Seizing Interdisciplinary Opportunities in the Changing Landscape of Health and Aging: A Social Work Perspective

Abstract

Purpose of the Study: This paper is a revision of the Kent Award Lecture given at the Annual Meeting of the Gerontological Society of America held in New Orleans, Louisiana, in November, 2010. Design and Methods: This paper looks at the evolution in geriatric social work assessment and outcomes research and concludes with observations of the changing landscape in health and aging. Results: Since the 1960s, the policies and the context of health care delivery have changed many times as have geriatric health screening and assessment of patients in need of social health care services. Research on social–behavioral and environmental factors critical in measurement of outcomes of health care has progressed significantly as theories of care and the research technologies that allow us to study these factors have become more sophisticated. Implications: Researchers from multiple disciplines need to study the questions which can build the evidence necessary for empirically supported social policy direction. Opportunities in interdisciplinary geriatric assessment and measurement of outcomes, which are presented to researchers today, are highlighted.

Donald Kent believed, and practiced his beliefs, in the importance of linking research on aging to practice and policy. He wrote eloquently on policy issues and the need for researchers from multiple disciplines to study the questions that can build the evidence necessary for empirically supported social policy direction (Kastenbaum & Sherwood, 1972; Kent, 1964). It is my hope that this paper will bring renewed meaning to his words.

Since the 1960s, when my academic trajectory began, the policies and the context of health care delivery have changed many times as have geriatric health screening and assessment of patients in need of social work services in the hospital and other health care settings. In addition, research on the social–behavioral and environmental factors critical in measurement of outcomes of health care has progressed significantly as theories of care and the research technologies that allow us to study these factors have become more sophisticated.

Research by social workers has been conducted in this arena for 40 years, with practice-based research efforts linked to the ever-changing realities of our health policies, care systems, patients, and providers. This paper looks at the evolution in geriatric social work assessment and outcomes research from my personal perspective and concludes with observations of the changing landscape in health and aging. Opportunities in interdisciplinary geriatric assessment and measurement of outcomes, which are presented to researchers today, are highlighted.

The 1960s

An appropriate place to begin is the 1960s, a time when the number of social workers in hospitals was growing exponentially in response to an increasing need for the services they offered. John F. Kennedy, who exuded youth and energy, was elected President in 1960. The country was elated when we walked on the moon, and there were also impassioned public disagreements over the appropriateness of our engagement in the Vietnam War. And when President Kennedy was assassinated in 1963, as horrible as that was, the transition to Lyndon Johnson was relatively uneventful (Ebert, 1986). However, the mid-60s was a time of the civil rights movement and growing public calls for creating a society that offered equity for all, particularly minorities, the poor, and the elderly patients. President Johnson responded to this growing public concern with his “unconditional war on poverty,” with access to health care being a major initiative. In 1965, the Older Americans Act was passed, establishing the Administration on Aging and state agencies to address, in part, the social service needs of older people. In 1966, Medicare and Medicaid legislation was passed, giving support to teaching hospitals with an emphasis on the health of older adults and of the poor. President Johnson’s Great Society social programs, which focused on community action and education, became inseparable from medical care (Starr, 1982). Social work began to be viewed as a significant part of health service delivery in the academic medical center and as a necessity for identifying and dealing with inadequacies related to social, behavioral, and environmental issues, which affect patient’s health care outcomes.

With the advent of Medicare and Medicaid, assessment of older hospitalized adults who were at risk for problems, which could negatively affect their health care outcomes, became extremely critical. Hospital floors were increasingly crowded with older patients whose discharges were being delayed because of the complexity of their health care and home care needs. Social work services were vital for such patients, but a deterrent to service delivery was the hospital social work case-finding system. Traditionally, this was a system that was dependent on referrals from doctors and nurses. Those patients whom they believed needed social work service were often referred during the last days of their long lengths of stay when their discharge planning was in crisis (Rehr & Gordon, 1967). The average length of stay of those patients who were referred to social workers was 39 days compared with 18 days for other older patients (Berkman, Rehr, Siegal, Paneth, & Pomrinse, 1971). In addition, there were concerns that because of the referral system, social work might be missing the most vulnerable of the older patients, specifically minorities and the poor (Gordon & Rehr, 1969). Hospitals needed a means of earlier valid identification and assessment of whom, among the increasing numbers of elderly vulnerable inpatients, would benefit from receiving social work services earlier in their hospitalizations.

Thus, a major question for social work research was to identify characteristics of elderly hospitalized patients who had social–behavioral problems (e.g., difficulties in family relationships or in ability to follow posthospital care plans) and/or environmental problems (such as unsuitable housing) that might negatively affect the outcome of their health care and who could be helped by social workers. This study had not been done before, and in 1967, supported by a grant from Health, Education, and Welfare, Berkman and colleagues (1971) began recording daily admission data on 5,280 older patients admitted to Mt. Sinai Hospital in New York City, with a focus on the 520 patients referred to social work services . These data were basically sociodemographic and found that 12 factors accounted for only 12% of the variance in who were referred and who were not. However, the most significant variable was extended lengths of stay, and the major question became whether there were social, psychological, and environmental factors, which could be identified early in a patient’s length of stay, which were associated with long hospitalizations.

The 1970s

By 1970, the halcyon days of available resources for health care service delivery were waning. The Vietnam War had drained federal monies and in the mid-70s, 10 years after Medicare and Medicaid came to be, there was an unprecedented rise in health care expenditures (Blendon, 1986). With a weakened economy, the war on poverty disappeared. President Richard Nixon spoke about a crisis in costs of health care (Starr, 1982). Medical care was under scrutiny, and regulation was the “only game in town.” Other deficiencies were also reported as the United States “had higher infant mortality rates and lower life expectancies than most Europeans” (Starr, 1982, p. 382).

By 1972, Medicare and the Joint Commission for Accreditation of Hospitals (JCAH) both required utilization review of the necessity of admission, duration of stay, readmissions, and professional services utilized (United States. Congress, Senate, 1972). The major health care issue was costly lengths of stay. Social work researchers believed that patient- and family-related social, behavioral, and environmental factors recorded in social work records would predict patients at high risk for extended lengths of stay and who needed social services. And so, in the early 70s, a series of studies to test those assumptions were conducted (Berkman & Rehr, 1973, 1974; Rehr & Berkman, 1973). These social work–specific research efforts identified some important social–behavioral factors. Stress related to becoming a dependent patient, psychological issues in adjusting to illness and hospitalization, and problematic social relationships were important factors in identifying high-risk patients. But still, these factors explained very little of the variance in who did, or did not, need social work services.

By the late 70s, many health professionals were beginning to try to explain the reasons for extended lengths of hospital stay and early preventable readmissions (Rothberg, Pinto, & Gertman, 1980; Sang-O, Lyons, & Payne, 1979; Sherman & Flatley, 1980; Westphal, Frazier, & Clinton Miller, 1979). Social work researchers realized that the perspective of a single discipline was not fruitful. What was needed were data from medicine and nursing. However, the practice in the hospital at that time was not collaborative. The professions were not there yet. And so, utilizing a broader database, through medical and social work records, social workers identified that severity and chronicity of illness were critical factors and were, of course, related to extended length of stay, complex patient posthospital needs, and the need for social work help (Berkman, Rehr, & Rosenberg, 1980). But these efforts did not account for a great deal of the variance in differentiating which elders were “at risk” and which were not. Instead of the originally reported 12%, these factors accounted for 28% of the variance, but there was still a high false-positive rate in the screening process. Two thirds of the time, our screening was inaccurate, and social workers did unnecessary time-consuming assessments on patients and families who did not need social work help.

The 1980s

By the 1980s, Ronald Reagan was President, and cost control was an even greater driving force. So the country enacted a number of cost containment measures, one of which, in 1983, was a prospective payment system through which hospitals received a fixed payment on the basis of a medical diagnosis. Diagnoses were grouped in categories called Diagnostic-Related Groups (DRGs). The diagnostic specificity (e.g., Diseases and Disorders of the Respiratory System, Diseases and Disorders of the Circulatory System) was recognized as a critical variable in determining reimbursement rates and in reviewing costly hospitalizations and readmissions. Specific disease categories were significant as biomedical factors, which could not be ignored in our thinking about length of stay and discharge planning.

In addition to meeting government mandates to control rising costs, quality assurance mechanisms were being implemented to address issues around quality of care. In 1980, the JCAH standards on quality assurance mandated that discharge planning should be initiated through early determination of need, such as use of early screening for high social risk patients (JCAH, 1980). By 1982, most hospitals had formed utilization review committees to monitor lengths of stay, readmissions, and quality of services delivered. These committees were frequently composed of hospital-based professionals such as nurses, physicians, and social workers who had similar research interests and skills and had been picked by their departments to collaborate on utilization reviews. There was a uniform interdisciplinary objective—to address the mandates for reducing length of stay and readmissions while protecting quality of care for older patients.

It was also during this time, in the 80s, that I joined the social work department at Massachusetts General Hospital (MGH), an affiliate of Harvard Medical School, and I became one of the founders of the MGH Institute of Health Professions. The philosophy of the Institute’s education for their students who were to be nonphysician health care providers (i.e., dieticians, nurses, physical therapists, and social workers) was interdisciplinary in that they believed that certain key knowledge elements should be taught across the board with disciplines learning together. And it was also during this same time period that I joined the interdisciplinary practice and research efforts, which were being conducted by members of the Harvard Medical School’s Division on Aging (Berkman, Campion, Swaggerty, & Goldman, 1983; Campion, Jette, & Berkman, 1983).

Concomitantly in the mid-80s, in their efforts to establish valid floor staffing patterns based on DRGs and disease-specific patient needs, nursing at MGH began to use the computerized Medicus Nursing Productivity and Quality Patient Care Classification (NPAC) data system, which was very sophisticated for the times. Nursing and social work researchers realized that multidisciplinary data inputs were essential to the study questions they were pursuing, and they believed that in order to improve the measurement of health care outcomes, disease-specific studies, as were available through the DRG classification system, were necessary. Now, new research avenues opened. Collaboration began between social work and nursing colleagues who started to study standardized functional data available on the NPAC system as well as additional social work and medical data. Studies of assessment and outcomes of discharge planning for elderly cardiac patients, a high-risk group in need of multiple services for posthospital care planning, were undertaken. The assessment data derived from one such study reduced the false-positive screening rate of cardiac patients needing social services from 65% to 26%—a very important step as social work staff time was becoming a precious resource (Berkman, Millar, Holmes, & Bonander, 1990). Again in collaboration among social work, medicine, and nursing, research focused on the rehospitalization of elders with cardiac disease reported that the rapidity of physical and functional deterioration was key assessment indicators predicting early recurrent rehospitalization (Berkman, Millar, Holmes, & Bonander, 1992).

And so, by the end of the 80s, the research perspective at MGH had now clearly become interdisciplinary in terms of education and patient care as well as multidisciplinary in relation to the data systems needed to answer significant patient care outcome questions. This model was biopsychosocial. In brief, the interdisciplinary contributions in the 80s were efforts that created a very exciting and fruitful research environment.

The 1990s

In the 1990s, we saw the dissolution of the Soviet Union, and President Bill Clinton was a dominant political figure. The United States experienced its longest period of economic expansion, and we began to have widespread use of computers and the Internet. The early research work of the 60s set in motion a research direction that basically remained the same for over thirty years. What changed by the 1990s was the health care policy and practice context of the research, the technological capabilities, and the firm conviction held by many researchers of the necessity of interdisciplinary collaboration in assessment and outcome research in clinical health. The fee-for-service payment system began to be under scrutiny, and a managed care capitation payment system was initiated (Shortell, Gillies, & Devers, 1995). And the use of managed care organizations for the administration of medical benefits was a growing trend (White, Simmons, & Bixby, 1993).

Prior to this time, our health care system had been based on a paradigm of unpredictable acute disease. In the 1990s, increasing numbers of elderly patients were living longer with multiple chronic health problems. In the new health care paradigm characterized by chronic illness, there was awareness that chronic illnesses and their progression are determined by multiple factors, such as an individual’s physical, social, and psychological status as well as genetic endowments, cultural backgrounds, and health care accessibility issues (Berkman, 1996). The focus of care became primary care with an emphasis on disease prevention and health promotion (Pawlson, 1994).

It was recognized that positive health care outcomes could only be effectively achieved with collaborative practice among multiple disciplines. There was a surge of interest in new models of hospital-based short-term care as well as models of long-term care, home care, and community care. Short-term collaborative practice teams were effectively used by physicians and nurses in hospital intensive care units. And the substantial evidence of the benefits of collaborative care in improving depression outcomes in primary care was reported (Richards et al., 2006). The accepted value of collaborative care in primary care was also clearly evidenced in the 1997 National Institute of Mental Health call for proposals for demonstration projects, which would prevent suicide among elderly patients in primary care, a trial which was to test a “collaborative care” model involving nurses and social workers with physicians in primary care practices working together to better manage chronic conditions and improve identification of older patients in need of treatment for depression (National Institutes of Health, 1997).

A prime example of collaborative long-term care beginning in the 90s was the Program of All-Inclusive Care for the Elderly (PACE). Originally authorized in the Omnibus Reconciliation Act of 1986 as a demonstration program, in 1997, PACE became a regular part of Medicare targeting individuals who meet Medicaid nursing home eligibility criteria. Through the use of a multidisciplinary team approach, PACE integrates social and medical services primarily through a combination of adult day health care and home care (Eng, Pedulla, Eleazer, McCann, & Fox, 1997). And another excellent example of collaborative long-term care was the Social/Health Maintenance Organization (S/HMOII) demonstration, implemented in 1996 in Nevada. It presented a unique coordination approach for collaborative care services, such as case management, personal attendant care, transportation, day care, and social services (Kane & Homyak, 2003; Kane, Homyak, Bershadsky, Lum, & Siadaty, 2003; Newcomer, Harrington, & Kane, 2002).

Educational multidisciplinary health care training projects were also introduced in the 90s. For example, in 1995, the John A. Hartford Foundation (JAHF) funded its Geriatric Interdisciplinary Team Training Program with the understanding that if health care professions were to continue to follow strict disciplinary lines, they would leave students unprepared for implementing patient care plans in an interdisciplinary setting (JAHF, 2007). However, the issue not yet addressed was that we would never achieve validity in our health care outcomes research without the input of standardized data from the perspective of multiple disciplines.

The 2000s

In this first decade of the 21^st century, we have experienced serious challenges as well as historic milestones. There have been horrific events; 9/11, the wars in Iraq and Afghanistan, Hurricane Katrina, and a global recession. But our first African American President was elected, and the continuing yearly debates (for almost 50 years) regarding health care reform resulted on March 23, 2010, with President Obama signing into law the Patient Protection and Affordable Care Act (PPACA; United States. Congress, Senate, 2010).

The practices of the 90s have had a significant impact on health care practice in the first ten years of the century and have influenced many of the provisions in the PPACA. The landscape of delivery of care to older adults has been evolving rapidly with the majority of patients receiving their medical care in ambulatory primary care settings (Westphall, Mold, & Fagnan, 2007). Disease prevention and health promotion are receiving more recognition with older adults because research has shown that adults who engage in evidence-based self-management programs can remain healthy longer than those who do not (Tilly, 2010).

Considerable efforts in the 1990s addressed quality and costs in Medicare, and teams became an important feature of health care delivery in acute and long-term care and in primary care. In the PPACA, interdisciplinary health teams are envisioned as a means to enhance quality while lowering health care costs. For example, starting this year, there is a new division within the Centers for Medicare and Medicaid Services known as the Center for Medicare Innovations, which is charged with establishing innovative methods of furthering the quality/cost initiative (Segal, 2010). One of the pilot programs being tested is “medical homes,” with its concept of a personal physician leading a team of practitioners and allied health professionals who are responsible for the health care of patients with chronic illnesses. In this model, community-based “health teams” are to be established as integrated primary care practices within community areas served by the hospital.

Evidence-based practice was also a growing movement in the 90s and is now being utilized in many community-based organizations serving older adults (Whitelaw, 2010). Valid outcome measures are critical to the development of evidence-based interventions as well as to the evaluation of the component parts of programs in aging services. For example, outcomes representative of the work of many disciplines should be included in the assessment of the effectiveness of the varied components in a health promotion program. An example of this is diabetes management, which offers a number of services such as nutrition guidance, physical exercise programs and counseling (Rahman & Applebaum, 2010).

The evolution of clinical assessment measures in health and aging has influenced the attention given to clinical health outcomes measures in the new health care reforms act. The PPACA mandates evidence-based changes aimed at improving the quality of health care, and so the Patient-Centered Outcomes Research Institute is now being created (H.R. 3590-625). Among this, Institute’s responsibilities are identifying: national priorities for research with an emphasis on chronic conditions, gaps in evidence in terms of clinical outcomes, and practice variations and health disparities in terms of delivery and outcomes of care. These objectives are very significant for our practice and research with older adults. The question is how should these responsibilities be interpreted if they are going to prove fruitful? First, in examining clinical health outcomes, it is not clear, based on the language of the bill, if, or how, the new Institute will conceptualize interdisciplinary health outcomes research.

Health outcomes have, too frequently, been defined narrowly, focusing on traditional clinical indices, which measure elements within the biological component of the biopsychosocial model. It is time to go beyond traditional physiological data to include psychological, social, and behavioral data in the assessment of patient needs and clinical outcomes. Significant among such research efforts has been the work of the Society of Hospital Medicine’s (SHM) Hartford Discharge Planning Improvement Initiative. SHM’s Project Boost Implementation Guide is used to identify patients with complex discharge needs. Such needs are to be addressed prior to discharge, or patients are considered at high risk for adverse clinical events. The Guide suggests attention to caregiver support issues as well as to language and cultural factors, which are necessary to bolster patient’s understanding and acceptance of services, such as home health care (Williams et al., 2010).

Measurement of outcomes in older adults is frequently, if not always, affected by the complex multidimensionality of their physical, psychological, and social challenges. Measuring outcomes of interventions usually should not be limited to a single factor or dimension. The high rates of cooccurring health and mental health conditions among chronically ill older adults complicate the assessment of outcomes for determining the effectiveness of interventions. As an example, older adults experiencing cognitive impairment may simultaneously confront significant psychological, social, and physiological impairments, and measuring outcomes of treatment activities may require an array of measures across multiple domains during both the assessment process and after the intervention (Berkman & Kaplan, in press). The multidimensional factors that are critical in determining patient care outcomes must be defined and corresponding priorities for research developed. Without the presence of multiple disciplinary practice perspectives, critical biopsychosocial–behavioral factors may be understated or overlooked and the validity of the outcomes compromised.

The other major focus of the Patient-Centered Outcomes Research Institute is to address health disparities. Disease risk factors, as well as responses to treatment, caregiving, and overall quality of life, may be affected by race, ethnicity, gender, and socioeconomic status (Berkman, Silverstone, Simmons, Howe, & Volland, 2000; MacKenzie, 2000; Maramaldi & Cadet, in press; Yee, 1997). As many researchers have demonstrated, issues of health disparities, in terms of both delivery of services and differential patient preferences, need to be “front and center” when examining clinical outcomes. Since 1997, managed care plans administered by Medicare have been reporting quality of care measures utilizing the Health Plan Employer Data and Information Set. In early studies, disparities in health care quality between black and white patients were reported. Data from 1997 to 2003 indicate that Black and White elders enrolled in Medicare-managed care plans showed improvements on a number of clinical performance measures (Trivedi, Zaslavsky, Schneider, & Ayanian, 2005).

And although many are committed to developing culturally sensitive approaches to service, particularly troubling is the lack of culturally sensitive outcome measures, which are necessary to develop evidence-based practices. The heterogeneity and demographic changes in the older adult population are associated with health care and service delivery challenges that demand more effective outcome measures that will deepen our understanding of the interactions between health and culture, such as the effects of cultural beliefs on self-care practices and acceptance of health services (Berkman & Kaplan, in press). Researchers must commit to working toward these needed outcome measures or our data will lack validity. When social–behavioral and cultural factors are included in assessment of patient needs, and appropriate interventions to address these needs are put in place, then clinical patient outcomes are significantly improved. Our assessments and outcomes research must reflect these important dimensions.

In light of current health care reforms, which are significant in design, potentially significant in effects, and probably subject to a number of sizeable alterations in health care legislation and oversight, collaborative research in geriatrics and gerontology on outcome measures is more essential than ever. The Gerontological Society of America can be a significant base for interdisciplinary collaborative action in this arena. To do this, professionals must leave comfortable discipline-specific research silos and seize the opportunity to work together on developing and improving biopsychosocial health care outcome measures and making clinical health research as valid and as relevant as possible.

References

Author notes