| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||
Correspondence: Address correspondence to Joseph Angelelli, PhD, Brown University, Center for Gerontology and Health Care Research, PO Box G-B2, Providence, RI 02912. E-mail: Joseph_Angelelli{at}Brown.edu
 |
Abstract |
|---|
 TOP Abstract Policy BackgroundTheoretical FrameworkMethodsResultsDiscussionReferences |
|---|
Key Words: Rural • Nursing home • Hospital • Strategy • BBA • Post-acute
The Medicare program has contributed to remarkable improvements in the health status of older adults over the past 35 years (Moon, 1999). However, the same provider incentives that have served Medicare beneficiaries so well in terms of increased access to services within particular care settings have often impeded the coordination of care across different settings (Kane, Finch, Blewett, Chen, Burns, & Moskowitz, 1996). Successful coordination of care across the array of acute and long-term care (LTC) services depends in part on the ability of hospitals to form relationships with LTC providers in the community so as to minimize adverse events such as preventable rehospitalizations. Linkages between hospitals and LTC providers are especially critical in many parts of rural America, where the continuum of care exists in concept only (Ricketts, 2000).
The policy developments over the past several years have increased the level of uncertainty in most rural health care environments. The purpose of our analysis is to describe how postacute care and LTC strategies of rural hospitals have changed in the face of this uncertainty and to establish a foundation for examining how such strategy changes have affected the care of Medicare beneficiaries. Our paper focuses on two important strategic questions facing rural hospitals. The initial decision to diversify into postacute care and LTC, and the subsequent decision concerning whether to pursue diversification by means of direct ownership or contractual relationships with others (the decision to integrate vertically or virtually; e.g., Bazzoli, Shortell, Dubbs, Chan, & Kralovec, 1999; Lehrman & Shore, 1998; Robinson & Casalino, 1996; Zajac & D'Aunno, 1994).
|
Policy Background |
|---|
|
TOPAbstractPolicy BackgroundTheoretical FrameworkMethodsResultsDiscussionReferences |
|---|
Coordinated care across settings has been challenged in the years since the passage of the Balanced Budget Act of 1997 (BBA97), as financial incentives for the treatment of certain types of high-cost patients have been altered significantly for hospitals, nursing homes, and home health agencies (General Accounting Office [GAO], 1999a, 1999b). An early analysis of the BBA97 predicted that small rural hospital profit margins would fall from 4.2% in 1998 to -5.6% in 2002 (HCIA-Sachs, 1999). The actual losses experienced by rural providers were even larger than anticipated in the first year of the BBA97 provisions, necessitating refinements to the new reimbursement methodology in 1999 and 2000.
However, only small portions of the BBA97 cuts were returned to rural hospitals (Stensland, Moscovice, & Christianson, 2000), despite the fact that the financial performance of rural hospitals was made worse by a 2.9% reduction in overall Medicare profit margins in 1999 (Medicare Payment Advisory Commission [MedPAC], 2001). The rationale is perhaps related to the fact that rural hospitals fared better overall when compared with urban hospitals in the late 1990s, mostly because private sector pressure and business sector losses occurred more frequently for urban hospitals.
Nonetheless, BBA97 featured several policy changes with the potential to alter rural hospital strategies. The adoption of a prospective payment system (PPS) for Medicare reimbursement to nursing homes in 1999 was perhaps the most sweeping BBA97-related change. The PPS policy had a clear impact on those rural hospitals that diversified into on-site nursing homes. There were 2,173 hospital-based nursing homes operating in 1998. In 2001, the number in business had declined by 20% to 1,762 (MedPAC, 2001).
Concurrent changes in the way home health care agencies (HHAs) are reimbursed by Medicare also contributed to the potential for fragmentation and disruption of postacute care. Overall, Medicare reimbursement to HHAs dropped 40% between 1997 and 1999, from $17.4 billion to $9.3 billion (McCall, Komisar, Petersons, & Moore, 2001). The number of Medicare patients discharged to home health care declined by 23% between 1997 and 1999 (Office of Inspector General, 2000).
Relationships between the acute care, postacute care, and LTC providers are complex and highly sensitive to environmental pressures like the BBA97. The potential for disruption is particularly high in rural areas, where a patchwork of providers exists to meet the health care needs of vulnerable older adults (Moscovice, Wellever, & Stensland, 1999; Trinh & Begun, 1999). Recent case studies of six rural communities suggest rural areas lag behind urban areas in terms of their ability to respond to new BBA97 provisions (Mueller et al., 1999).
Our paper focuses on the early postacute care and LTC strategic responses to BBA97 among rural hospitals. The results of our initial 1997 nationally representative survey of 540 rural hospitals represent a unique baseline from which to gauge changes associated with the BBA97 provisions. Our 97% response rate to the follow-up survey in 2000 attests to the critical importance of the issue among rural hospitals.
|
Theoretical Framework |
|---|
|
TOPAbstractPolicy BackgroundTheoretical FrameworkMethodsResultsDiscussionReferences |
|---|
Swing beds represent a hybrid decision fitting neither the make nor buy categorization. A "swing bed" is defined as a hospital bed that can be used to provide either acute care or LTC. The advantages of swing beds include more efficient utilization of unused rural hospital beds and staff, the ability to cover unmet LTC needs in rural areas without constructing additional nursing homes, and the facilitation of rural hospitals to "become the link between traditional acute and long-term care" (Shaughnessy & Schlenker, 1986).
Rural hospitals can engage in LTC service provision through participation in the swing-bed program without allocating scarce resources to either establish internal LTC units and staff or to monitor an external relationship with a LTC partner. The other strategic options available to rural providers more clearly reflect the internal–external dynamic.
We also use the Miles and Snow (1978) typology of strategic behavior to classify rural hospitals according to their overall postacute care and LTC strategic profile in 1997 and 2000. The typology was originally developed on the basis of chief executive officer (CEO) responses to a 7-point scale reflecting the respondent's perceptions concerning the rate at which his or her organization makes changes in its offerings (products, services, or markets). Each of the four types of "environmental enactment" portrays a relatively distinct pattern of organization–environment interaction.
The four main Miles and Snow strategy types are as follows. Prospectors continuously search for market opportunities and experiment with responses to emerging environmental trends; Analyzers use stable product-market domains but watch competitors closely for new ideas and then rapidly adopt those that appear most promising; Reactors are organizations in which top managers frequently perceive change and uncertainty occurring in their organizational environments but are unable to respond effectively; and Defenders use narrow product-market domains with little searching for new opportunities outside their domains.
For this study, membership in the four strategy types was determined according to each rural hospital's activity in any or all of six major postacute care and LTC market domains: (a) swing-bed program participation, (b) hospital-based nursing home ownership, (c) freestanding nursing home ownership, (d) home health care ownership, (e) formal linkages to external nursing homes by means of transfer agreements, and (f) formal linkages by means of sharing arrangements for facilities, staff, or both.
Actual environmental conditions in rural health care markets play an acknowledged role in shaping the postacute care and LTC strategic orientation of rural hospitals. For example, a rural hospital's options to pursue formal linkages with external nursing homes may be foreclosed because of a lack of viable strategic partners in the community. Similarly, home health care ownership may become more or less attractive in response to the viability of independent HHAs in the market. These environmental conditions may confound the opportunity structure for classification into the Miles and Snow typology, yet our focus on the overall changes in the distribution of facilities in our sample of rural hospitals lessens the importance of explicitly controlling for environmental conditions. We are concerned with examining gross patterns of change (e.g., the number of prospectors in both periods) as an alternative way of classifying overall strategic orientation of rural hospitals (i.e., in addition to specific adoptions or discontinuations).
|
Methods |
|---|
|
TOPAbstractPolicy BackgroundTheoretical FrameworkMethodsResultsDiscussionReferences |
|---|
We compared characteristics of sample facilities with the population of all nonfederal hospitals located in nonurban areas, using the Area Resource File (ARF) data on county characteristics and the POS data on hospital size, type, service offerings, and geographic location. Sampled hospitals were slightly more likely to be local government operated than rural hospitals nationwide (45% vs. 40%), and slightly less likely to be for profit (7% vs. 10%). Sampled hospitals were also slightly more likely to operate home care units (55% vs. 50%). On all other characteristics (region, bed size, medical school affiliation, percent participating in swing beds, and percent part of a corporate health care system), there was no significant difference between the sample and rural hospitals nationwide. Both the large sample size and high response rate to both surveys contribute to the generalizability of the study findings.
The telephone survey of discharge planners took approximately 35–45 min to complete. Detailed information was collected on the types and years of adoption for various LTC strategies, including swing-bed participation, ownership of hospital-based or freestanding nursing homes and HHAs, and various formal and informal relationships with independent LTC providers. These included formal contracts with an LTC provider for the transfer of patients and formal contracts with LTC providers for the sharing of administrative entities such as staff (medical, administrative, or nursing), medical records, board members, lab or management services, physical plant facilities, or fixed equipment. In both interviews we asked whether such formal linkages existed between the rural hospital and its "most used nursing home" and with the rural hospital's "second most often used nursing home."
We classified rural hospitals into the Miles and Snow typology on the basis of the number of their postacute care and LTC strategies. We also considered the timing of each strategy adoption, characterizing strategies adopted in the previous 3 years to be evidence of recent strategic behavior. Rural hospitals were classified on the basis of actual behavior rather than CEO perceptions. The decision to use organizational behavior as a primary organizing principle is supported by other hospital research using the Miles and Snow typology, where the number of diversified services (e.g., ambulatory surgery, home health care, and health promotion) and the number of new services were used to validate perceptions and were found to be highly consistent with the typology (Zajac & Shortell, 1989).
Hospitals were classified as Prospectors if they reported at least four out of the six strategies or if they adopted two or more strategies in the previous 3 years. Hospitals were classified as Analyzers if they had three of six strategies or if they had two strategies, one of which was adopted in the previous 3 years. Reactors were facilities with two strategies, neither of which were adopted in the previous 3 years, and Defenders were classified as such if they had zero or only one postacute care or LTC strategy.
|
Results |
|---|
|
TOPAbstractPolicy BackgroundTheoretical FrameworkMethodsResultsDiscussionReferences |
|---|
Table 1 summarizes specific LTC strategy changes between 1997 and 2000. Proportions of hospitals either adding or dropping LTC strategies in 2000 are displayed by type of strategy. The strategies include participation in the swing-bed program, ownership of hospital-based or freestanding nursing homes, ownership of a HHA, and various forms of external linkages with independent LTC partners. Overall, a considerable amount of strategy change occurred over the brief 3-year period.
|
Of the 168 rural hospitals that did not participate in the swing-bed program in 1997, more than 26% (n = 44) had chosen to do so by 2000. Only 3% of those using swing beds in 1997 had eliminated them by 2000 (12/346 hospitals). In addition, a sizable proportion (28%) of the rural hospitals that owned and operated a hospital-based nursing home in 1997 had divested their units by 2000. However, a comparable number of rural hospitals actually added this strategy by 2000 (n = 55, or 17% of the 317 rural hospitals not operating a hospital-based skilled nursing facility [SNF] in 1997). Similarly, the overall number of hospitals owning freestanding nursing homes increased by 2000.
As expected, a sizable number (n = 56) of rural hospitals had divested ownership of a HHA between 1997 and 2000, whereas relatively few rural hospitals (n = 23) reported new ownership of a HHA in 2000. In the area of patient transfer contracts, 88 rural hospitals reported such a contract with their most-used nursing home partner in 1997. By 2000, 39 of those rural hospitals had terminated their contracts. However, of the 425 hospitals in 1997 lacking a formal patient transfer contract with their most-used nursing home partner, 74 had initiated such a contract by 2000. The pattern of transfer contracts with the second most-used nursing home partner confirms the general trend toward more external LTC linkages. Whereas 19 of the 54 rural hospitals with secondary contracts in 1997 reported dropping their contracts 3 years later, 148 of the 459 hospitals without patient transfer contracts in 1997 had arranged contracts by 2000. A similar pattern is seen with formal sharing arrangements. More formal sharing linkages were added in 2000 than were dropped, especially involving each rural hospital's "second most used" nursing home partner.
Most of the strategic changes reported in 2000 were not related empirically to other strategic decisions. The lone exception is the case of swing-bed participation and hospital-based nursing home operation. Among the 54 rural hospitals closing their hospital-based SNF units by 2000, 11 began participation in the swing-bed program between 1997 and 2000. The association was significant with a chi-square value of 10.74 (p <.05). Conversely, of the 55 rural hospitals that began operating a hospital-based SNF in the study period, 46 were participating in the swing-bed program in 1997 (all continued to do so in 2000) and an additional 3 hospitals had begun participation by 2000.
The decisions of hospitals to begin or discontinue postacute care and LTC strategic options are conditioned on local market characteristics. We examined the postacute market characteristics of subgroups of rural hospitals organized according to strategic behavior between 1997 and 2000. The 54 hospitals discontinuing ownership of a hospital-based SNF between 1997 and 2000 resided in counties that had a significantly higher average proportion of hospitals owning HHAs in 1997 (65%) when compared with the 55 facilities that added a hospital-based SNF (48%). Those facilities that maintained their hospital-based SNF (n = 145) resided in counties with an intermediate average proportion of hospital ownership of HHAs in 1997 (53%). In addition, the 44 hospitals adding swing beds between 1997 and 2000 were in counties with a lower average proportion of hospitals owning HHAs in 1997 (49%) compared with those relatively few (n = 12) that dropped swing-bed participation (78%).
Table 2 summarizes the typology membership in 1997 and 2000. Forty-two rural hospitals were classified as Prospectors in 1997. By 2000, 85 rural hospitals met the criteria for Prospectors. A relatively large number of the facilities classified as Defenders in 1997 (n = 158) had become Prospectors or Analyzers by 2000 (33 and 23, respectively), suggesting a relatively high level of increased strategic development on the part of these previously static rural hospitals. However, there were also facilities that did not aggressively purse postacute care and LTC linkages in the 3 years between surveys, as evidenced by the still sizable number of Defenders in 2000 (127). The 88 hospitals that remained Defenders in both periods resided in counties with the lowest average proportion of hospitals owning HHAs (28%) and nursing homes (15%) in 1997. By 2000, the average proportion of hospital-owned HHAs in the counties of these consistent Defenders was 24%, and nursing home ownership was 13%.
|
|
Discussion |
|---|
|
TOPAbstractPolicy BackgroundTheoretical FrameworkMethodsResultsDiscussionReferences |
|---|
Taken together, our findings illustrate the importance of swing-bed participation as a buffer for rural hospitals challenged by the uncertainty of the BBA97-related reimbursement changes. The swing-bed program continues to represent an attractive option for rural hospitals to care for patients who may previously have received care in a hospital-based SNF, though strategic choice is confined by local market conditions such as the availability of other postacute options. Flexibility of hospitals maintaining swing-bed programs may make it easier for them to initiate and continue the operation of a hospital-based SNF in the post-BBA97 period. Specific types of costly patients may be difficult to care for in hospital-based units under the new PPS reimbursement rules, but participation in the swing-bed program may make it possible to manage these patient types in a cost-effective manner. BBA97-related contraction in the home health market may make such an option a necessity.
The increase in the breadth of formal external linkages in the post-BBA97 period is noteworthy for this reason. Formal arrangements may be more attractive for hospitals that are having difficulty placing patients in postacute settings. For example, anecdotal reports describe how some hospitals develop arrangements with local nursing homes to "sidestep cost issues by selling high-cost products (such as IV drugs) to nursing homes at the hospital's (usually much lower) cost" (Gebhart, 1999). The quid pro quo for such an arrangement is guaranteed nursing home admission for patients from that hospital. It may be that, in the post-BBA97 era, rural hospitals are attempting to link more formally to more than one LTC partner as a way of coping with the uncertainty of the LTC market.
The postacute care and LTC strategies of rural hospitals have the potential to significantly affect the course of postacute care for Medicare beneficiaries. Rural hospital strategies such as swing-bed utilization and formal relationships with nursing homes may reduce the likelihood of potentially harmful "underfunded" extended hospital stays in which financial incentives under the diagnosis-related group system limit the provision of care. Participation in the swing-bed program and the establishment of formal external linkages with LTC providers have real-world implications for hospital discharge planners confronted with persistent pressures to find postacute placement for patients. However, the fundamental importance of strategic linkages relate ultimately to their effect on the well-being of older adults and their families, for whom the hospital discharge process is often chaotic and uncertain—unfolding as it does in a time of personal and family crisis.
The BBA97 may also have affected rural hospital financial performance through its impact on freestanding nursing homes in rural areas and closures of home health care providers. The cumulative effect of freestanding nursing home admission decisions on the financial performance of discharging hospitals is unknown. However, rural hospitals that are part of an integrated delivery system may be able to work in concert with their affiliated freestanding nursing homes to better manage the hospital-to-nursing home transition for Medicare beneficiaries.
The major provision of the Balanced Budget Refinement Act of 1999 (BBRA99) in terms of rural hospital access involved the expansion of the swing-bed program. Participation in the swing-bed program formerly required only that the hospital obtain a certificate-of-need from the state health planning and development agency, and reimbursement was limited to 5 days of LTC per patient per stay. Discharge from an acute care bed and admission to the swing bed is essentially a paper transaction, with no physical movement of the beneficiary. However, discharge to a swing bed makes the patient eligible for an additional per diem reimbursement that the hospital would not otherwise receive were the patient to remain as an acute care patient, thereby increasing the resources for postacute care. The BBRA99 removed the certificate-of-need requirement as well as the length of stay restriction to facilitate wider participation in the program, though the BBRA99 provisions did not take effect until April 2000, after the fielding of our follow-up survey.
Furthermore, in response to arguments that more BBA97 relief was needed, Congress passed the Medicare, Medicaid and SCHIP Benefits Improvement and Protection Act of 2000 (BIPA00). This second BBA refinement legislation lessened the difference between rural and urban hospitals in disproportionate share hospital payment—monies designed to subsidize caring for a disproportionate share of poor Medicare patients. BIPA00 also extends the Medicare Dependent Hospital program for rural areas (though it did not include a decrease in the percent Medicare needed to qualify), and it modifies the payment schedule to include a mileage payment that could increase rural patients' access to emergency and ambulance services.
Given how complex the BBA97, BBRA99, and BIPA00 changes have made the policy environment, there is good reason to expect continued differences in patterns of strategy changes over time. Consistent with earlier research on the effects of the hospital PPS system (Zajac & Shortell, 1989), hospitals appear to have altered their strategic focus in response to the environmental jolt of the BBA97 policies. There remains a need to compare the patterns of strategy changes evident in the early BBA97 period (1997–2000) and a later BBRA99–BIPA00 period (2000–2003). The early period may be characterized by more variation in LTC strategy changes and patterns of multiple strategies than the later period, after BBRA99 changes and earlier LTC market contraction have "settled in."
|
Footnotes |
|---|
We acknowledge Bethany Maher and Gina Ruggieri for their assistance in the preparation of this manuscript.
1 Center for Gerontology and Health Care Research, Brown University Providence, RI.
Laurence G. Branch,, PhD, Decision Editor
Received for publication December 10, 2001. Accepted for publication August 16, 2002.
|
References |
|---|
|
TOPAbstractPolicy BackgroundTheoretical FrameworkMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
