This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation

Hospital Provision of Institutional Long-Term Care

Pattern and Correlates

^a School of Public Health, University of Illinois at Chicago
^b Department of Sociology, University of Illinois at Chicago
^c School of Public Health, University of Michigan

Correspondence: Naoko Muramatsu, PhD, School of Public Health, University of Illinois at Chicago, 2035 West Taylor Street, Chicago, IL 60612-7259. E-mail: naoko{at}uic.edu.

Decision Editor: Laurence G. Branch, PhD

	Abstract

TOP Abstract Background Conceptual Framework Methods Results Discussion References

 
Purpose: This study examined the pattern and correlates of institutional long-term care provision among U.S. community hospitals, differentiating two categories of services: (1) skilled nursing and rehabilitation (SN-R) and (2) other long-term care (O-LTC). Design and methods: Multinomial logistic regression analysis was used to examine the associations of hospital and community characteristics with the pattern of long-term care provision (SN-R only, O-LTC only, both SN-R and O-LTC, and None) among 3,842 hospitals. Results: The pattern of long-term care provision was significantly associated with hospitals' mission (for-profit and teaching status) and their internal and external resources. Implications: Results suggest the importance of considering hospital and community characteristics in predicting the impact of policy changes and in envisioning the role of hospitals in long-term care.

Key Words: Skilled nursing care • Rehabilitation services • Acute care • Continuum of care • Organizational study

Hospitals in the United States have significantly diversified into long-term care, partly in response to changes in Medicare reimbursement policy since the early 1980s and market competition in hospitals' core business of acute inpatient care (Paone 1996; Paone and Jack 1993; Robinson 1994; Shortell et al. 1986). Hospitals' provision of long-term care provides opportunities for much needed coordination of acute and long-term care (Campion, Bang, and May 1983). On the other hand, it presents challenges to the delivery and financing of services for older Americans. According to a study of a random sample of Medicare beneficiaries in 1995 (Office of Inspector General 1997), patients discharged to the hospital's own nursing facility reported better continuity of care than those discharged to a free-standing nursing facility; however, they also had shorter hospital stays and longer nursing home stays. This finding suggests that hospitals may be shifting cost from the prospective payment system (PPS) in inpatient acute care to a cost-based system in post-acute care. The PPS for Medicare inpatient benefits, which provides incentives to shorten hospital length of stay, has also prompted hospitals and nursing care facilities to develop so-called "subacute care" (Banaszak-Holl, Zinn, Brannon, Castle, & Mor, 1997; Kane et al. 1996; Lewin-VHI, Inc., 1995) to accommodate increasing levels of institutional care for people discharged from hospitals. The possible cost-shifting and the rapid development of subacute care have contributed to the recent growth in Medicare post-acute care expenditures (Scanlon 1997). These opportunities and challenges at the intersection of acute and long-term care require better understanding of the provision of institutional long-term care among hospitals, the hub of acute care.

The purpose of this study was to examine the pattern and correlates of the provision of institutional long-term care services among U.S. community hospitals as of 1995. Apart from previous research, we differentiated skilled nursing and rehabilitation services (SN-R) and other long-term care (O-LTC), which are subject to different technical and institutional environments. Our results indicate whether and how hospital provision of SN-R and O-LTC varies by hospital mission and internal and external resources. Our findings provide the basis for policy makers to take into consideration hospital and community characteristics in predicting the impact of current policy changes on hospital provision of long-term care and in planning the role of hospitals in the spectrum of care needed for older and disabled Americans.

	Background

TOP Abstract Background Conceptual Framework Methods Results Discussion References

 
Hospitals may provide a wide variety of long-term care, ranging from skilled nursing and rehabilitation services to less intensive long-term care. Skilled nursing care involves medical and skilled nursing care, therapy, and social services under the supervision of a licensed registered nurse on a 24-hour basis. Physical rehabilitation services encompass a comprehensive array of restoration services for the disabled and all support services necessary to help patients attain their maximum functional capacity. Other long-term care includes intermediate nursing care for people with physical or functional disabilities who do not require a hospital or skilled nursing facility care, but need supervision and support services as well as residential care, such as assisted living and retirement housing (American Hospital Association 1996a). Although there is no obvious way to categorize these partly overlapping services, SN-R services differ from O-LTC in two important ways. First, SN-R services involve an intensive level of nursing for restorative or rehabilitative purposes that O-LTC services do not provide. Second, SN-R, when used by Medicare beneficiaries who demonstrate the need for the intensity of such services, is reimbursed by Medicare. On the other hand, O-LTC is not reimbursed by Medicare, unless it is provided to beneficiaries who also have skilled nursing or rehabilitation needs as defined by Medicare. Payment for O-LTC comes from the patient's pocket, Medicaid, or, in exceptional cases, long-term insurance (Wiener and Illston 1996).

Differentiation between SN-R and O-LTC is important, because these categories of services require different sets of skills and technology and are subject to different financial and institutional environments. As such, hospitals may adopt different strategies to provide these different services. For the purpose of this study, we use "long-term care" as an umbrella term for SN-R and O-LTC, and so-called subacute care is considered as part of SN-R. Although SN-R and O-LTC can be provided in the home setting, our study focuses on services provided in the institutional setting so that we can discuss them in relation to their specific technological and financial/institutional environments that are different from those of home care.

There are several economic and social incentives for hospitals to provide institutional long-term care (Blewett, Kane and Finch 1995–1996; Wheeler, Burkhardt, Alexander, and Magnus 1999). Economically, providing long-term care may help hospitals to (1) secure post-acute beds to shorten acute inpatient length of stay; (2) shift the cost from the prospective payment system of acute inpatient services to a cost-based reimbursement of skilled nursing and rehabilitation services; and (3) reduce direct competition in the acute inpatient market, when demand for this core hospital business is declining. Socially, hospitals may provide long-term care to promote social causes that are valued by the community (e.g., improving access to needed post-acute medical and social services). Social and economic benefits, of course, are related. Hospitals providing long-term care for economic reasons may produce socially valued outputs; and hospitals motivated by social objectives may use their public image as a marketing tool to attain economic objectives. In the current reimbursement system, the provision of SN-R is likely to contribute to both the economic and social objectives of hospitals, whereas the limited profitability of O-LTC may make the provision of services in that area a more socially rather than economically motivated activity. To the extent this is true, hospitals that emphasize both social and economic objectives would be likely to provide both SN-R and O-LTC, while those that are oriented toward economic goals would be likely to provide only SN-R.

Research indicates that the provision of institutional long-term care services varies among hospitals and that such variation is strongly associated with hospital and local community characteristics (Shortell et al. 1986; Wheeler et al. 1999). However, prior research is limited in two ways. First, it tends to bundle all types of long-term care into one broad category. The lack of differentiation between SN-R and O-LTC has prevented us from understanding (1) the extent to which hospitals have ventured into not only SN-R but also O-LTC, and (2) whether and how the pattern of the provision of these two categories of services depends on hospital and community characteristics. Second, existing literature defines the provision of long-term care based on the ownership of long-term care beds, disregarding other important modes of service provision. For example, hospitals may provide long-term care through their health system or network, or through formal contractual arrangements or joint ventures with other providers. These modes of provision may be effective alternatives for hospitals to diversify their services and to avoid unnecessary duplication of services, especially if similar services are provided elsewhere in their health systems or neighboring health care organizations. Excluding them would underestimate hospitals' role in the provision of long-term care.

	Conceptual Framework

TOP Abstract Background Conceptual Framework Methods Results Discussion References

 
Provision of long-term care is conceptualized as a strategy for hospitals to position themselves in their market in order to achieve their goals and assure their survival (Clement 1988; Clement, D'Aunno and Poyzer 1993). We adopted the strategic management perspective to investigate the pattern and correlates of hospital provision of long-term care. The strategic management perspective is derived from theoretical and empirical research in organizational studies and in economic and health services literatures. It assumes in its core that organizations are rational actors; their behavior is determined by their strategic goals and is adaptive to not only their internal resources but also constraints and opportunities in their environments (Child 1972; Mahoney and Pandian 1992; Shoemaker 1993; Shortell and Zajac 1990). Applying this perspective to hospitals, we contend that hospitals' provision of long-term care is likely to be correlated with their organizational mission or goals and that the extent of provision will be a function of their internal as well as external resources or opportunities. In the following sections, we discuss our hypotheses on the overall pattern of hospital provision of institutional long-term care, followed by our hypothesis on how such pattern is associated with the organizational and environmental factors of hospitals.

Overall Pattern of Hospital Provision of Institutional Long-Term Care
Organizations most often adopt strategies that do not require a significant departure from the organizations' current values, norms, practices, policies, and mix of abilities and skills (Bowman and Hurry 1993). Following the notion of "strategic comfort zone" (Shortell and Zajac 1990), we contend that, all things being equal, hospitals would be more likely to provide SN-R than O-LTC. SN-R is closer to hospitals' traditional core competence in acute care and therefore requires less divergence from hospitals' existing organizational structure and resources. Two environmental factors also compel hospitals to secure and control post-acute beds for patients requiring skilled nursing and rehabilitation services after hospitalization. First, the inpatient PPS, which was originally established for Medicare patients and has now spread into many other insurance programs, provides hospitals a strong financial incentive to reduce inpatient length of stays. Second, the rapid growth of managed care organizations has also pressured hospitals to reduce utilization of costly inpatient acute services. Thus, we expect that, in general, hospitals are more likely to provide SN-R than O-LTC.

Once hospitals master the provision of SN-R, they may expand their activities into O-LTC, which, by then, may have fallen into their strategic comfort zone. By the same token, hospitals would rarely venture into O-LTC without first providing SN-R. Thus, we expect that hospitals are more likely to provide "both SN-R and O-LTC" than "O-LTC only."

Organizational and Environmental Correlates of Hospital Provision of Institutional Long-Term Care
Organizational Mission.
Organizational mission or goals direct and constrain organizational decision making and action (Scott 1998). Hospital ownership type (investor-owned vs nonprofit) is generally considered a major indicator of organizational mission (Hollingsworth and Hollingsworth 1987; Lee and Weisbrod 1977; Weisbrod 1998). It determines what the hospital wants to accomplish and how. Investor-owned (IO) hospitals are profit-maximizers. As a primary objective, they have to satisfy the expectations of their investors or stockholders by maximizing the returns on investment and achieve long-term growth in earnings per share. Nonprofit hospitals are likely to have a broader mission, mainly to satisfy the needs of various groups of the local community, sometimes regardless of profitability. Local boards of trustees or public officials scrutinize nonprofit hospitals for the consistency of their strategies with their mission (Hollingsworth and Hollingsworth 1987; Weisbrod 1998). Given the requirement to satisfy various constituencies and the scrutiny by community groups, we expect nonprofit hospitals, whether private or public, to provide a wider range of services including both SN-R and O-LTC. By contrast, IO hospitals, in the absence of local scrutiny, could narrowly focus on profit-making services. Because SN-R is considered more profitable than O-LTC (Sabatino 1991), IO hospitals are more likely to provide SN-R but less likely to offer O-LTC.

Another important indicator of organizational mission in hospitals is teaching status. The core mission of teaching hospitals includes high-tech treatment of complicated acute cases and training of physicians and other health professionals. Teaching hospitals may be less likely to provide long-term care services than nonteaching hospitals, because these services are outside the "strategic comfort zone" of their core business of tertiary acute care. Furthermore, in the current medical education and research system that values cutting-edge, high-tech programs, long-term care may have a small chance of competing successfully with more valued activities for internal resources in teaching hospitals.

Internal Resources.
An organizational strategy has little chance of success without sufficient resources for implementation. In addition, slack resources—especially unused financial, human, and physical resources—reinforce the organizations' decision to embark on a new investment (Child 1972; Shoemaker 1993). Thus, more resources are likely to enhance hospitals' ability to diversify into long-term care services, other things being equal.

In this study, we use organizational scale, occupancy rate, cash flow per asset, interorganizational linkages, and patient mix to measure the variety of hospitals' internal resources. Larger hospitals, compared to their smaller counterparts, tend to have more slack resources to explore and invest in unconventional service areas (Alexander and Amburgey 1987; Levinthal 1990). Thus, we expect larger hospitals to have a higher chance of providing both SN-R and O-LTC. Similarly, hospitals with higher occupancy and cash flow per asset—both indicators of superior performance—may have more financial resources to invest in SN-R and O-LTC.

Research has suggested that interorganizational linkages with multihospital systems and hospital networks may improve hospitals' access to financial, technical, personnel, and information resources that are critical for the provision of long-term care (Dranove and Shanley 1995; Shortell and Zajac 1990; Succi, Lee, and Alexander 1997). For example, given that the long-term care market is highly regulated and surrounded by rapidly changing environments, developing and sustaining long-term care services requires specialized knowledge and information. Hospitals embedded in relationships with relevant organizations may have a better chance of accessing key information and knowledge. Thus, we expect linkages with multihospital systems and hospital networks to facilitate hospitals' provision of long-term care.

Patient mix determines hospitals' revenue sources. As Medicare and Medicaid fees are set at lower levels than the fees for private pay and private insurance patients, hospitals with a higher proportion of Medicare and Medicaid patients are expected to receive less revenues for the same set of services, and thus have less resources available to diversify into long-term care (Dranove and White 1998). On the other hand, evidence suggests that hospitals where Medicare represents a larger proportion of revenue are more likely to discharge patients to post-acute care covered by Medicare (i.e., skilled nursing services and rehab services) in order to shift the cost of inpatient services (Blewett et al. 1995–1996). Thus, we expect that Medicare dependence may increase the probability of providing SN-R services while reducing the probability of providing O-LTC not covered by Medicare.

External Conditions.
Every organization exists in a specific physical, technological, regulatory, and social environment to which it must adapt (Scott 1998). Because hospitals mainly serve the local community, they are especially sensitive to their socioeconomic conditions and health care resources. These environmental factors may constrain funds and demand for services that are required for hospitals to initiate and sustain long-term care services. Six variables—rural location, percentage of elderly population, unemployment rate, availability of hospital beds, availability of nursing home beds, and physician supply are used in this study to assess the relationships between environmental conditions and hospital provision of long-term care.

In rural communities that generally lack a well-developed health and social service infrastructure, hospitals constitute a critical source of services. To satisfy the emerging needs of an aging population, Federal health policies have encouraged provision of long-term care in rural hospitals. For example, as a result of federal legislation implemented in 1982, small rural hospitals are eligible to have "swing beds" that can swing between acute and skilled nursing care use according to changing patient needs (Richardson and Kovner 1987; Shaughnessy and Schlenker 1986). Thus, we expect hospitals in rural communities to be more likely than their urban counterparts to provide long-term care.

To the extent that hospital services are determined by rational assessment of the community's health care needs, we expect hospitals located in communities with a higher proportion of elderly people to be more likely to diversify into long-term care. Although the local community's health needs might promote hospital diversification, a depressed local economy, represented by a higher unemployment rate, may reduce residents' ability to pay for heath services, especially long-term care services that are not covered by health insurance (Nelson, Thompson, and Bland 1998). A poor local economy also diminishes hospital revenues (either patient or nonpatient revenues such as donations) and may hinder the development of long-term care structure in the community. As such, we anticipate a negative relationship between a high unemployment rate and hospitals' provision of long-term care.

Health care resources in the community, such as the availability of hospitals, nursing homes, and physicians, may significantly affect whether hospitals provide long-term care. We expect a high level of hospital bed supply in the community to increase the focal hospitals' provision of long-term care, because such diversification may provide hospitals with opportunities to establish a viable market niche and thereby reduce direct competitive pressure from surrounding hospitals in the acute care market (Succi et al. 1997). The presence of nursing homes in the community can work in two, and possibly contradictory, ways in affecting hospital provision of SN-R and O-LTC services. First, nursing homes can compete with hospitals for the provision of both SN-R and O-LTC services, thus reducing the incentive of hospitals to move into the long-term care market. On the other hand, a strong presence of nursing homes that provide SN-R and/or O-LTC may provide opportunities for affiliations or joint ventures, making it easier for hospitals to move into the long-term care market. We offer no definite hypothesis on the relationship between the presence of nursing homes in the community and hospitals' provision of long-term care, and leave it for empirical examination. To the extent that current U.S. physicians focus on acute care services and their decisions determine the amount and the type of services provided to patients (Cassel 1985; Gonzalez and Rizzo 1991; Javalgi, Joseph, Gombeski, and Lester 1993), we expect that a high level of physician supply in the community would lead to a strong acute care orientation in the local care market, discouraging hospital provision of long-term care

	Methods

TOP Abstract Background Conceptual Framework Methods Results Discussion References

 
Study Sample and Data
Data for this study were obtained from four sources. The first was the 1995 American Hospital Association (AHA) Hospital Survey, which provided most of the hospital-level variables. The survey, considered the best source of information on American hospitals, was administered to all hospitals in the United States and the response rate was over 90%. Our analysis focused on community general hospitals, or nonfederal short-term general hospitals whose facilities and services were available to the public. The second source of data was the 1996 Bureau of Health Professions Area Resource File (ARF), a compilation of longitudinal data pertaining to all U.S. counties. This database provided information on social, demographic, and health care characteristics of the county where each hospital was located. The third source of data was the American Hospital Association Guide to the Health Care Field (Part B: Multihospital System) for coding multihospital system membership (American Hospital Association 1996b). Fourth, the Health Care Financing Administration (HCFA) hospital cost report in 1994, which contains financial and statistical records submitted by Medicare-certified hospitals, provided the information to construct the ratio of cash flow to total assets, an indicator of financial performance. The 3,842 community general hospitals with complete information on studied variables constituted the sample for this study.

Dependent Variables
We created a dependent variable involving four categories of long-term care provision: (1) SN-R only, (2) O-LTC only, (3) both SN-R and O-LTC, and (4) none of SN-R or O-LTC. SN-R was assessed by the hospital's self-report of provision of institutional skilled nursing care and/or physical rehabilitation as defined by the AHA Hospital Survey. O-LTC was assessed by whether the hospital provided one or more of the following institutional services: intermediate care, long-term care other than skilled or intermediate care, and residential services, such as assisted living and retirement housing. Given our focus on institutional long-term care often used by elders, we did not include services designated for people with special acute and long-term care needs, such as drug abuse/dependency and psychiatric care. The 1995 AHA survey did not specifically ask about swing beds, which can be used for both acute and long-term care services in rural hospitals in the Federal swing-bed program (Richardson and Kovner 1987; Shaughnessy and Schlenker 1986). However, we assumed that swing beds were captured in the survey questions on various types of long-term care.

Independent Variables
For hospital ownership, we used IO hospitals as the reference category and created two dummy variables, private nonprofit and public, to assess the difference between nonprofit and IO hospitals. Teaching status was measured by a dummy variable indicating membership on the Council of Teaching Hospitals, the Association of American Medical Colleges. Hospital size was measured by the average number of hospital beds including cribs and pediatric bassinets regularly maintained for inpatients during the one-year reporting period. Interorganizational linkages were measured by two variables, indicating membership in a multihospital system (MHS) or a network. An MHS is a corporate body that operates two or more hospitals (Alexander 1990). A network is a group of hospitals, physicians, other providers, insurers, and/or community agencies that work together to coordinate and deliver a broad spectrum of services to their community (American Hospital Association 1996a). Medicare and Medicaid dependency was measured by the ratio of Medicare/Medicaid inpatient days to the total number of hospital inpatient days during the reporting period. Occupancy was the ratio of average daily census to the total number of hospital beds set up and staffed. The ratio of cash flow (net income plus depreciation) to total assets indicated the amount of cash flows generated per dollar of the hospital's assets. It measures the financial return; a greater value indicates that the hospital is more effective in earning a higher return on its assets and has more slack resources for realizing its strategies (Wheeler et al. 1999).

Measures of community characteristics were assessed at the county level. We used the metropolitan/nonmetropolitan dichotomy to define rurality. If the hospital was located in a nonmetropolitan county, we designated it as rural. The percentage of elderly people was the number of residents aged 65 or older divided by the total county population. The unemployment rate, a proxy for local economy, was the ratio of the unemployed to the civilian labor force. Presence of other hospitals and nursing homes was represented by two variables: the number of community hospital beds (excluding those of the focal hospital) and nursing home beds (skilled nursing beds and other long-term care beds) per 10,000 population. Physician supply was measured by the number of active physicians per 10,000 population.

Analysis
Multinomial logistic regression analysis was used to examine the associations of hospital and community characteristics with the four categories of institutional long-term care provided by hospitals: "SN-R only," "O-LTC only," "both SN-R and O-LTC," and "none" (neither SN-R nor O-LTC). We used the model for "none" as the reference to obtain three nonredundant sets of parameter estimates. The first set compared the likelihood of "SN-R only" with that of "none"; the second set "O-LTC only" with "none"; and the last "both SN-R and O-LTC" with "none." In order to present the results of all the possible contrasts (6 nonredundant combinations), we reestimated the models two additional times using "SN-R only" and "O-LTC only" as references.

	Results

TOP Abstract Background Conceptual Framework Methods Results Discussion References

 
Pattern of Hospital-Based Institutional Long-Term Care
The majority of community hospitals (65%) provided at least one type of institutional long-term care. As shown in Table 1 , the pattern of hospital-based institutional long-term care was consistent with our expectations that hospitals were more likely to provide services close to their core, acute-care domain. Hospitals that provided "SN-R only" accounted for 36% of the study hospitals, followed by those with "both SN-R and O-LTC" (23%) and those with "O-LTC only" (7%). Hospitals were more likely to provide SN-R (59%) than O-LTC (30%).

In general, hospital size, occupancy, and interorganizational linkages were positively associated with hospital provision of long-term care, supporting our expectation that hospitals' internal resources promote long-term care provision, especially "both SN-R and O-LTC." Conversely, higher proportions of Medicare and Medicaid patients were associated with a lower likelihood of providing long-term care, with a notable exception of the positive association of Medicare dependency with "SN-R only." The ratio of cashflow to assets was not statistically associated with the dependent variable.

Independent of hospital characteristics, several community factors were significantly associated with hospital provision of long-term care. Consistent with our expectations, rural hospitals and hospitals located in a community with a lower unemployment rate were more likely to provide long-term care across the board. Interestingly, a higher proportion of the elderly population was negatively related to hospital provision of "SN-R only" and "both SN-R and O-LTC." The supply of hospital beds was positively associated with the provision of at least one type of long-term care and, most strikingly, with the provision of "both SN-R and O-LTC." The supply of nursing home beds was a weak correlate for the probability of "SN-R only" and "O-LTC only," but was strongly associated with the provision of "both SN-R and O-LTC." A higher level of physician supply in the community was negatively associated with the provision of long-term care, in general, and with "O-LTC only" and "both SN-R and O-LTC," in particular.

	Discussion

TOP Abstract Background Conceptual Framework Methods Results Discussion References

 
Consistent with our hypotheses, results show that variables indicating hospitals' organizational mission, internal resources, and external conditions were strongly associated with the pattern of hospital provision of institutional long-term care. These findings suggest significant organizational and geographic variation in the role of hospitals in long-term care. Our study also extended previous research by differentiating two types of services, skilled nursing and rehabilitation services (SN-R) and other long-term care (O-LTC). This differentiation provided several insights into specific aspects of hospital provision of institutional long-term care.

Specifically, results supported our contention based on the strategic management perspective that most hospitals were unlikely to venture far away from their core technology (acute care) and tended to remain in their "strategic comfort zone" (Shortell and Zajac 1990). More than half of all general community hospitals provided SN-R, but less than one third provided O-LTC in 1995. Furthermore, our study revealed important findings missed in previous research that did not differentiate types of institutional long-term care. For example, Wheeler and colleagues 1999 concluded that IO hospitals provided a lower level of institutional nonacute care, including both SN-R and O-LTC. Our study showed that the differences between IO and nonprofit hospitals varied by the type of services provided. Compared to their nonprofit counterparts, IO hospitals tended to provide "SN-R only" and refrained from venturing into other less intensive long-term care. This finding is consistent with the observation that IO hospitals are swift to mobilize resources to lucrative markets (e.g., subacute care) but tend to avoid services that are unprofitable yet essential for meeting the health needs of the community (Shortell et al. 1986; Wheeler et al. 1999). In contrast, nonprofit hospitals—private or public—displayed a more even-handed pattern of long-term care provision, suggesting that they are better positioned to play a facilitative role of bridging the two types of long-term care services than IO hospitals.

Several other findings warrant further discussion. Teaching hospitals' lack of involvement in post-acute care was expected, given their core mission of teaching and research in cutting-edge medical technologies. Nonetheless, the finding reiterates the acute-care orientation of current medical education and the concern that training in teaching hospitals may not adequately prepare medical professionals to meet the growing needs of older, chronic, and disabled populations. Given the significant role of teaching hospitals in training and socializing physicians, nurses, and other health professionals, it is important that teaching hospitals expand training opportunities in long-term care.

Among the community factors, rural location had a pronounced association with hospitals' provision of long-term care. This underscores the importance of hospitals in rural health care. Declined economies, population emigration, and insufficient hospital reimbursement rates have threatened the survival of rural hospitals (American Hospital Association 1991; Succi et al. 1997). Our findings suggest that closures of rural hospitals are likely to result in the loss of both acute and long-term care services in the community. Although this result seems inconsistent with the findings of Wheeler and associates 1999, who reported a lower level of institutional nonacute care provision in rural hospitals, the difference may be due to their inclusion of swing beds, which exist only in rural hospitals, as a covariate in their model.

A high supply of hospital and nursing home beds in the community appeared to promote the provision of both skilled nursing/rehabilitation services and other long-term care. This result suggests that, in order to have a competitive edge, hospitals might be compelled to not only provide skilled nursing and rehabilitation services but also venture into other, less intensive long-term care. This finding also suggests that the availability of nursing home beds may enhance hospitals' ability to provide long-term care, possibly through contractual arrangements. A greater supply of physicians, on the other hand, was negatively related to hospital provision of long-term care. Although no causality can be established in our cross-sectional analysis, the findings suggest a lack of physician participation in long-term care and point to the necessity of educating (and re-educating) physicians regarding the changing nature of health care and older adults' health care needs. Further research is needed to elucidate why a greater supply of physicians is associated with lower hospital provision of long-term care.

Similar to Wheeler and colleagues 1999 research, we found little relevance of the community's proportion of elderly residents with hospitals' provision of long-term care. This result held even in an additional analysis that excluded the availability of nursing home beds, which was likely to correlate highly with the percentage of elders in the community (results available from the first author). To the extent that the proportion of older adults in the community represents the potential need and demand for long-term care, our finding raises a serious concern that, rather than responding to the health needs of the community, hospitals' provision of services is based mainly on the considerations of market opportunities, financial solvency, competitive pressure from neighboring hospitals, and alternative priorities of physicians.

Future Research
Hospitals' role in delivering a wide range of services to manage health of the community is an important area for research (Shortell 1999). Our study suggests several directions for future research. First, the cross-sectional nature of our study limits our ability to examine the process and dynamics of hospital provision of long-term care (e.g., a hospital's entering and exiting the skilled nursing and rehabilitation services markets). Although Medicare, with a strong focus on acute care, has never been intended to cover long-term care, it has greatly influenced, and will continue to influence, the delivery of long-term care. Under the Balanced Budget Act of 1997 (BBA97), Medicare reimbursement for skilled nursing and rehabilitation services is undergoing a series of changes. For example, the PPS for skilled nursing facilities is being phased in over three years. Hospital-based facilities that emphasize nursing services are likely to face disproportionately reduced reimbursement as compared to free-standing nursing facilities that focus on rehabilitation, particularly in rural areas (Angelelli 1999). In addition, for the 10 diagnostic related groups (DRGs) selected based on a high volume of post-acute discharges, discharges from hospitals to post-acute care are now considered "transfers," rather than "discharges" that can qualify for full DRG payments. This "transfer rule" may decrease hospitals' financial incentive to discharge patients immaturely to skilled nursing or rehabilitation settings and reduce hospitals' motivation to secure discharge outlets. Our research suggests that the impact of these changes is likely to be uneven, depending on hospitals' mission and internal/external resources. For example, with declining economic incentives to provide skilled nursing and rehabilitation services, hospital provision of long-term care may decrease, especially among IO hospitals that are motivated by economic incentives. Furthermore, even if hospitals are motivated to provide a wide range of services for the community, hospitals with limited resources may find it more difficult to continue to provide long-term care. A longitudinal study involving data collected before and after the passage of BBA97 will help us understand how health policy changes may interact with organizational and environmental conditions to produce the dynamics of hospital provision of various long-term care services.

Second, hospital decisions regarding long-term care provision may involve two stages: first, decisions on whether to provide certain types of services; and, second, decisions on how to provide services—internally (making their own services) or externally (providing services through their systems or entering contractual agreements with other organizations). Given our primary interest in examining the overall picture of hospital provision of long-term care, this study focused on the first part of the decision making. Compared to external provision, internal provision allows hospitals to exert more control over finance, organization, technology, and quality in providing long-term care services; however, it requires significant organizational investment and commitment, which some hospitals, especially those in financial distress, may not afford. External provision, by contrast, may represent an efficient alternative strategy for hospitals to satisfy their patients' long-term care needs and diversify their services. The diminishing economic benefits of owning SN-R may lead hospitals to divest from SN-R or to attain their social objectives of providing long-term care through alternative ways. Future research incorporating alternative ways to provide services would be able to identify various strategies appropriate for different types of hospitals and communities to provide and coordinate acute and long-term care services needed for the aging population.

Third, because our study focused solely on hospitals, it did not address the role of hospitals in institutional long-term care in relation to other long-term care facilities. It is unclear whether hospitals and nursing care facilities will continue to provide overlapping services or which is taking the lead in the provision and coordination of acute and long-term institutional care in the future. What we know is that, currently, the presence of hospitals in long-term care (nursing homes' core technology) exceeds that of nursing homes in acute care (hospitals' core technology). As of 1995, 13% of all the nursing facilities in the United States are hospital-based (American Health Care Association 1998). On the other hand, there are few hospitals, if any, that are nursing home based. Further research using both hospital and nursing home data would be needed to understand the current roles of hospitals and nursing homes in long-term care. Moreover, our analysis was constrained by the American Hospital Association data and could not differentiate hospitals providing only skilled nursing home health services from those also providing custodial type home care. Given that home- and community-based services constitute an essential part of long-term care and are increasingly important links between acute and long-term care, hospital provision of home- and community-based services warrants future investigations.

Fourth, new organizational forms (e.g., integrated health systems, health care networks) are quickly emerging in the health care sector. Our results suggested that membership in multihospital systems or networks may be key in promoting hospital provision of both skilled nursing/rehabilitation services and other long-term care. Future research needs to consider how those organizational forms may coordinate or incorporate traditional acute and long-term care providers, including hospitals and nursing homes. In addition, to the extent integrated health systems, rather than individual hospitals, make decisions regarding entry into and exit from post-acute care services, analysis incorporating system-level characteristics is required. Research on system characteristics that explains hospitals' commitment to, and the coordination of member hospitals in, the provision of long-term care is also needed.

Fifth, we chose a strategic management perspective and assumed rational behaviors. This perspective, however, is one of many theories that can explain hospital behavior (Scott 1998). Use of alternative theoretical frameworks may reveal questions not addressed in our study (Shortell 1999). For example, given high levels of regulations and uncertainties in the long-term care market, hospitals may mimic other providers and follow norms and regulations without rational deliberation. This possibility may be better understood by institutional theory (Scott 1995).

Finally, hospital provision of long-term care does not necessarily mean organizational and clinical integration of services. Future research needs to examine whether and how hospitals' long-term care provision impacts various dimensions of patient outcomes, especially access to long-term care and health status. A study using the data of 2,572 patients and 52 hospitals in 1988–1989 indicated that hospitals' characteristics, including direct provision of institutional and in-home skilled-nursing and rehabilitation services, did not have strong impact on patients' use of post-acute services, controlling for patient characteristics (Blewett et al. 1995). More research that incorporates longitudinal patient-level data linked with facility-level data is needed to examine the recent and future impacts of hospital provision of long-term care on patients.

Conclusion
Despite uncertainties and complexities associated with changes in the health care systems and policies, one thing seems certain: Coordination and integration of acute and long-term care will be an increasingly important issue for health care providers, policy makers, and patients. With the older population experiencing multiple chronic health problems, it is becoming more important to recognize that acute and long-term care constitutes but one episode of care. As key acute care institutions and important long-term care providers, hospitals will continue to be an essential part in the continuum of care, although whether they will assume the leadership role in coordinating various services in the community remains to be seen. Some have argued against the idea, fearing that giving hospitals a greater role in post-acute care may "medicalize" those services and increase their cost. Others contend that hospitals' provision of post-acute care does not guarantee smooth transitions between acute and long-term care. Past debate on the role of hospitals in long-term care tended to ignore the heterogeneity of hospitals and their surrounding environments. Our finding that hospital provision of long-term care varies across hospitals and communities suggests that hospitals' role in service integration may depend upon hospitals' organizational and environmental characteristics. For example, larger, nonteaching, nonprofit hospitals and hospitals in rural communities may have a larger role in providing and coordinating acute and long-term care.

Integration of acute and long-term care has been a long-standing issue in gerontological research and practice. There have been results from various demonstration programs (e.g., social HMOs, On Lok/PACE) that attempt to financially integrate acute and long-term care in limited geographical areas (Wiener and Skaggs 1995). Instituting experimental programs on a large scale will require information about the contexts and settings in real situations, in particular, where existing resources (services and skills) are available. Studies based on national data on health service organizations like our current study, on the other hand, describe and explain the organization of acute and long-term care and its geographic variations in "natural settings" under the existing reimbursement arrangements. Synthesizing findings from these two types of research will benefit efforts to integrate acute and long-term care for the U.S. aging population.

	Acknowledgments

 
The authors thank Richard T. Campbell, Susan Hughes, Marshall Chin, and Thomas Prohaska for their helpful comments, and Melissa Succi for her assistance with data acquisition. This research was partially supported by Suntory Cultural Foundation, Japan.

Received for publication September 16, 1999. Accepted for publication May 23, 2000.

	References

TOP Abstract Background Conceptual Framework Methods Results Discussion References

 

• Alexander J. A., 1990. Diversification behavior of multihospital systems: Patterns of change, 1983–1985. Hospital & Health Services Administration 35:83-102.
• Alexander J. A., Amburgey T. L., 1987. The dynamics of change in the American hospital industry: Transformation or selection?. Medical Care Review 44:279-321. [Medline]
• American Health Care Association1998. The nursing facility sourcebook, 1998 Author, Washington, DC.
• American Hospital Association1991. 1990 closure report: A statistical profile Author, Chicago, IL.
• American Hospital Association1996a. Annual survey of hospitals database: Documentation for 1995 data Author, Chicago, IL.
• American Hospital Association1996b. American Hospital Association guide to the health care field Author, Chicago, IL.
• Angelelli, J. (1999). Evaluating SNF prospective payment: A descriptive foundation. Poster presentation at the 16th Annual Meeting of Association of Health Services Research, Chicago, IL.
• Banaszalk-Holl J., Zinn J. S., Brannon D., Castle N. G., Mor V., 1997. Specialization and diversification in the nursing home industry. Health Care Management 3:91-99.
• Blewett L. A., Kane R. L., Finch M., 1995. Hospital ownership of post-acute care: Does it increase access to post-acute care services?. Inquiry 32:457-467.
• Bowman E. H., Hurry D., 1993. Strategy through the option lens: An integrated view of resource investments and the incremental-choice process. Academy of Management Review 18:760-782.
• Campion E., Bang A., May M. I., 1983. Why acute-care hospitals must undertake long-term care. New England Journal of Medicine 308:71-75. [Medline]
• Cassel C. K., 1985, 564. Doctors and allocation decisions: A new role in the new Medicare. Journal of Health Politics Policy: (and Law) 10-549.
• Child J., 1972. Organizational structure, environment and performance: The role of strategic choice. Sociology 6:1-22. [Abstract]
• Clement J. P., 1988. Corporate diversification: Expectations and outcomes. Health Care Management Review 13:7-13.
• Clement J. P., D'Aunno T., Poyzer L. M., 1993. The financial performance of diversified hospital subsidiaries. Health Services Research 27:741-763. [Medline]
• Dranove D., Shanley M., 1995. Cost reductions or reputation enhancement as motives for mergers: The logic of multihospital systems. Strategic Management Journal 16:55-74.
• Dranove D., White W. D., 1998. Medicaid-dependent hospitals and their patients: How have they fared?. Health Services Research 3:163-185.
• Gonzalez M. L., Rizzo J. A., 1991. Physician referrals and the medical market place. Medical Care 29:1017-1027. [Medline]
• Hollingsworth J. R., Hollingsworth E. J., 1987. Controversy about American hospitals: Funding, ownership, and performance American Enterprise Institute for Public Policy Research, Washington, DC.
• Javalgi R., Joseph W. B., Gombeski W. R., Jr. Lester J. A., 1993. How physicians make referrals. Journal of Health Care Market 13:6-17.
• Kane R. L., Finch M., Blewett L., Chen Q., Burns R., Moskowitz M., 1996. Use of post-hospital care by Medicare patients. Journal of the American Geriatrics Society 44:242-250. [Medline]
• Lee A. J., Weisbrod B. A., 1977. Collective goods and the voluntary sector: The case of the hospital industry. Weisbrod B. A., , ed.The voluntary nonprofit sector. Lexington, MA: D.C. Heath
• Levinthal D. A., 1990. Organizational adaptation, environmental selection, and random walk. Singh J. V., , ed.Organizational evolution: New directions. Newbury Park, CA: Sage
• Lewin-VHI, Inc. (1995). Subacute care: Policy synthesis and market area analysis. Report submitted to the Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation, Washington, DC.
• Mahoney J. T., Pandian J. R., 1992. The resource-based view within the conversation of strategic management. Strategic Management Journal 13:363-380.
• Nelson D. E., Thompson B. L., Bland S. D., 1998. Cost as a barrier to medical care in relation to unemployment rates. New England Journal of Medicine 339:1644-1645. [Free Full Text]
• Office of Inspector General1997. Medicare hospital discharge planning Department of Health and Human Services (OEI-02-94-00320), Washington, DC.
• Paone D. L., 1996. Hospitals. Evashwick C. J., , ed.The continuum of long-term care: An integrated systems approach. Albany, NY: Delmar
• Paone D. L., Jack C. M., 1993. Hospital owned and operated nursing facility report American Hospital Association, Chicago, IL.
• Retherford R. D., Choe M. K., 1993. Statistical models for causal analysis John Wiley & Sons, New York.
• Richardson H., Kovner A. R., 1987. Swing-beds: Current experience and future directions. Health Affairs 6:61-74. [Medline]
• Robinson J. C., 1994. The changing boundaries of the American hospital. The Milbank Quarterly 72:259-275. [Medline]
• Sabatino F., 1991. Home health, diagnostic centers were financial winners in 1990. Hospitals 65:27-29.
• Scanlon W., 1997Testimony before the Committee on Finance, U.S. Senate, April 9, 1997. United States General Accounting Office Available: http://www.goa.govHealth Care Financing and Systems Issues, Health, Education, and Human Services Division. Statement of William J. Scanlon Director:
• Scott W. R., 1995. Institutions and organizations Sage Publications, Thousand Oaks, CA.
• Scott W. R., 1998. Organizations: Rational, natural, and open systems 4th ed. Prentice Hall, Upper Saddle River, NJ.
• Shaughnessy P. W., Schlenker R. E., 1986. Hospital swing-bed care in the United States. Health Services Research 21:477-498. [Medline]
• Shoemaker P. J. H., 1993. Strategic decisions in organizations: Rational and behavioral views. Journal of Management Studies 30:107-129.
• Shortell S. M., 1999. The need for interdisciplinary-based theory. Health Services Research 34:1-5.
• Shortell S. M., Morrison E. M., Hughes S. L., Friedman B., Coverdill J., Berg L., 1986. The effects of hospital ownership on nontraditional services. Health Affairs 5: (Winter) 96-111. [Medline]
• Shortell S. M., Zajac E. J., 1990. Health care organizations and the development of the strategic-management perspective. Mick S. S., Associates , ed.Innovations in health care delivery. San Francisco: Jossey-Bass
• Succi M. J., Lee S. D., Alexander J. A., 1997. Effects of market position and competition on rural hospital closures. Health Services Research 31:679-699. [Medline]
• Weisbrod B. A., 1998. Instititional form and organizational behavior. Powell W. W., Clemens E. S., , ed.Private action and the public good. New Haven, CT: Yale University Press
• Wheeler J. R. C., Burkhardt J., Alexander J. A., Magnus S. A., 1999. Financial and organizational determinants of hospital diversification into subacute care. Health Services Research 34:61-81. [Medline]
• Wiener J., Illston L. H., 1996. The financing and organization of health care for older Americans. Binstock R. H., George L. K., , ed.Handbook of aging and the social sciences 4th ed. 427-445. Academic Press, San Diego, CA.
• Wiener J., Skaggs J., 1995. Current approaches to integrating acute and long-term care financing and services American Association of Retired Persons, Washington, DC.

This article has been cited by other articles:

				N. G. Castle Innovation in Nursing Homes: Which Facilities Are the Early Adopters? Gerontologist, April 1, 2001; 41(2): 161 - 172. [Abstract] [Full Text]

This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation