| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
Correspondence: Address correspondence to Denise Burnette, Columbia University School of Social Work, 622 West 113th Street, New York, NY 10025. E-mail: jdb5{at}columbia.edu
 |
Abstract |
|---|
 TOP Abstract Background and PurposeThe Delphi MethodDelphi Studies in GerontologyMethodsResultsDiscussionReferences |
|---|
Key Words: Research agenda • Research priorities • Gerontological social work • Intervention research • Delphi method
Aging issues are prominent in the research agendas of the National Institutes of Health (NIH). In close collaboration with the National Advisory Council on Aging and other public and private organizations, the National Institute on Aging (NIA) marked its 25th anniversary in 1999 by identifying four 5-year research priorities, each with several subgoals (see NIA, 2000). Agendas of other institutes, such as the National Eye Institute, the National Cancer Institute, and the National Institute for Nursing Research, also include explicit priorities concerning other adults. Some topics are discipline specific; others are crosscutting. An important example of the latter is the recent report by the National Research Council (NRC) on the NIH Committee on Future Directions for Behavioral and Social Science Research (Singer & Ryff, 2001). The report lists 10 priorities that span institute boundaries, thereby underscoring the significance of social and behavioral science research for multiple disease outcomes and health promotion. Relevant aspects of the NIA and NRC agendas are discussed more fully in the paragraphs that follow vis-à-vis the social work research priorities identified herein.
As chair of the National Institute of Mental Health (NIMH) Task Force on Social Work Research (NIMH Task Force, 1991), Austin (1998) called for a strategic approach to social work research based on specific domains of expertise. Specifically, he recommended that each domain assess the state of its own knowledge, identify critical knowledge gaps, and establish research plans and priorities to address those gaps. Despite having a vital, specialized role in the development and application of knowledge in gerontology, social workers have yet to conduct such an assessment.
Gerontological social workers, for example, provide extensive individual, group, family, and community services to older adults and their families in a wide range of settings (Schneider, Kropf, & Kisor, 2000). They are highly active in the formulation, application, and evaluation of public policy, and they bring an essential psychosocial perspective to interdisciplinary teams in medical, psychiatric, and long-term-care settings. Expert knowledge about eligibility, availability, quality, and barriers to health and social services is a hallmark of professional social work, and social workers' frontline experience in service delivery to older adults affords them a distinctive perspective on the knowledge needs and practice competencies required for grounded research (Schneider, 1995). However, social work's contribution to applied research in aging has been incommensurate with its integral role in intradisciplinary and interdisciplinary practice (Morrow-Howell & Burnette, 2001; Scharlach, Damron-Rodriguez, Robinson, & Feldman, 2000).
The John A. Hartford Foundation of New York City is currently sponsoring a national, multipronged initiative: Strengthening Geriatric Social Work. A major aim of this program is to increase the number of highly skilled gerontological social work researchers through its Faculty Scholar and Doctoral Fellows Programs (Berkman, Silverstone, Simmons, Volland, & Howe, 2000; Lubben & Harootyan, 2002). Establishing a prioritized research agenda will help focus, organize, and synthesize the work of these and future social work researchers in aging. This article describes the development of such an agenda and identifies high-priority research areas whose investigation is expected to yield social-work-related knowledge for improving the lives of older persons and their families and to complement and enhance the research agendas of other disciplines in aging.
|
Background and Purpose |
|---|
|
TOPAbstractBackground and PurposeThe Delphi MethodDelphi Studies in GerontologyMethodsResultsDiscussionReferences |
|---|
|
The Delphi Method |
|---|
|
TOPAbstractBackground and PurposeThe Delphi MethodDelphi Studies in GerontologyMethodsResultsDiscussionReferences |
|---|
Adler and Ziglio (1996) discussed the pros and cons of Delphi techniques. Advantages are obviating the need for geographic proximity, focusing attention on a specific topic, minimizing group pressure on individual opinions (challenged by Woudenberg, 1991), allowing panelists equal opportunity for influence, providing precise documentation of data used, and avoiding some of the pitfalls of public group discussion. Drawbacks include use of subjective judgments rather than objective data, potential difficulty determining who qualifies as an expert, less scientific rigor than other methods, and possible heavy time expenditures to complete questionnaires. Others have also questioned the accuracy of forecasts produced by such methods (Dietz, 1987). Van Dijk (1990) concluded from a study of accuracy of large-group techniques, for instance, that different techniques may yield different responses, and that a mix of individual and group interviews plus a mailed questionnaire may yield the most accurate results.
|
Delphi Studies in Gerontology |
|---|
|
TOPAbstractBackground and PurposeThe Delphi MethodDelphi Studies in GerontologyMethodsResultsDiscussionReferences |
|---|
Delphi panels on practice have identified key social, environmental, and economic aspects of geriatric assessment (Saltz, Schaefer, & Weinreich, 1998) and core decision-making features of adult protective services workers (Bergeron, 1999). Others have identified the primary needs of rural elders (Powell & Thorson, 1989); developed taxonomies of elder abuse (Hudson, 1991); devised typologies of practitioner problems in health services to elders (White & Pearlman, 1986); and established guidelines for a graded license for older drivers (Malfetti & Winter, 1990). Finally, Delphi techniques have been used to develop research and policy priorities in specific areas of gerontology. Austin, Durham, Hedrick, Pearlman, and White (1984) established research topics to meet health needs of aging veterans, whereas Cassel (1988) assessed guidelines to protect human subjects in long-term-care research. In addition, in the epilogue of a book on social policy issues in long-term care, Binstock, Cluff-Leighton, and von-Mering (1996) derived a series of prognostications about the future of long-term care in the United States.
In considering our study objectives, the potential pros and cons of Delphi techniques, and the productive use of these methods in gerontology, we selected this as the most appropriate method to draw on the expertise of gerontological social work researchers to develop a set of research priorities for the field.
|
Methods |
|---|
|
TOPAbstractBackground and PurposeThe Delphi MethodDelphi Studies in GerontologyMethodsResultsDiscussionReferences |
|---|
Procedures
We designed a Web-based platform to collect data via the Internet, and then we pilot tested the instrument and Web site with three senior geriatric social work researchers. This process resulted in minor revisions in wording to clarify instructions to the panelists. Figure 1 depicts the study process. Briefly, in Round 1, panelists were asked to list up to 5 high-priority research topics for gerontological social work and to provide a rationale for each choice. The objective at this stage was to develop an exhaustive enumeration of research topics. A total of 181 topics was submitted. The authors and two additional gerontological social work researchers with experience in coding open-ended data then independently coded topics into nonoverlapping categories. In order to ensure that all enumerated topics were represented in the condensed list, raters were instructed to err in favor of too many rather than too few categories. Resolution of minor discrepancies in coding yielded a refined list of 49 discrete research topics.
|
Descriptive statistics (mean, median, standard deviation, range, and interquartile range) were calculated for each topic along with the number of times each was identified as one of five top priorities. As a way to further condense the data and move toward consensus, 14 topics whose lower bound of the interquartile range dropped below 3 were eliminated and the remaining 35 topics were conceptually reorganized into 12 broad subheadings. In the third and final round, panelists reviewed this reorganized list and descriptive statistics for each topic. Taking into account this information, they were again instructed to rate each (1 = highest priority to 5 = lowest priority) and to generate a rank-ordered list of five top priorities. They were also invited to register and explain divergent opinions from the majority consensus at this final stage, although few did so.
As a way to facilitate comparisons among topics, numeric priority ratings were next grouped by dividing the sum of the scale values by the number of ratings, a procedure that treats nominal scales as interval data. The value 2.5 was regarded as a neutral point, with boundaries distributed accordingly to avoid polarization of scores (Linstone & Turoff, 1979). Mean priority ratings were thus grouped as follows: highest (1.0–1.49), high (1.5–2.49), moderate (2.5–3.49), or low (3.5–4.49). For example, the mean score for "developing psychosocial outcomes mea-sures" was 2.11 (see Table 1), meaning that panelists judged it to be a high-priority research topic.
|
Given the close clustering of scores with a deviation of less than 1 point on a 5-point Likert scale, a quartile deviation of 1.00 conventionally indicates group consensus. When the distribution of scores in this project was examined, topics with a quartile deviation less than or equal to.60 were deemed high consensus, those greater than.60 and less than or equal to 1.00 were deemed moderate consensus, and those greater than 1.00 were deemed low consensus (Faherty, 1979; Raskin, 1994). Following on the previous example, the consensus score for "developing psychosocial outcomes measures" was 0.875 [(2.75 - 1)/2], indicating that, by the third and final round of data collection, panelists reached only moderate consensus on their ratings of this topic.
Finally, the two evaluative criteria used to determine the relative importance of topics—priority (highest, high, moderate, low, or lowest) and consensus (high, moderate, or low) ratings—yielded a matrix of 15 possible combinations of endorsement levels for each topic.
|
Results |
|---|
|
TOPAbstractBackground and PurposeThe Delphi MethodDelphi Studies in GerontologyMethodsResultsDiscussionReferences |
|---|
Table 2 depicts the 35 highest ranked topics, grouped by substantive area, which panelists rated in Round 3. In addition to the proportion of panelists who endorsed the topic as one of five top priorities, the table presents consensus and priority levels for each topic.
|
|
|
Discussion |
|---|
|
TOPAbstractBackground and PurposeThe Delphi MethodDelphi Studies in GerontologyMethodsResultsDiscussionReferences |
|---|
These 16 priorities are highly compatible with those of other national research agendas on aging, yet they also underscore social work specific expertise. As seen in Table 3, they overlap and complement many items in the NRC and NIA agendas. These commonalties highlight the interdisciplinary nature of gerontology and the shared goals of its contributing disciplines. They also signify specific contributions that skilled social work researchers can make in developing and translating knowledge in these critical areas. The NRC lists health inequalities and NIA identifies health disparities, for example, whereas the social work agenda includes three items on mechanisms to improve outcomes for a diverse older population: determining needs of ethnically diverse elders; developing and testing ethnically sensitive assessment strategies and interventions, and enhancing service use by ethnic elders.
In a study of funding for education and training of geriatric personnel, Dawson and Santos (2000) noted "national shortages of geriatric-care personnel in the medical, mental health, and social service professions who are prepared to provide effective services for the nation's older population" (p. 1; also see Klein, 1996). The NRC and NIA agendas address the need to attract and train high-quality researchers in aging, whereas the social work agenda identifies professional workforce needs, that is, recruiting and training of highly skilled professionals to serve an aging population—no doubt a salient issue for this predominantly academic panel. The U.S. Department of Labor (2000) projects at least a 36% increase in the number of social workers by 2008, with the highest demands in aging and health. Nevertheless, only 16% of baccalaureate social workers and 4% of masters prepared social workers are employed specifically in services to the aged (Gibelman & Schervish, 1997; also see Barth, 2001). The aforementioned Hartford Foundation initiative also encompasses the SAGE-Social Work program, housed at the Council on Social Work Education, which is addressing issues of workforce preparation (Council on Social Work Education, 2000).
Four of the top 16 priorities concern intervention research, a topic also shared with the NRC and NIA agendas. The NRC specifically lists intervention research, and NIA implies this priority in terms of improved health and mental health outcomes. The highest priority for social work research was developing and testing psychosocial interventions across specific populations and conditions, such as different racial or ethnic groups, varying cognitive abilities, and the continuum of care environments. These interventions, which range from care management and discharge planning, for example, to caregiver psychoeducation groups and intergenerational family therapy, are central functions of gerontological social work.
There is evidence that social work interventions with older adults and their families are effective for a myriad of problems (see Grenier & Gorey, 1998). However, in a review of social work research published during the past 5 years, Morrow-Howell and Burnette (2001) found few rigorous outcome and effectiveness studies. The panelists' clear call for more refined and rigorous intervention studies is thus timely. It is also highly consistent with the profession's current efforts to establish practice guidelines and competencies and evidence-based practice modalities (Council on Social Work Education, 2000; Gambrill, 1999; Rosen & Proctor, in press), and the NIH's emphases on practice research and effectiveness research (NIH, 1999).
Two topics that attained high priority and high consensus ratings concern housing or living arrangements and care settings. Ensuring safe, appropriate, affordable housing is an essential social work function, the urgency and complexity of which may grow as housing options and preferences expand with population aging. The identification of transitions across care settings as a unique topic is of particular note because older adults often move within and among care settings as their capacities and resources fluctuate. These transitions may also become more common and more complex as new levels and types of care evolve with changing needs, care philosophies, and policies. Current examples are the rapid growth of palliative care and assisted living settings. Social workers are well situated to examine the risks and challenges associated with these transitions.
Four high-priority research topics are in the area of service delivery to older adults and their families. Assessing the need for, ensuring access to, and managing the delivery of health, mental health, and social services vis-à-vis informal support networks is a hallmark of social work practice. These, too, are vital and challenging functions, as many older adults have concurrent needs requiring services across multiple sectors of a fragmented service delivery system (Morrow-Howell, Proctor, Rubin, Li, & Thompson, 2000). This category also identifies service needs and service use by ethnic elders, reflecting well-established patterns of low use despite high levels of chronic medical conditions, functional impairment, and economic hardship that suggest high need (Mui & Burnette, 1994). Knowledge is needed to inform the development of feasible, effective guidelines and policies to improve service planning, delivery, coordination, and evaluation.
Systems theories fundamentally underpin social work practice philosophy, meaning that assessment and intervention strategies typically include clients' families and often other members of their social network. Social work researchers have contributed substantially to the knowledge base on family caregiving. Two high priority, high-consensus topics concern family caregiving, both of which underscore the need for targeted interventions for specific caregiver populations, such as ethnic minority, older, male, and grandparent caregivers.
High-priority health and mental health topics are improved knowledge of and services for psychosocial aspects of health-related quality of life, especially associated with chronic conditions and disabilities, and improved detection and treatment of depression. These health-related topics are commensurate with research objectives of the Centers for Medicare and Medicaid Service's (formerly the Health Care Financing Administration) research and demonstration program, which include health access, outcomes, quality and performance, and health services for vulnerable populations (Health Care Financing Administration, 2002).
The specification of depression from among the numerous mental health conditions that affect older adults is probably attributable to its high prevalence (and low detection) in community and institutional settings served by social workers, its high comorbid occurrence with physical and functional problems, and its debilitating effects on quality of life. As many social workers practice in health and mental health settings, their experience and expertise on these and related problems make them particularly valuable investigators and partners in applied research on these topics.
Three topics attained high priority but low consensus (long-term-care policy; case finding and interventions in specific settings, and end-of-life care), meaning that panelists judged these items to be important but the deviations of their ratings from their mean scores did not decrease between Rounds 2 and 3. Indeed, long-term-care policy and end-of-life care ranked second and ninth, respectively, in Round 2 (see Table 2). A recent report by Morrow-Howell and Burnette (2001) that few social workers appear to be conducting policy research in aging supports the low consensus on long-term-care policy. The dearth of social policy topics in the NRC and the NIA agendas is also conspicuous, given the profound impact of social policies on the lives of older people and their families. Perhaps social policy is implied in other topics, or perhaps the development, implementation, and evaluation of social policy is viewed as an extension or outcome of rather than a component of research. The Soros Foundation's Project on Death in America has greatly increased attention to end-of-life care in social work (Christ & Sormanti, 1999). This topic may have achieved high priority because panelists recognize its importance to social work, but low consensus because they saw it as perhaps more specialized than other, broader topics or as a domain of health care that is legitimately shared by other professions.
Conclusions
The research priorities identified in this national Delphi study reflect the judgments of an expert panel of 46 gerontological social work researchers who play major roles in the academic gerontology community. The aforementioned companion study with a national expert panel of nominated gerontological social work practitioners is expected to improve our understanding of the knowledge needs of providers who work directly with older adults and their families. The extent and manner in which priorities identified by these two groups overlap and diverge can inform our efforts to close existing gaps between research and practice.
Several limitations to this study are noted. First, the expert panelists were drawn from a comprehensive national listserve intended to tap a wide range of gerontological social workers in academic and research settings. However, researchers who were not involved in the Hartford initiative were not included. It is difficult to estimate how many qualified panelists were thus excluded, as no exhaustive list of such individuals exists, but their numbers are expected to be quite small compared with those on the panel. Second, there are numerous methods for conducting Delphi studies. We relied on prior studies, especially those on gerontology or setting research agendas, for our design. This strategy worked well for our purposes, although it afforded us few explicit procedural guidelines. Following this literature, for example, we permitted maximum flexibility in panelists' responses. As a result, we concluded with a mix of basic and applied research topics expressed at varying levels of abstraction. Rather than being a shortcoming, this may genuinely reflect the research needs in an applied profession. Third, we explicitly urged panelists to base their judgments on social work research needs in the broad field of aging, but we could not determine how personal commitments and preferences may have influenced topic selection or priority ratings. The range and distribution of topics enumerated and the progressive winnowing of the list suggest that study procedures may have controlled for this concern to some extent, as a personal interest would have to be endorsed by a sufficient number of panelists to remain practicable.
In conclusion, although they are not a blueprint for research, findings from this national Delphi study of social work research priorities in aging are expected to provide valuable information for social workers as they respond to the growing mandate for a prioritized research agenda. The 16 top-priority topics represent areas of critical knowledge need and suggest specific foci for research. Such an agenda must fit with and advance a profession's philosophical underpinnings, knowledge bases, and practice commitments. These topics, which are more proximal and specific than those in other research agendas and which focus largely on ensuring the development and delivery of effective services to older adults and their families, fit these criteria well. The study findings are also expected to enhance and refine social work's collaboration with other disciplines on topics of mutual interest and expertise.
|
Footnotes |
|---|
1 Columbia University School of Social Work, New York, NY.
2 George Warren Brown School of Social Work, Washington University, St. Louis, MO.
3 University of Houston Graduate School of Social Work, Houston, TX.
Decision Editor: Laurence G. Branch, PhD
Received for publication November 15, 2001. Accepted for publication March 15, 2002.
|
References |
|---|
|
TOPAbstractBackground and PurposeThe Delphi MethodDelphi Studies in GerontologyMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  D. T. Levy, E. A. Mumford, K. M. Cummings, E. A. Gilpin, G. Giovino, A. Hyland, D. Sweanor, and K. E. Warner The Relative Risks of a Low-Nitrosamine Smokeless Tobacco Product Compared with Smoking Cigarettes: Estimates of a Panel of Experts Cancer Epidemiol. Biomarkers Prev., December 1, 2004; 13(12): 2035 - 2042. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
