Pope (1992) asserted that there was a significant need to examine secondary conditions among individuals with development disabilities. In the present study we focused on that need. The development of a secondary conditions surveillance instrument is described, as are the results of a pilot survey conducted with adults receiving state developmental disabilities program supports and with their direct-care service providers. Results of a pilot survey are presented to illustrate how survey data might be used to improve systems of services and supports to enhance the health and participation of adults with developmental disabilities in community life.
Editor in charge: Steven J. Taylor
Over the past decade, there has been an explosion of interest in the health and wellness of people with disabilities (e.g., Horowitz, Kerker, Owend, & Zigler, 2001; National Institute on Disability, 1999; U.S. Department of Health and Human Services, 2000; Williams, 1998). Central to this interest is an awareness that individuals with disabilities may be at increased risk for a number of preventable health problems. These are referred to as secondary conditions (Marge, 1988) and are receiving significant attention in the form of a national disability prevention agenda (Brandt & Pope, 1997; Houk & Thacker, 1989; National Council on Disability, 1986; Pope, 1992; Pope & Tarlov, 1991). People with disabilities constitute only 17% of the noninstitutionalized population of the United States but account for 47% of total medical expenditures (Rice & Trupin, 1996). On average, their medically related expenditures are more than four times those of people without disabilities (Rice & Trupin, 1996). Arguably, it is the secondary conditions that may be responsible for a significant proportion of health problems and expenditures for persons with disabilities. Importantly, many of these conditions are preventable at the primary, secondary, and tertiary levels (Pope & Tarlov, 1991).
Marge (1988) defined secondary conditions as those conditions that may be experienced by individuals after they have a primary (or first) disability. Conditions range from medical complications (e.g., such as pressure sores and urinary tract infections) to problems of psychosocial adjustment (e.g., depression), and to environmental and quality of life issues (e.g., disability associated with access barriers) (Seekins, Smith, McCleary, Clay, & Walsh, 1990). The functional consequences of a chronic health condition are often more severe for a person who already has an impairment (Sutton & DeJong, 1998). In this framework, having a disability is viewed as a risk factor contributing to one's vulnerability to a variety of preventable problems that can limit health, function, independence, and participation in life activities.
Developmental disabilities hold a significant place in this new paradigm of health promotion for people with disabilities (Pope, 1992). Currently, there are 2 to 4 million people with developmental disabilities in the United States (Pope & Tarlov, 1991). Many of them are supported through state and local partnerships with community service providers. Prouty and Lakin (2000) reported that between the years 1977 and 1999, the number of individuals eligible for and receiving residential services increased 45.8%, from 247,780 to an estimated 361,172 nationally. They also reported that on June 30, 1999, these services were being provided in 113,633 residential settings, of which 93.2% had 6 or fewer residents. Our purpose in this report is to briefly describe the development of a tool for assessing current experiences of secondary conditions among adults living in these residential settings and to suggest ways in which resulting data may be used to shape health promotion programs that can accompany other community-based resources.
Historically, the development of the community-based model of supported living focused heavily on behavioral, social, employment, and educational goals. Generally, as people moved from institutions into the community, health concerns were equated with and assigned to medical professionals. A wide range of forces (e.g., the maturing of the model, pressures for cost containment of special programs, the general public interest in health promotion and wellness, the emergence of managed care) is creating conditions for the development of a more comprehensive approach to health within the community-based system of supported living arrangements. A critical component in the development of effective health promotion programs is a system that monitors health-related problems within a defined population (Graitcer, 1987; Pope & Tarlov, 1991; Teusch & Churchill, 1994; Thacker & Berkelman, 1988). A health-monitoring system can conduct surveillance activities to provide measurements of the incidence, severity, and prevalence of health problems. Individual and system planners can use these measurements as they prioritize and target specific issues, areas of concern, geographic areas, or population subsets and as they monitor changes over time. This type of information also is important to investigations of the protective and risk factors affecting the participation and well-being of persons with disabilities.
Figure 1 outlines the components that would be involved in a behavioral health-monitoring and promotion system. The model illustrates a method for systematically addressing the prevention and management of secondary conditions among adult consumers of state developmental disabilities service programs. Such a model would be anchored on one end with a site-based (e.g., group home, case-manager) health assessment and management system and, on the other end, by a state's developmental disabilities and health agencies. These agencies would monitor population health and risks and would develop systems-wide interventions to target selected conditions of high prevalence or cost.
Although methods, procedures, instruments, and normative information about secondary conditions are necessary for implementing such a model, they are scarce within the developmental disabilities literature. Data about secondary conditions experienced by people with developmental disabilities are extremely limited (Lollar, 1994; Pope, 1992; Pope & Tarlov, 1991; Strauss, Kastner, & Shavelle, 1998). Although specific secondary conditions are addressed in research and as part of current programs and services, they do not appear to be dealt with or measured under a unified conceptual framework or using systematic surveillance practices. As a result, we could find no studies in defined populations in which investigators report the incidence, prevalence, or severity of a range of secondary conditions. One exception is a recent study by Strauss et al. (1998) in which secondary conditions (i.e., functional limitations, such as limited motor and self-care skills) were measured for the purpose of assigning relative risk for mortality. However, those data cannot be used to estimate the incidence or the average severity associated with specific secondary conditions. These additional estimates would allow policymakers and program planners to target the conditions of highest priority for the most people or to identify low-incidence conditions of the greatest severity. A method for assessing the overall health and participation of a population across multiple settings is needed to establish baselines of limitation associated with health problems and then to evaluate changes over time and location. The purpose of this report is to describe our preliminary efforts to fulfill that need.
The instrument development phase involved the participation of consumers and service providers within the developmental disabilities service system. In addition, members of the Montana Developmental Disabilities Planning and Advisory Council reviewed each version of the instrument. Data were reported on 119 adults with developmental disabilities who were living in independent residences, family homes, supported living arrangements, group homes, or a residential care facility. Directors of 11 private service corporations and one state Intermediate Care Facility for the Mentally Retarded (ICF/MR) recruited consumers willing to participate and secured informed consent. Fifty direct-care providers were responsible for the completion of surveys for the consumers. Informed consent was secured by corporations for these participants as well. On average, reporters each completed two surveys, with one survey being the fewest and eight surveys the most completed by any one reporter.
The development of a Health and Secondary Conditions Surveillance Instrument for Adults With Developmental Disabilities (hereafter called the Seconday Surveillance Instrument) has been a collaborative and dynamic effort between researchers and relevant constituents (e.g., consumers, direct-care providers, case managers, work supervisors, and parents). The current product continues to be an evolving measurement tool responsive to the expanding literature describing health and health-related conditions, the needs and concerns of relevant constituents, and the advancing means of behavior measurement and analyses. The following describes the processes by which measurement protocol was defined and by which instrument content and scales of measurement were constructed.
A surveillance instrument for assessing the status of secondary conditions among adults with disabilities related to mobility impairments, such as spinal cord injury (see Seekins et al., 1990; Seekins, Clay, & Ravesloot, 1994; Seekins, 1992) served as a model in the development of the current surveillance instrument for adults with developmental disabilities. With a basic instrument framework in hand, we conducted literature reviews and identified additional secondary conditions for potential inclusion in an instrument for assessing the health of adults with developmental disabilities. A structured focus group involving women with developmental disabilities further contributed to the identification of important secondary conditions. Smaller focus groups were held at community-based group homes and sheltered workshops and involved consumers, group home managers, case managers, work supervisors, and parents. Finally, interviews were conducted with a developmental pediatrician and a health care coordinator for a private service corporation. Using these methods, we identified 65 secondary conditions as major sources of limitation for adults with developmental disabilities. For each condition, we developed descriptions that were based on medical resources (e.g., Batshaw & Perret, 1992; Dell Orto & Marinelli, 1995; Glanze et al., 1986; Goldman, 1996), the developmental disabilities literature, and information we collected in focus groups from consumers and direct-care staff members.
The 65 secondary conditions were then formatted to include a rating scale for each item, a box to check if an individual plan objective was targeting the condition, and the descriptions of the conditions. The resulting secondary conditions surveillance items were reviewed by relevant constituents for further refinement. To decrease completion time while maintaining content, constituents recommended that approximately one third of the items be included in broader categories or in descriptions of other secondary conditions. For example, asthma and respiratory infections were combined into respiratory problems, and the specific condition, rashes/skin irritations, was included in the descriptions of side effects due to medications, allergies and allergic reactions, and pressure sores.
Researchers next addressed the issue of who should complete the instrument. The Project Advisory Council suggested sources that could potentially provide health-related information about the consumer: the consumer, the case manager, the direct-care provider, a medical professional, a work supervisor, a family member, or the Individual Plan team. Two main considerations led to the decision to have direct-care providers as the primary reporters for the instrument. First, the Council envisioned the survey completion process as a self-directed training on the specific health concerns affecting adults with developmental disabilities and decided that consumers would benefit if direct-care providers were the recipients of such training. Second, the Council was concerned about the reliability and validity of the data and was influenced by research in the literature on aging, which has shown self-reports of functional status to be inaccurate when respondents have cognitive impairments (e.g., DeBettignies, Mahurin, & Pirozzolo, 1990). The Council's decision to have direct-care providers act as surrogate respondents included resolutions to emphasize collaboration with the consumer and others having knowledge about the consumer's health in the survey directions and training content. The first author trained (in person or via teleconference) direct-care staff members to complete the instrument. Reporters completed the instrument for each consumer within 6 weeks of the training sessions.
The final instrument included a demographic section, a health care utilization section, an inventory of life events, and 44 secondary conditions that were divided into three sections: observable secondary conditions, medically related secondary conditions, and secondary conditions related to mobility impairments. (These sections were developed to ease survey completion and to accentuate for raters and consumers conditions requiring medical attention; they do not indicate theoretical assumptions of homogeneity within sections.) Raters were asked to estimate the amount of limitation associated with each secondary condition on a 4-point scale that included the following severity rating options: 0 (no limitation), 1 (mild/infrequent limitation; limits activity 1 to 5 hours per week), 2 (moderate/occasional limitation; limits activity 6 to 10 hours per week), and 3 (significant/chronic limitation; limits activity 11 or more hours per week). Finally, 15 raters reported their attitudes toward health promotion and the Secondary Surveillance Instrument on an 11-item Reporter Reaction Survey.
From a psychometric perspective, instrument development activities followed three steps suggested for achieving content validity: domain identification, item generation, and instrument formation (Nunnally, 1978). Internal consistency reliability for the sum of the severity ratings associated with the 44 secondary conditions was acceptable, with Cronbach's alpha equal to .81 (SE = .01). That is, very little item-specific variance contributed to raw summary scores representing severity of limitation associated with secondary conditions. This alpha, in conjunction with item total correlations ranging from −.35 to .80 (M = .09), indicated that the item set of secondary conditions was relatively interrelated and lacked unidimensionality (for a review of these issues, see Cortina, 1993). Ongoing construct validation studies of this instrument will further inform the meaning of the summary measures representing overall limitation associated with secondary conditions. For the purposes of this report, summary scores related to independent items are presented, and implications of these descriptions within a behavioral health monitoring and promotion system are suggested.
Summary scores to describe how the pilot sample of consumers experienced secondary conditions during the past year were calculated as follows: A problem index for each secondary condition was calculated by dividing its sum of the severity ratings by the total number of consumers reported on for that specific condition. A prevalence rate for each secondary condition was calculated to estimate the proportion of individuals per 1,000 who may identify a specific condition as a concern. The prevalence rate was based upon the number of individuals in the sample who reported some level of limitation for that specific condition. The average severity of each secondary condition was calculated by dividing the sum of the severity ratings by the total number of consumers who reported some level of limitation for that specific condition. The percentage receiving treatment for a secondary condition reflects the number of individuals with Individual Plan objectives addressing the condition divided by the number of individuals who either reported some level of limitation for that specific condition or indicated no limitation but had an Individual Plan objective for the condition. In addition, the numbers of persons with Individual Plan treatments targeting specific secondary conditions were represented in percentages across the four levels of severity estimated for each secondary condition. Such distributions suggest, in descriptive terms, how severity of limitation associated with a secondary condition may relate to the existence of a corresponding Individual Plan objective. Finally, two percentages representing individuals experiencing limitation due to a secondary condition, but without Individual Plan objectives, were calculated to illustrate relative need for access to services and supports. Those percentages represent proportions of (a) the total number of individuals estimating any level of limitation severity associated with specific secondary conditions and (b) the total number of individuals estimating moderate or severe limitation severity associated with specific secondary conditions.
Preliminary results of the survey were presented to the Project Advisory Council, the state Developmental Disabilities Planning and Advisory Council, and small focus groups composed of case managers and direct-care staff members who had participated in the pilot survey. The significance of the data to longitudinal assessment and planning was discussed following these presentations.
Pilot Sample Characteristics
Table 1 shows the demographic characteristics of respondents. The sample represented approximately equal numbers of men and women, and ages of participants closely approximated normal distribution around the mean age of 50 (standard deviation [SD] = 14.8). Twenty-five percent of the sample was rated as having “poor” or “fair” health, and 75%, as having “good” or “excellent” health. Thirty-one percent of the sample was rated as having “poor” or “fair” independence, and 69%, as having “good” or “excellent” independence. The majority of the sample indicated mental retardation as their primary disability (76%), were employed in sheltered workshops (72%), and lived in supported independent living arrangements, group homes, or institutions (87%).
Table 2 presents the average rating of limitation for each condition, the number who reported experiencing the condition, the estimated prevalence of the condition per 1,000 consumers, and the problem index and overall ranking for each of the secondary conditions. In addition, the percentage of consumers receiving treatment (i.e., experiencing limitation and having an Individual Plan objective) for each condition is reported. Reading across columns in Table 2, one can see that (a) for those in the pilot sample who experienced some limitation or need for support due to communication difficulties, the condition was associated with mild to moderate limitation; (b) 62 of the 119 consumers sampled experienced this condition as limiting; (c) this proportion may be used to estimate a prevalence rate of 573 consumers limited by the condition for every 1,000 consumers; (d) this was the condition associated with the greatest limitation for the most people; and (e) only 31% of consumers who were identified as limited by communication difficulties had an Individual Plan objective to address this secondary condition. Although some conditions, such as diabetes, may have relatively low prevalence rates and corresponding problem indices, the individuals experiencing them may be greatly limited, as indicated by the high average limitation rating.
Table 3 reports the same summary measurements for nine supplementary secondary conditions that were only rated for a subset of consumers who had a related mobility impairment (n = 48). Of this subset of nine, access problems were the most common and had the highest problem index, whereas experiences with abnormal spinal curvatures (e.g., scoliosis kyphosis, lordosis) caused the most severe limitation.
Table 4 presents the 11 most commonly reported conditions, with the results of one method for estimating relative treatment effectiveness. It shows the percentages of those persons receiving treatment for a condition distributed across the four levels of limitation severity estimated for each item. Although longitudinal assessment is necessary to validly investigate Individual Plan effectiveness (Stancliffe, Hayden, & Lakin, 1999), for the purposes of interpreting these cross-sectional pilot data, we have reasoned that an Individual Plan treatment for a condition is effective if the rating of the limitation is no limitation or mild/infrequent limitation. Conversely, if a consumer has an Individual Plan treatment for a condition that is reported as producing moderate or severe limitation, we have reasoned that the plan may not be effective. In addition, Table 4 shows results of two methods for estimating relative need for treatment. First, it shows the percentage of respondents who reported any limitation due to a condition but who do not have a relevant objective in their Individual Plan. Second, it shows the number of consumers experiencing moderate or significant limitation due to a condition but who do not have a relevant Individual Plan objective. This latter column suggests potential need for treatment.
Table 5 presents results from 15 participants who also completed the Reporter Reaction Survey. They rated the importance of health issues discussed on the surveillance instrument as very important on an 11-point Likert scale that had a negative anchor of 0 and a positive anchor of 10 (M = 9.4) and gave lower ratings to their own levels of preparedness for addressing health promotion and management of secondary conditions on a similar 11-point scale (M = 7.27). Reporters' feedback informed refinements of item descriptions as well as the emerging content for a behavioral risk and protective factors instrument being developed to accompany the surveillance instrument (Caruso, Traci, Seekins, Szalda-Petree, & Ravesloot, 2002).
In this report we have described the development and field testing of an instrument to assess the prevalence and extent of limitation due to secondary conditions among adults receiving services from community-based developmental disabilities service providers. Results suggest that adults in our pilot sample experienced the conditions included on the instrument. In addition, results demonstrate that the conditions can be ranked by various summary measurements. These rankings can be used to identify health issues of relative importance to individual and system planners. Finally, these results suggest that the health of individuals receiving or not receiving treatment for secondary conditions can be evaluated and, to some extent, the effectiveness of the treatment can be interpreted. Although these data were collected on a relatively small, select sample, the rankings of communication difficulties as the highest rated problem for all respondents, and access problems as the most significant problem for those with mobility impairments add to judgments of the instrument's content validity.
All 10 items presenting the most limitation for the most respondents involved significant behavioral or life-style aspects, including communication difficulties; problems with physical fitness and conditioning; persistence problems; weight problems; problems with personal hygiene and appearance, dental hygiene, fatigue, and depression; problems with mobility and sleep disturbances. The high ranking of items associated with lifestyle and behavioral issues is also similar for adults with mobility impairments. It appears that conditions that are typically treated medically may be more likely to receive treatment than those with behavioral, environmental, or life-style treatments. For example, 86% of those with gastrointestinal problems were reported to have Individual Plan treatment addressing that condition. Only 15% reporting limitations due to depression and 3% with limitations due to access barriers were reported to have Individual Plans addressing those conditions. Medical procedures are often better established than strategies for addressing lifestyle and environmental issues.
Data collected from health assessments such as these may have several practical applications. First, from a public health perspective, a state system may collect data about specific health conditions to identify secondary conditions that are either the most problematic for the most people or that are the most severe and costly. Such data would allow program policymakers, administrators, and program designers to develop programs and staff training to address those problems. A health surveillance system of periodic assessments could then be used to monitor the effects of such system-wide interventions across the population. Second, data about the existence of Individual Plan objectives for specific secondary conditions may be useful in monitoring the level of treatment for these conditions. Such data could also be used to identify patterns of treatment. For example, in program planning it may be useful to know the proportion of consumers who report significant limitations due to a secondary condition and the proportion who are receiving some systematic treatment or support targeting that condition. Further, it may be useful to identify Individual Plan objectives that coincide with low ratings of limitation due to secondary conditions because these may include effective secondary and tertiary prevention strategies for the condition.
In addition, identified strategies could be analyzed to assess their utility for prevention among other individuals. The lack of information about empirically tested prevention strategies may be reflected in the disturbing findings of Stancliffe et al. (1999). They investigated how effectively consumers' Individual Habilitation Plan (IHP) objectives reduced functional limitation as measured by the Inventory for Client and Advisory Planning (Bruininks, Weatherman, & Woodcock, 1986). They found that IHP objectives did not have a significant effect on functional limitation, regardless of the quality of such objectives. If empirically tested prevention strategies and resources were available for incorporation into IHP objectives along with the quality indicators outlined by Stancliffe et al., then longitudinal assessments of secondary conditions should reveal a reduction in the amount of limitation experienced by individuals.
Finally, these data and the system for collecting them might also include measures of risk factors and medical service utilization associated with secondary conditions (e.g., Eyman, Chaney, Givens, Lopez, & Lee, 1986; Miller & Eyman, 1978). These data could help in setting priorities for system improvement and evaluating the cost-effectiveness of services and supports. Research is being conducted to examine these latter issues (Caruso et al., 2002).
For several reasons, the results of this preliminary report must be interpreted with caution. First, these data come from a relatively small number of individuals living in one state who were selected by service providers. This significantly limits the external validity of the data and points to the need for similar data from a larger sample. Second, the data were reported by representatives of the primary consumer. Although respondents were selected based on their familiarity with each consumer and were directed to confer with the consumer while completing the instrument, their reporting may have had systematic biases that threaten the data validity (Goode & Hogg, 1994) and was contrary to the principles of participatory action research (Fenton, Batavia, & Roody, 1993; Seekins, Mathews, Fawcett, Jones, & Budde, 1988; Whyte, 1991). In participatory action research, there are no hard and fast rules for determining which constituents and how many should participate, how much participation is enough, or how to best evaluate constituents' impact. We recruited individuals with a broad range of perspectives who would be affected by the process and results. This application of participatory action research shaped the content and procedures of the surveillance instrument and afforded important perspectives on the interpretation of these data.
Perhaps most important, the dynamic interplay between researchers and potential adopters led to the inclusion of a measure for assessing which conditions were being addressed by Individualized Plan objectives as a way to estimate the effectiveness of services delivered through the Individualized Plan process. Applied components of participatory action research shaped the content and format of the instrument in ways that facilitated the implementation of the pilot survey and improved the educational benefits of participating. These contributions led to high ratings on the Reporter Reaction Survey. Direct participation of consumers in the Advisory Council discussions and the survey completion process continues to be a challenge for this project and has been reported by other researchers applying participatory action research with populations who have disabilities (e.g., Balcazar, Keys, Kaplan, & Suarez-Balcazar, 1998; Balcazar, Mathews, Francisco, Fawcett, & Seekins, 1994; Sample, 1996; Seekins, 1995; Stevens & Folchman, 1998).
In addition to the surveillance protocol described in this report, in our current research we are attempting to directly assess consumer ratings of secondary conditions using computer presentations of secondary conditions as the survey medium. A future reliability analysis comparing surrogate ratings with consumer ratings will aid in the identification of reporter biases, which could inform staff and consumer training. The direct input of consumers in health surveillance and in Advisory Panel discussions (see Kavin, Sligar, & Oellette, 1997) will be critical to the application of participatory action research within a Behavioral Health Promotion and Monitoring System.
NOTE: Preparation of this manuscript was supported by grants from the Office on Disability and Health, Center for Environmental Health of the Centers for Disease Control and Prevention (R04/CCR814162–02), by the Montana Developmental Disabilities Planning and Advisory Council (DDPAC), and by the National Institute on Disability and Rehabilitation Research (NIDRR). The opinions expressed are those of the authors and do not necessarily reflect the position of any agency. The authors acknowledge the early contributions of Phil Mattheis to the development of this project. We also thank Greg Olsen, Deborah Swingley, and the Montana DDPAC for their consistent support and Greg Ragee, Ken Brown, and the staff of Opportunities Resources, Inc., without whose help this project would not have been possible. Finally, we thank Lisa Brennan and Diana Spas for their careful reading of manuscript drafts.
Meg Ann Traci ( email@example.com) PhD, Project Director; Tom Seekins, PhD, Director; Ann Szalda-Petree, PhD, Research Consultant; and Craig Ravesloot, PhD, Program Director, The University of Montana Rural Institute, 52 Corbin Hall, The University of Montana, Missoula, MT 59812. Requests for reprints should be sent to the first author