is needed can potentially have negative outcomes.
Sensitivity and Specificity
Although the small number of problems analyzed affects sensitivity significance levels, sensitivity trends alone are encouraging for both remote protocols, particularly for analyzing problem categories where the number of potential items is relatively small, such as getting in and out of the house and going up and down stairs. In contrast, sensitivity rates are somewhat lower for those categories, such as moving from room to room and using the bathroom, where the number of potential solutions is large. In these instances, there is a greater potential for disagreement between the specialists on what problems are most important to consider.
Nonetheless, we consider agreement rates as low as 70% in this study to be acceptable, as they can be attributed, at least in part, to methodological constraints imposed upon these strategies to avoid potential contamination of results. As a result, unlike the traditional in-home protocol, the remote protocols did not include typical postassessment follow-up with clients and health care providers to gather additional information needed to verify data and discuss potential solutions. It is expected that when additional follow-up information is solicited in real-world situations, remote assessments will be more effective than they were under the more constrained testing conditions. On the other hand, a remote protocol capable of identifying 70% of the problems responses might be acceptable to home-modification experts when compared to the alternative of unknowledgeable providers making decisions about home modifications.
Home Modifications Solutions
Even though the remote protocols had somewhat low rates of solution agreement, the implications are probably minor. In real-world situations, the range of acceptable solutions for any one problem can vary widely. In instances where several modifications will meet a client's functional needs, alternatives are generally recommended to help an older adult, caregivers, and/or family members select the one that best fits the particular circumstances (e.g., cost or preferences). Therefore, singular recommendations that were considered a mismatch for study purposes might both be recommended as functionally equivalent alternatives in an actual situation.
Implications for Practice
Advantages of remote assessments
The anticipated advantage of remote assessments over in-home protocols is that they have the potential to reduce the cost of home assessment, particularly for people living in remote areas, by eliminating the barriers of time and distance for experts. These protocols are intended to be implemented on-site by providers of home-health services, who have greater access to and live in closer proximity to clients in need of home modifications and may have a keener insight as to the typical functional abilities and individual goals of these older individuals. As a result, these protocols also have the potential to permit increased teamwork among health care providers, particularly for providing opportunities for collaboration between providers of aging services who have a better understanding of elders' needs and abilities and providers of home modifications who have greater insights into the barriers and facilitators in the home environment.
Differences in remote assessments
The major difference between the two remote protocols is access to both the home and the client in the home. Regardless of how detailed the CASPAR data collection might be, specialists are limited in their understanding of the environment and the client's problems because they never actually see the home or observe the client interacting with it. Televideo not only brings the specialist into the home but also provides the opportunity to observe activity within the context in which it occurs.
Measuring the home
In the CASPAR protocol, specialists are dependent on someone else to provide them with data about the home from which they can identify problems and solutions. Therefore, defining situations and standardizing measurements are critical to correctly identifying problems. If a less experienced assessor measures a unique or unconventional situation in a manner other than a specialist expects it to be done, problems could be misidentified or missed altogether. To minimize the effects of these discrepancies in CASPAR the specialist relies on a comparison of the measurement data to photographs taken of the home as well as follow-up interviews with the client. In contrast, the televideo assessment protocol does not rely on someone else to determine what should be measured and how. The data collection is directed entirely by the specialist who identifies problems and recommends solutions. As a result, the specialist is not dependent on someone else's interpretation of the home environment. He or she can observe the data collection, verify measurement accuracy, and compensate for unique or unconventional design features. This minimizes the need to set up additional appointments with the client and data collector to verify data measurements and client goals.
Observing person–environment interactions
In contrast to CASPAR, which relies on self-reported difficulties in task performance to identify problems, televideo gives specialists the opportunity to observe actual task performance. This process is not only more reliable in identifying problems but in determining the best solutions. Moreover, it can be used to conduct follow-up evaluations of modifications made as well as any new barriers that might be encountered as an individual's functional abilities change. While observations of clients interacting with the home environment were not part of the televideo study reported in this article, having demonstrated acceptable equivalence of the televideo, its application for providing a more comprehensive set of home interventions can be determined. As a result, a randomized controlled trial sponsored by the VA Rehabilitation R & D Service is currently being conducted to determine the efficacy of televideo to provide a variety of in-home rehabilitation services.
Limitations of the Study
The major limitations of the study are the lack of comparability between the two remote protocols and the potential for measurement error resulting from the reliance on individual judgment of the specialists. First, it is important to recognize that despite the common objective of demonstrating equivalency between the remote and in-home assessments, the two comparisons reported here had very different goals and anticipated outcomes. CASPAR was a Phase II development project that was designed to determine efficacy of the new assessment protocol. In contrast, the televideo comparison was a Phase I project that was designed to show feasibility of concept. As a result, the data-collection protocols, instruments, and samples from the two strategies were not intended to be comparable. In addition, the small frequency of problems identified in some problem categories resulted in sensitivity data that was too small to produce significance, even in cells with 100% agreement. For example, in the televideo comparison, using the bedroom was insignificant even though 3 out of 3 problems were correctly identified. The trends, however, indicate that the remote assessments are promising, particularly given the constraints of the research methodologies.
Second, whereas each matched pair of specialists in the two comparisons was matched on background and experience, the use of different specialists may have affected the accuracy of problem identification and solutions. This methodology was used to counter rater bias of each specialist performing both assessments in each home. However, the methodology also presumed that matched pairs of specialists would assess the same home in a similar manner. Alternatively, each specialist could have performed both remote and in-home assessments in every home, with the type of assessment randomly assigned to counterbalance the effects of bias. Unfortunately, such a methodology would have necessitated twice as many assessments, which were beyond the scope of either comparison.
Finally, the study was limited by its focus on demonstrating equivalence of each remote protocol with a traditional in-home assessment. While this is the necessary first step in developing effective remote-assessment services, the usability of these protocols by local aging service providers who would implement these strategies and acceptance of the potentially invasive photographic and documentation methods by older adults who would receive the services are unknown. The rationale for this limitation is that if equivalency could not be demonstrated, then acceptance and usability were less relevant. On the other hand, because the data suggest that the remote assessments are promising, further studies of receptivity by older adults as well as the usability, training required, and user satisfaction are warranted.
|
Footnotes |
|---|
1 Atlanta VA Medical Center, Decatur, GA.
2 Extended Home Living Services, Inc., Wheeling, IL.
Decision Editor: David E. Biegel, PhD
Received for publication September 30, 2003. Accepted for publication December 1, 2004.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
